Final Report Summary - SYDDARTA (SYstem for Digitization and Diagnosis in ART Applications)
Executive Summary:
Cultural heritage monitoring comprehends a sum of equipments, technologies and protocols to achieve the most detailed information about how the time eras and the environment contribute to deteriorate, change or weaken the artworks. Nowadays, the rapidness, accuracy and versatility of monitoring techniques become critical in estimating an asset state and determining the future durability of an artwork, depending on the materials of which it is composed. Detecting molecular changes due to complex processes as represent ageing and environmental climate effects is indispensable for the purpose of a proper restoration and conservation. The main objective of SYDDARTA is to develop a methodology and corresponding prototype for monitoring the state and deterioration of works of art, and in particular, paintings from the Baroque period. This has been achieved by using combined 3D - hyperspectral imaging, in a noninvasive and automatic way, where hyperspectral images are acquired whilst simultaneously scanning the 3D profile of the object. The proposed new combination of 3D scanning techniques and hyperspectral imaging will widely reduce and increase inspection accuracy, while reducing time and costs in the characterization of art assets, speeding up the analysis procedures and minimizing the impact of manipulating and transporting the artwork. The project has implemented a successful structure of work packages distributed in two areas of scientific-technical and demo activities: on one side, heritage study, cultural experimentation and investigation on deterioration procedures and degradation effects on paintings, and on the other, the development of a complete novel prototype which implements the procedures of data capturing, 4D representation generation corresponding to the artworks, data treatment and data processing and analysis, to infer maps of physical surface deterioration and chemical distribution (representing either substance identification, mixtures, or mixture states, such as a unaffected or degraded compounds). The work developed included, the production of historic physical models of paintings on canvas and woods, some of which were aged and degraded by UV radiation, relative humidity and temperature cycle treatment, and pollution exposure. In addition a collection of real painting form the Baroque period was selected for further in-depth understanding of the deterioration of the art assets due to climate and environmental conditions. The technical developments focused on the production of a prototype, consisting of two optical channels for combined 3D digitisaton and hyperspectral imaging data capturing. The system worked in both spectral ranges (400-750nm and 900-2500nm). It permitted the registration of 3D and hyperspectral data in the form of images, which later on were processed to produce 3D point clouds and hypercubes which were directly related to the object’s surface and chemical composition. The data would be ultimately processed in an automatic way to help understand the current state of the art work, and any alterations or degradation suffered. 3D point cloud data was analysed to obtain degradation maps, by means tri-dimensional plane local approximations. Hyperspectral imaging was used to produce automatic classification by means of nonlinear supervised modeling, with the implementation of support vector machines, which helped in the identification and quantification of substances, mixtures and deterioration and degradation types, as defined by the user. The resulting system incorporated some of the most advanced technology for optical treatment: digital light projectors, based on digital micromirror devices, acousto-optic tunable filters, liquid crystal tunable filters and high-end focal plan array sensors working in the visible and short wave infrared. A novel optical, mechanical and electronic integration was also required for the completion of this transportable prototype. The SYDDARTA SYSTEM was tested on the real paintings selection with successful results. Interesting scientific and technical foregrounds were also identified at project completion, which will establish the bases of future commercialization among the heritage community and other areas of industrial activity. Finally a complete set of actions for project content and results dissemination were carried out by means of a dedicated webpage, publications in scientific and technical congress proceedings, the implementation of dedicated workshops in different EU countries and the production of a project biannual news letter reflected in the EU Heritage Portal.
Project Context and Objectives:
According to the EGMUS - European Group on Museum Statistics there are, approximately, 21,000 institutions across Europe that exhibit permanent or temporal art collections and act as a key role for the heritage sustainability by joining the importance of culture dissemination and the tourism as an exploitation market. For these organizations, cultural heritage monitoring comprehends an essential
routine and requires employing a sum of equipments, technologies and protocols to achieve the most detailed information about how the past of time and the environment contribute to deteriorate, change or weaken the artworks. Moreover, the accuracy of monitoring techniques nowadays becomes more important at the time that the durability of an artwork depends on the materials of which it is composed and detecting molecular changes due to complex processes as represent ageing and environmental climate effects is indispensable for the purpose of a proper restoration and conservation. In front of this, restorers and collection owners are working and facing a truth constrained by four aspects:
1. Handling restrictions: the artwork transport could mean important risks due to unknown data about the initial state and stability.
2. Several steps in the characterization: the complete measurement of deterioration parameters requires the application of some different techniques separately with a direct consequence in time consumption and wider data analysis.
3. Bad cost-benefit ratio: the use and maintenance of these equipments represent a high expenditure/investment respecting the data achieved in the time spent, and that produces a poor exploitation of these techniques and a gap in the artwork study.
4. Imprecision due to extrapolation: the study of whole artworks is difficult to perform since the current routine takes random or limited samples.
The main objective of SYDDARTA is to develop a prototype instrument for monitoring the deterioration of works of art and other heritage artefacts. This will be achieved by using “3D
Hyperspectral imaging”; where hyperspectral images are acquired whilst simultaneously scanning the 3D profile of the object. Being optical, the technique will be non-invasive and therefore non-destructive. The proposed new combination of 3D scanning techniques and hyperspectral imaging could widely reduce and increase inspection accuracy, while reducing time and costs in the characterization of art assets, speeding up the analysis procedures and minimizing the impact of manipulating and transportation on the artwork. To successfully develop such prototype, a series of technical and project operational considerations must be stated, as follows:
• 3D digitization. The new system will include 3D imaging capabilities by means of a digital light project (DLP), by means of Digital Micromirror Devices (DMDs) which would project light-on-light-off patterns on the objet for 3D digital surface reconstruction.
• Hyperspectral imaging in the visible and near-IR ranges. Many of the pigments considered for some historical periods can be properly distinguished in the visible and near-infrared part of the spectrum.
The prototype will be sensitive to the range 400nm to 2500nm.
• Types of evaluated cultural assets: Paintings on canvas and wooden panels. Within the Baroque period, the methodology development will focus on painting on canvas and on wooden panels. The prototype will produce a hyperspectral-3D model of each of the assets analyzed, for deterioration information extraction.
• Deterioration evaluation due to environmental conditions. Affection by temperature, relative humidity, UV radiation and pollutants deposition such as Carbon, Calcium Carbonate and dust, will be considered within the SYDDARTA project. Models for technical evaluation will be produced following receipts from the historical time selected. The choice of these two environmental parameters is done because it is the most important for painting degradation
Project Results:
SUBSTANCE DATABASE AND SPECTROSCOPIC INSPECTION KNOWLEDGE
General overview
A database of substance spectral behaviour (in ranges from 400nm to 2500nm by means of single point diffuse reflectance, dispersive spectroscopy) has been create in the project, with information on historic recipes substances prepared in the project. The database is accompanied with FTIR and Raman information and on the samples analysed (documentation). This database was produced by means of the middle demonstrator and single point spectroscopic systems).. The analysis included degraded and not degraded samples.
Foreground contents and specification
Inspection equipment used to generate this database: PerkinElmer 1500.
Spectral configuration of the database: spectral range covered (400-2500nm). Spectra resolution: 1 nm. Spectral curve form: Diffuse reflectance.
Number of spectral samples included in the database: 160.
Sample distribution: 40 spectra corresponding to the canvas reference sampled, 40 spectra corresponding to the UV-exposed canvas sample, 40 spectra corresponding to the RH-T treated canvas sample, 40 spectra corresponding to the polluted canvas sample.
ST Foreground format: digital document and txt. files.
SYDDARTA PROTOTYPE
General Overview
The SYDDARTA prototype is the main project foreground. It is a single piece of equipment, for the combined hyperspectral and 3D digitization of art assets, in particular paintings - for this first version. The novel system is particularly suitable for noninvasive inspections tasks of objects. The system is suitable for capturing data regarding surface status and spectral reflectance. Internally the system is capable of generating the 3D point clouds of the object under inspection and storing the spectral reflectance data as hyperspectral images of the covered field of view. The system can be directly used for the assessment of the state of an art asset. This state can be correlated with the degree of deterioration the asset may present. Object alterations not related to environmental and climate influence may also be assessed with the SYDDARTA prototype. The prototype implements a novel optical inspection arrangement, and a collection of processing algorithms for the purpose of evaluating the structure, surface and chemical composition of the object under study.
Foreground contents
The prototype consists of two physical units: (i) an optical module and (ii) power and control unit. The units can be controlled via a standard PC. There are also two software modules associated to the prototype: (a) a control application, which is used to control the hardware and enables data capturing and (b) a processing application, which is used to visualise and process the data captured with the control application. The two applications are run independently. The applications run on Windows OS environment.
The optical module includes two inspection channels, from which all 3D and hyperspectral data is captured. The two channels are named (1) the 3D imaging channel (3DI-channel) and the (2) shortwave infrared hyperspectral imaging channel (SHI-channel). The purpose of the first channel is to capture hyperspectral imaging data in the visible spectral range (400-750nm), and also 3D fringe pattern images to construct the 3D point clouds later on. The aim of the second channel is to capture hyperspectral imaging in short wave infrared range (900-2500nm). The whole system can be controlled
The processing application is used to pre-treat the data captured by the control application. The principal pre-treatment processes are:
For the hyperspectral data: data normalisation (via white reference and embedded multiplicative scattering correction algorithms), channel data merging (affine transformation to c-register the two image streams from the two channels), and dark current noise suppression for the SHI-channel hyperspectral data only.
For the 3D digitization data: 3D point cloud generation in real xyz coordinates.
The processing application has a collection of built in procedures and algorithms for data analysis and results generation:
For the hyperspectral pre-treated data: modeling environment for future substance and spectral occurrence automatic classification based on the following steps: a. data sampling from syddarta generated raw information, b. model training based on support vector machines, based on linear polynomial and radial basis function kernels, and c. model automatic classification and display of class maps. Sampled and modeled substances can be pigments, binders, supports, or else mixtures of substances, as they are normally presented in a painting. Spectral samples for model construction may also correspond to complex combinations which may present some alteration, or degradation, hence modeling the identification of such alteration on a single asset, for quantification purposes, or else on a collection of assets, for identification of occurrence, to say but a few applications.
For the 3D pre-treated data: 3D data processing by means of a multi-scale approach in which 3D data is segmented in areas of a given radius (the scale) and planes are approximated to them. The collection of locally approximating planes define a corresponding collection of normal vectors with a given direction in the 3D space, which then are compared with a standard planar reference. Deviations from the standard determine the non-flatness at a local level. For assets which are supposed to be flat, such as paintings, this deviation can be considered a definition of physical surface degradation.
Both control and processing applications have a GUI which allow user operation and data manipulation. In particular, de processing application presents 7 interfaces for general capturing control, system check and control and configuration for the different active components (such as the AOTF, DLP, light source, VIS and SWIR cameras and the LCTF). The processing application GUI implements a working desktop where data can be managed (load, save, display bands, display 3D point cloud, zoom in, rotate, scroll) and also visualize spectral curves on mouse movement on the hyperspectral image display. There is a panel on the left of the main GUI for modeling purposes, so the user can i. buil the spectral databases of interest by sampling on the displayed image and automatic spectrum recordin, ii. Train SVM models and, iii. Classify images hyperspectral images by means of the generated models, as many models as the user may require. There is also a small panel on the right of the interface for 3D data degradation map generation, based on the mutli-scale planar approach, embedded in the application.
The software modules are completed with two available functionalities: a. 4D stitching, to produce large digital representations of single objects which, due to their size, have been register in different captures, and b. the calculation of a confidence index for hyperspectral data classification.
Technical features:
Spectral ranges covered: 400-750nm and 900-2500nm.
Spectral resolution obtained on each range: nominal 20nm (VIS), nominal 5nm (SWIR).
Spectral capture sequence configuration:
Spectral configuration - VIS:
Bandwidth configuration: N.A.
Step size configuration: minimum of 1nm
Spectral configuration - SWIR:
Bandwidth configuration: configurable by means of RF generation signals in the control application. Recommended configuration: 10nm, 20nm and 40nm bandwidth.
Setp size configuration: minimum of 1nm
Field of view: VIS 40x40cm2; SWIR 6x6cm2; concentric.
Working distance: nominal 50 cm.
Object minimum measurable reflectance: 5% (below this the system captures noise, specially in the SHI-channel).
Communications and signals (LCTF (USB hyperterminal), VIS camera (GigE), SWIR camera (USB), AOTF (USB), DLP (USB), Light (DC signal).
TUNABLE SWIR ILLUMINATION MODULE
The SWIR illumination unit, as integrated in the prototype with minor optical adjustment at the output, can be considered an S/T foreground, as it has been specifically constructed within SYDDARTA and for the purpose of the prototype construction. This illumination unit allows the wavelength selection of the projection beam. In this way, only radiation of interest reaches the object to be inspected. This illumination source can be used to implement fast hyperspectral scanning arrangements. Other applications may be illumination of samples for photomicrography.
FOREGROUND CONTENTS
The illumination source is made up of a halogen tungsten bulb (30 W and a specifically design reflector to optimise the light bulb emittance and convey the radiation to the AOTF input, and two optical stages (a collimating lens and a SWIR accepting bandpass filter). The following stage consists of a linear grid polarizer in the SWIR region which orientates the oscillation plane of the electromagnetic radiation.
Radiation enters an acousto-optic tunable filter, with an acceptance angle of 3 degrees and an optical window of 1x1 cm2. The AOTF is made up of a birefringent crystal (TeO2) which produces a double wavefront in the crystal structure, each front with a different polarisation (ordinary and extraordinary). The acout-optic effect produces the extraordinary polarisation to be diffracted from the main optical path, with an angle of 7 degrees. The radiation outputting the AOTF is then split into two paths, one with undesired radiation and the other the patch driven to the object for single band illumination. The 0-order cancellation (i.e. unwanted, non-monochromatised light) is cancelled by means of a Rochon prism a the AOTF output and a physically generated beam dump in the output optical path.
Technical features
Spectral ranges covered: 900-2500nm.
Spectral capture sequence configuration:
Spectral configuration: Bandwidth configuration: configurable by means of RF generation signals in the control application. Recommended configuration: minimum of 5nm, 10nm, 20nm and 40nm bandwidth.
Setp size configuration: minimum of 1nm
Field of view at 50 cm: SWIR 6x6cm2.
Optimal working distance: nominal 50 cm.
Communications and signals (AOTF (USB), Light (DC signal).
Power needs: 220AC (converted to 24V DC internally).
Specific casing and output optical configuration for stand-alone operation not developed in the project.
THE PROCESSING APPLICATION
The processing application is a stand-alone piece of software which can be run independently from the SYDDARTA prototype. The application has been developed to work with standard image formats (such as jpg). The application implements a number of pre-treatment features, particularised for the SYDDARTA control application raw data. This functionality can be bypassed and the application is capable of loading standard hyperspectral imaging format .bsp. Likewise, embedded pretreatment can be either enabled or disabled, in relation to white reference compensation, normalisation and dark current removal. The application structure is configured in such a way that it can either load the fringe pattern images - and generate the 3D point cloud - or load a vrml and XYZ 3D data file formats. The application is conceived as a working bench to manually inspect 4D data representations of objects (by displaying non processed 3D point clouds or hypercubes) or apply processing techniques for identification of both 3D patterns and hyperspectral trends. The features of the processing motor are the usage of SVM modeling techniques with different kernels (linear, polynomial and radial basis functions) to produce classifiers for automatic spectral fingerprint identification and local deviations of a 3D surface form a planar standard.
FOREGROUND CONTENTS:
The two main sets of processing algorithms are implemented in this application:
3D data processing algorithms for the identification of non-planar surface areas at different scales
Hyperspectral automatic modeling for substance, mixture and alterations identification.
The application consists of a single GUI interface which allows the user: i. to manage 4D data (load, save, select) form a pane on the left part of the window, ii. visualise and display data (3D point cloud and hyperspectral imaging) on a central image control, iii. visualise spectral curves from the hyperspectral image in real time on a central lower pane, iv. spectral digital sample, model and classify with built in functionality from a left positioned panel, and v. apply the degradation estimation functionality on a left bottom panel. Displaying of 3D degradation maps, classification maps and confidence maps is one in the main image display control (central to the GUI).
Technical features
Windows OS application
Programming language C++
File acceptance: JPG, TIFF, .bsq, .txt, .vrml
File maximum size: 1GB
SECONDARY FOREGROUND
As part of the development of the prototype, different components were produced and adjusted by the technical partners, which resulted in variations of already existing commercial products. This is the case of the following optical components, integrated as part of the prototype:
Second foreground - DLP projector:
Based on the DLP technology by Texas Instruments, the DLP produced in the project was improved with respect to the already existing version by the implementation of a HD DMD chip (1080x1920 pixels) with a high power LED in a luminescent reflector which produces a constant continuous emmittance between 420 to beyond 700nm. Implementation of electronics, projecting optics and encapsulation where considered a novelty and hence a secondary foreground.
Second foreground - SWIR camera:
Based on previous camera models commercialized, the MCT SWIR camera unit integrated in the prototype was specially design to optmise quantum efficiency. In fact, this parameter was almost kept constant throughput the range of interest, according to manufacturing measurements. The camera incorporated also an improved procedure for image adjustment (compensation and removal of dark and white pixels, typical in MCT sensor after manufacturing).
Potential Impact:
SYDDARTA POTENTIAL MARKET
The main output of the project is the achievement of a pre-industrial prototype with highly innovative features (product family 1) that will have a modular character. This means that the equipment may be understood as a whole system but also the features may be separate into specific devices (product family 2 and 3) and a complete spectral database (product family 4) in order to estimate market sales.
As a consequence of the previous analysis, the consortium has defined two markets: the direct one composed by cultural heritage end-users and the secondary one composed by other sectors to explore.
I. Principal market: cultural heritage end-users
According to the EGMUS - European Group on Museum Statistics there are, approximately, 21,000 institutions across Europe that exhibit permanent or temporal art collections and act as a key role for the heritage sustainability by joining the importance of culture dissemination and the tourism as an exploitation market.
The European Fine Art Foundation (TEFAF) has released its annual report on the international art and antiques market. The study for 2013, carried out by Arts Economics , used data gathered from a range of sources: art dealers, auction houses, collectors, fairs, financial databases, industry experts and many others involved in the international trade of art. Although the report focused its attention on the US and Chinese art markets in particular, this year's headline finding was that the global art market reached €47.4bn (£39.7bn) in total sales of art and antiques in 2013, close to its highest ever recorded total and advancing 8% year-on-year.
The project has applicability for different types of entities, all in relation to the arts and heritage in two principal ways.
- Conservation
- Authentication
The first impact of SYDDARTA is to contribute to overall preventive strategy to maintain and secure the conservation and management of movable assets in public and private cultural heritage collections, regarding "natural" deterioration processes and human interventions.
CONSERVATION
Developed around two parallel lines of work:
1. Preventive Conservation: risk analysis and implementation of appropriate measures for storage, handling, transport and exhibition of art works, to ensure their proper preservation. Preventive conservation is now an essential element of any policy on heritage conservation, including;
-Preventive measures: optimal storage conditions, storage, handling, packaging, transport, exposure and maintenance works, taking into account parameters such as room temperature, relative humidity, radiation, visible light, ultraviolet radiation, vibration, shocks, disasters and plagues, etc..
-Action protocols in case the risk materializes.
2. Study of techniques and materials of construction, as well as causes and degradation of the latter, in order to establish the necessary conservation measures and, where appropriate, consider the most appropriate treatment or interventions.
Each work must be comprehensively analyzed to identify technical and physical characteristics, to establish its state of conservation and develop the best strategy capable of ensuring its preservation. In parallel, some have set a specific program storage, handling, transport, exposure and maintenance for each work.
In SYDDARTA, IPCHS and RABASF are end-users representing heritage and collection owners which have the need to incorporate better cost-benefit conservation equipment and ensure the sustainability exploitation of the artworks.
Moreover, the business plan contemplates the option of renting the prototype in terms of providing a service to the cultural heritage sector in Europe. Due to the fact that some end-users will require the use of SYDDARTA in a concrete period or for a specific study of an art work, the consortium has contemplated the possibility of renting the SYDDARTA prototype for different institutions and entities. This option will give an opportunity to avail from the SYDDARTA technology without having to invest in equipment. This option is particularly recommended for medium and small preservation centers, museums, and research institutions.
Taking into account the specific characteristics of the system and the importance of the objects to be inspected by SYDDARTA, the consortium will create a group of technicians formed by representatives of project participants that will elaborate a prototype manual for technical use.
AUTHENTICATION
The expertise and authentication play an increasingly important role in the art market. The evaluation of the price of an art object (or collection) depends primarily on the degree of authenticity. There are still many identification and allocation tables errors, with financial and legal consequences. Scientific techniques are becoming increasingly sophisticated, and go further in the exploration of materials.
Many players can make the cost among them: sellers, buyers, companies Sale Volunteers (SVV), research laboratories, restaurant, insurance companies, lawyers and judges, historians of and finally the art experts.
Determining authenticity is the mission of the expert. The art expert, with experience gained over many years, is a specialist capable of determining the nature, origin and time of manufacture of the object subject to judgment. He is asked to authenticate, but also to detect alterations, conversions and repairs that may be incurred by this and indicate a range of values to be used for trading on a sale or estimate, for example, an insurance company.
The expert operates in two business areas:
1. Activity authentication and estimate for the certificates insurance
2. Activity expertise and estimation in the context of public sales
It is often faced with many difficulties: firstly, it is technically easy to produce fake. The result of expertise then depends on the importance of research that the expert will complete. Finally, the state of the art and knowledge are constantly changing, challenging entrenched beliefs.
The European Fine Art Foundation (TEFAF) has released its annual report on the international art and antiques market. The study for 2013, carried out by Arts Economics , used data gathered from a range of sources: art dealers, auction houses, collectors, fairs, financial databases, industry experts and many others involved in the international trade of art.
II. Secondary markets
SYDDARTA system can be applied into different sectors other than conservation and restoration of cultural heritage. Depending on the specific field where it is going to be applied the system will need a specific adaptation to be implemented in a concrete market, due to the fact that each industry requires different functionalities and operates with special compounds and substances.
Therefore the system must implement a specific database for the concrete sector where it is going to be applied; this means that following the market sector where project will be used, it will be created a database for the characterisation and composition of the substances that compounds the object to be inspected. In addition it will be elaborated different algorithms specifics for the software development in order to better detect and identify the presence of the interesting components and provide a quantitative estimation of them.
SYDDARTA system will obtain the physic-chemical and 3D information of the objects and will inspect them. The system will be implemented in the following sectors:
• Agro-food – Under this market, the system will be able to detect the presence of harmful chemicals components, bacteria, virus, organic and inorganic contamination, food deformities, food good conditions (humidity, temperature, ripening, decay, etc.) and to distinguish the composition and volumetric estimation of food.
• Biomedical – Under this market, the system will be able to detect the characterisation and study of human body movements by analyzing the mechanics involved on them. The system will help on the detection of 3D and spectral information related to the human body behavior in a non-invasive way. The system can detect and provide: biometric estimations, size parameters, presence of oxygen, hemoglobin, etc.
• Pharmaceutical – Under this market, the system will be able to help on the process for medicine manufacturing, estimation of active components and excipients.
• Part manufacturing industry – Under this market, the system will be used for the inspection of part dimensionality as well as overall structural and chemical quality.
Based on the above considerations, the following includes an estimation of the Syddarta foreseen source of income per product family (see Business Plan D7.7):
Commercialization product family 1: Prototype Full Equipped
- Expected sales: First 2 years 15 units; First 5 years 100 units
Commercialization product family 2 : Extended spectral range imaging and 3D digitization applications for art heritage assets and other manufacturing industries
- Expected sales: First 2 years 15 units; First 5 years 80 units
Commercialization product family 3 : Hyperspectral and 3D digitization applications for art heritage management industry and research
- Expected sales: First 2 years 15 units; First 5 years 80 units
Commercialization product family 4 : Spectral database for the characterization of art assets
- Expected sales: First 2 years 30 units; First 5 years 120 units
EXPLOITATION OF RESULTS
Depending on the nature of the entity, the way of exploiting results can be quite different, oriented more or less directly to immediate profitability.
Non-profit entities
The mission of a non-profit entity is to investigate, to transfer knowledge in education, to transfer knowledge to economy through spin-offs and to disseminate knowledge through publications and congress contributions.
It is committed to scientific research and the highest international academic standards. Due to its non-profit nature, its potential exploitation activities mainly fall into the spheres of education, research and knowledge transfer.
In a particular way, the orientation for each entity will be the following:
• AIDO - Interest in patenting and exploitation of results through licensing to third parties.
• FORTH - Interest in patenting and exploitation of results through licensing to third parties.
• TUD - Interest in patenting and exploitation of results through licensing to third parties.
• ISAC - Interest in patenting and exploitation of results through licensing to third parties.
• IPCHS - Interest in patenting and exploitation of results through licensing to third parties.
• RABASF – Interest in patenting and exploitation of results through licensing to third parties.
For all of them, there is an interest in using the system once it has passed the end of the project, to study and produce scientific and technical documentation in order to make more publications and disseminate the
Companies and profit entities
After completion of the project, the business interest lies in the rapid exploitation of results, for maximum profitability. There are two main ways to achieve this profitability: intellectual or industrial protection, lease, or sale of property; immediate sale or license, without protection.
In a particular way, the orientation for each entity will be the following:
• SIGNINUM – Interest in patenting and use of the resulting product for generating inspection services business.
• AVANTES – Interest in patenting and production of lighting components (reflector and light source SWIR) on the line developed in the project, either as part of the prototype SYDDARTA either as standalone product marketing.
• G&H – Interest in patenting and production of lighting components (ATOF developed SYDDARTA)) on the line developed in the project, either as part of the prototype SYDDARTA either as standalone product marketing.
• VIALUX – Interest in patenting and production of lighting components (DLP projector) on the line developed in the project, either as part of the prototype SYDDARTA either as standalone product marketing.
• XENICS - Interest in patenting and production of sensing components (SWIR camera developed for the project) on the line developed in the project, either as part of the prototype SYDDARTA either as standalone product marketing.
DISSEMINATION ACTIVITIES
The Project Consortium has developed different activities focused in the dissemination of results, reported in the Periodic Report 1 and 2, classified as follows;
Press Releases in Web sites/Applications:
AIDO; https://arquivo.pt/wayback/20191218154828/http://www.syddarta.eu/ AIDO'S Newsletter, Twitter of Otri grupo, Facebook of Fedit (Federation of Spanish Technological Centers), Web Page and Newsletter of Fedit (Federation of Spanish Technological Centers), Diario de Salamanca, Agencia SINC, http://www.zvkds.si/sl/varstvo-kulturne-dediscine/raziskave-in-projekti/
Articles published in the popular press:
EL MUNDO Newspaper (Spain), LEVANTE Newspaper (Spain).
Presentations, Oral presentations to a scientific event and Exhibitions
TUD: NWO Getty Panel Paintings Initiative Meeting, Optical Metrology Techniques for Diagnostics of Artwork, Fotonica exhibition, International Conference of the Ageing of Materials & Structure
RABASF: Information to cultural institutions
SIGNINUM; Jornadas ARP; Universidade Católica Porto, Conference Heritage Colors: Faculdade Ciência e Tecnologia Lisboa, SPIE Photonics Europe; Syddarta Portuguese Tour
AIDO: Syddarta stand in the eighth edition of AR&PA Biennial, under the title "Innovation in Cultural Heritage"
VIALUX MESSTECHNIK + BILDVERARBEITUNG GMBH Symposium on Emerging and Industrial DLP Applications, SPIE Photonics West, Laser World of Photonics, 8th International Symposium on Emerging on Emerging and Industrial Texas Instruments DLP Technology Applications
CONSIGLIO NAZIONALE DELLE RICERCHE; Ambiente e patrimonio culturale: degrado, previsioni e strategie di conservazione. XIV CONGRESSO NAZIONALE DI CHIMICA DELL'AMBIENTE E DEI BENI CULTURALI; VIII Congresso Nazionale di Acheoetria. Participation in the round table “Prospettive e Strumenti per il finanziamento della ricerca applicata al Patrimonio Culturale”;
AIDO: VIII Congresso Nazionale di Archeometria. Poster Novel 3D hyperspectral methodology to estimate deterioration in paintings: Syddarta. Syddarta workshop: Novel 3D-hyperspectral methodology to estimate deterioration in paintings (Valencia, Spain)
FOUNDATION FOR RESEARCH AND TECHNOLOGY HELLAS: Syddarta: Development and Applications of novel technologies in Cultural Heritage
Publications;
SYDDARTA: new methodology for digitization of deterioration estimation in paintings. Proc. SPIE 8790, Optics for Arts, Architecture, and Archaeology IV, 879011
Deterioration estimation of paintings by means of combined 3D and hyperspectral data analysis. Proc. SPIE 8790, Optics for Arts, Architecture, and Archaeology IV, 879011
“Combined 3D-hyperspectral Analysis for art Asset Deterioration Assessment – SYDDARTA Project “
“A Aplicação do Sistema Syddarta na Conservação Preventiva”
List of Websites:
Project website address: https://arquivo.pt/wayback/20191218154828/http://www.syddarta.eu/
Contact details;
E-mail 1: ribes@aido.es
E-mail 2: eribes@aido.es
Cultural heritage monitoring comprehends a sum of equipments, technologies and protocols to achieve the most detailed information about how the time eras and the environment contribute to deteriorate, change or weaken the artworks. Nowadays, the rapidness, accuracy and versatility of monitoring techniques become critical in estimating an asset state and determining the future durability of an artwork, depending on the materials of which it is composed. Detecting molecular changes due to complex processes as represent ageing and environmental climate effects is indispensable for the purpose of a proper restoration and conservation. The main objective of SYDDARTA is to develop a methodology and corresponding prototype for monitoring the state and deterioration of works of art, and in particular, paintings from the Baroque period. This has been achieved by using combined 3D - hyperspectral imaging, in a noninvasive and automatic way, where hyperspectral images are acquired whilst simultaneously scanning the 3D profile of the object. The proposed new combination of 3D scanning techniques and hyperspectral imaging will widely reduce and increase inspection accuracy, while reducing time and costs in the characterization of art assets, speeding up the analysis procedures and minimizing the impact of manipulating and transporting the artwork. The project has implemented a successful structure of work packages distributed in two areas of scientific-technical and demo activities: on one side, heritage study, cultural experimentation and investigation on deterioration procedures and degradation effects on paintings, and on the other, the development of a complete novel prototype which implements the procedures of data capturing, 4D representation generation corresponding to the artworks, data treatment and data processing and analysis, to infer maps of physical surface deterioration and chemical distribution (representing either substance identification, mixtures, or mixture states, such as a unaffected or degraded compounds). The work developed included, the production of historic physical models of paintings on canvas and woods, some of which were aged and degraded by UV radiation, relative humidity and temperature cycle treatment, and pollution exposure. In addition a collection of real painting form the Baroque period was selected for further in-depth understanding of the deterioration of the art assets due to climate and environmental conditions. The technical developments focused on the production of a prototype, consisting of two optical channels for combined 3D digitisaton and hyperspectral imaging data capturing. The system worked in both spectral ranges (400-750nm and 900-2500nm). It permitted the registration of 3D and hyperspectral data in the form of images, which later on were processed to produce 3D point clouds and hypercubes which were directly related to the object’s surface and chemical composition. The data would be ultimately processed in an automatic way to help understand the current state of the art work, and any alterations or degradation suffered. 3D point cloud data was analysed to obtain degradation maps, by means tri-dimensional plane local approximations. Hyperspectral imaging was used to produce automatic classification by means of nonlinear supervised modeling, with the implementation of support vector machines, which helped in the identification and quantification of substances, mixtures and deterioration and degradation types, as defined by the user. The resulting system incorporated some of the most advanced technology for optical treatment: digital light projectors, based on digital micromirror devices, acousto-optic tunable filters, liquid crystal tunable filters and high-end focal plan array sensors working in the visible and short wave infrared. A novel optical, mechanical and electronic integration was also required for the completion of this transportable prototype. The SYDDARTA SYSTEM was tested on the real paintings selection with successful results. Interesting scientific and technical foregrounds were also identified at project completion, which will establish the bases of future commercialization among the heritage community and other areas of industrial activity. Finally a complete set of actions for project content and results dissemination were carried out by means of a dedicated webpage, publications in scientific and technical congress proceedings, the implementation of dedicated workshops in different EU countries and the production of a project biannual news letter reflected in the EU Heritage Portal.
Project Context and Objectives:
According to the EGMUS - European Group on Museum Statistics there are, approximately, 21,000 institutions across Europe that exhibit permanent or temporal art collections and act as a key role for the heritage sustainability by joining the importance of culture dissemination and the tourism as an exploitation market. For these organizations, cultural heritage monitoring comprehends an essential
routine and requires employing a sum of equipments, technologies and protocols to achieve the most detailed information about how the past of time and the environment contribute to deteriorate, change or weaken the artworks. Moreover, the accuracy of monitoring techniques nowadays becomes more important at the time that the durability of an artwork depends on the materials of which it is composed and detecting molecular changes due to complex processes as represent ageing and environmental climate effects is indispensable for the purpose of a proper restoration and conservation. In front of this, restorers and collection owners are working and facing a truth constrained by four aspects:
1. Handling restrictions: the artwork transport could mean important risks due to unknown data about the initial state and stability.
2. Several steps in the characterization: the complete measurement of deterioration parameters requires the application of some different techniques separately with a direct consequence in time consumption and wider data analysis.
3. Bad cost-benefit ratio: the use and maintenance of these equipments represent a high expenditure/investment respecting the data achieved in the time spent, and that produces a poor exploitation of these techniques and a gap in the artwork study.
4. Imprecision due to extrapolation: the study of whole artworks is difficult to perform since the current routine takes random or limited samples.
The main objective of SYDDARTA is to develop a prototype instrument for monitoring the deterioration of works of art and other heritage artefacts. This will be achieved by using “3D
Hyperspectral imaging”; where hyperspectral images are acquired whilst simultaneously scanning the 3D profile of the object. Being optical, the technique will be non-invasive and therefore non-destructive. The proposed new combination of 3D scanning techniques and hyperspectral imaging could widely reduce and increase inspection accuracy, while reducing time and costs in the characterization of art assets, speeding up the analysis procedures and minimizing the impact of manipulating and transportation on the artwork. To successfully develop such prototype, a series of technical and project operational considerations must be stated, as follows:
• 3D digitization. The new system will include 3D imaging capabilities by means of a digital light project (DLP), by means of Digital Micromirror Devices (DMDs) which would project light-on-light-off patterns on the objet for 3D digital surface reconstruction.
• Hyperspectral imaging in the visible and near-IR ranges. Many of the pigments considered for some historical periods can be properly distinguished in the visible and near-infrared part of the spectrum.
The prototype will be sensitive to the range 400nm to 2500nm.
• Types of evaluated cultural assets: Paintings on canvas and wooden panels. Within the Baroque period, the methodology development will focus on painting on canvas and on wooden panels. The prototype will produce a hyperspectral-3D model of each of the assets analyzed, for deterioration information extraction.
• Deterioration evaluation due to environmental conditions. Affection by temperature, relative humidity, UV radiation and pollutants deposition such as Carbon, Calcium Carbonate and dust, will be considered within the SYDDARTA project. Models for technical evaluation will be produced following receipts from the historical time selected. The choice of these two environmental parameters is done because it is the most important for painting degradation
Project Results:
SUBSTANCE DATABASE AND SPECTROSCOPIC INSPECTION KNOWLEDGE
General overview
A database of substance spectral behaviour (in ranges from 400nm to 2500nm by means of single point diffuse reflectance, dispersive spectroscopy) has been create in the project, with information on historic recipes substances prepared in the project. The database is accompanied with FTIR and Raman information and on the samples analysed (documentation). This database was produced by means of the middle demonstrator and single point spectroscopic systems).. The analysis included degraded and not degraded samples.
Foreground contents and specification
Inspection equipment used to generate this database: PerkinElmer 1500.
Spectral configuration of the database: spectral range covered (400-2500nm). Spectra resolution: 1 nm. Spectral curve form: Diffuse reflectance.
Number of spectral samples included in the database: 160.
Sample distribution: 40 spectra corresponding to the canvas reference sampled, 40 spectra corresponding to the UV-exposed canvas sample, 40 spectra corresponding to the RH-T treated canvas sample, 40 spectra corresponding to the polluted canvas sample.
ST Foreground format: digital document and txt. files.
SYDDARTA PROTOTYPE
General Overview
The SYDDARTA prototype is the main project foreground. It is a single piece of equipment, for the combined hyperspectral and 3D digitization of art assets, in particular paintings - for this first version. The novel system is particularly suitable for noninvasive inspections tasks of objects. The system is suitable for capturing data regarding surface status and spectral reflectance. Internally the system is capable of generating the 3D point clouds of the object under inspection and storing the spectral reflectance data as hyperspectral images of the covered field of view. The system can be directly used for the assessment of the state of an art asset. This state can be correlated with the degree of deterioration the asset may present. Object alterations not related to environmental and climate influence may also be assessed with the SYDDARTA prototype. The prototype implements a novel optical inspection arrangement, and a collection of processing algorithms for the purpose of evaluating the structure, surface and chemical composition of the object under study.
Foreground contents
The prototype consists of two physical units: (i) an optical module and (ii) power and control unit. The units can be controlled via a standard PC. There are also two software modules associated to the prototype: (a) a control application, which is used to control the hardware and enables data capturing and (b) a processing application, which is used to visualise and process the data captured with the control application. The two applications are run independently. The applications run on Windows OS environment.
The optical module includes two inspection channels, from which all 3D and hyperspectral data is captured. The two channels are named (1) the 3D imaging channel (3DI-channel) and the (2) shortwave infrared hyperspectral imaging channel (SHI-channel). The purpose of the first channel is to capture hyperspectral imaging data in the visible spectral range (400-750nm), and also 3D fringe pattern images to construct the 3D point clouds later on. The aim of the second channel is to capture hyperspectral imaging in short wave infrared range (900-2500nm). The whole system can be controlled
The processing application is used to pre-treat the data captured by the control application. The principal pre-treatment processes are:
For the hyperspectral data: data normalisation (via white reference and embedded multiplicative scattering correction algorithms), channel data merging (affine transformation to c-register the two image streams from the two channels), and dark current noise suppression for the SHI-channel hyperspectral data only.
For the 3D digitization data: 3D point cloud generation in real xyz coordinates.
The processing application has a collection of built in procedures and algorithms for data analysis and results generation:
For the hyperspectral pre-treated data: modeling environment for future substance and spectral occurrence automatic classification based on the following steps: a. data sampling from syddarta generated raw information, b. model training based on support vector machines, based on linear polynomial and radial basis function kernels, and c. model automatic classification and display of class maps. Sampled and modeled substances can be pigments, binders, supports, or else mixtures of substances, as they are normally presented in a painting. Spectral samples for model construction may also correspond to complex combinations which may present some alteration, or degradation, hence modeling the identification of such alteration on a single asset, for quantification purposes, or else on a collection of assets, for identification of occurrence, to say but a few applications.
For the 3D pre-treated data: 3D data processing by means of a multi-scale approach in which 3D data is segmented in areas of a given radius (the scale) and planes are approximated to them. The collection of locally approximating planes define a corresponding collection of normal vectors with a given direction in the 3D space, which then are compared with a standard planar reference. Deviations from the standard determine the non-flatness at a local level. For assets which are supposed to be flat, such as paintings, this deviation can be considered a definition of physical surface degradation.
Both control and processing applications have a GUI which allow user operation and data manipulation. In particular, de processing application presents 7 interfaces for general capturing control, system check and control and configuration for the different active components (such as the AOTF, DLP, light source, VIS and SWIR cameras and the LCTF). The processing application GUI implements a working desktop where data can be managed (load, save, display bands, display 3D point cloud, zoom in, rotate, scroll) and also visualize spectral curves on mouse movement on the hyperspectral image display. There is a panel on the left of the main GUI for modeling purposes, so the user can i. buil the spectral databases of interest by sampling on the displayed image and automatic spectrum recordin, ii. Train SVM models and, iii. Classify images hyperspectral images by means of the generated models, as many models as the user may require. There is also a small panel on the right of the interface for 3D data degradation map generation, based on the mutli-scale planar approach, embedded in the application.
The software modules are completed with two available functionalities: a. 4D stitching, to produce large digital representations of single objects which, due to their size, have been register in different captures, and b. the calculation of a confidence index for hyperspectral data classification.
Technical features:
Spectral ranges covered: 400-750nm and 900-2500nm.
Spectral resolution obtained on each range: nominal 20nm (VIS), nominal 5nm (SWIR).
Spectral capture sequence configuration:
Spectral configuration - VIS:
Bandwidth configuration: N.A.
Step size configuration: minimum of 1nm
Spectral configuration - SWIR:
Bandwidth configuration: configurable by means of RF generation signals in the control application. Recommended configuration: 10nm, 20nm and 40nm bandwidth.
Setp size configuration: minimum of 1nm
Field of view: VIS 40x40cm2; SWIR 6x6cm2; concentric.
Working distance: nominal 50 cm.
Object minimum measurable reflectance: 5% (below this the system captures noise, specially in the SHI-channel).
Communications and signals (LCTF (USB hyperterminal), VIS camera (GigE), SWIR camera (USB), AOTF (USB), DLP (USB), Light (DC signal).
TUNABLE SWIR ILLUMINATION MODULE
The SWIR illumination unit, as integrated in the prototype with minor optical adjustment at the output, can be considered an S/T foreground, as it has been specifically constructed within SYDDARTA and for the purpose of the prototype construction. This illumination unit allows the wavelength selection of the projection beam. In this way, only radiation of interest reaches the object to be inspected. This illumination source can be used to implement fast hyperspectral scanning arrangements. Other applications may be illumination of samples for photomicrography.
FOREGROUND CONTENTS
The illumination source is made up of a halogen tungsten bulb (30 W and a specifically design reflector to optimise the light bulb emittance and convey the radiation to the AOTF input, and two optical stages (a collimating lens and a SWIR accepting bandpass filter). The following stage consists of a linear grid polarizer in the SWIR region which orientates the oscillation plane of the electromagnetic radiation.
Radiation enters an acousto-optic tunable filter, with an acceptance angle of 3 degrees and an optical window of 1x1 cm2. The AOTF is made up of a birefringent crystal (TeO2) which produces a double wavefront in the crystal structure, each front with a different polarisation (ordinary and extraordinary). The acout-optic effect produces the extraordinary polarisation to be diffracted from the main optical path, with an angle of 7 degrees. The radiation outputting the AOTF is then split into two paths, one with undesired radiation and the other the patch driven to the object for single band illumination. The 0-order cancellation (i.e. unwanted, non-monochromatised light) is cancelled by means of a Rochon prism a the AOTF output and a physically generated beam dump in the output optical path.
Technical features
Spectral ranges covered: 900-2500nm.
Spectral capture sequence configuration:
Spectral configuration: Bandwidth configuration: configurable by means of RF generation signals in the control application. Recommended configuration: minimum of 5nm, 10nm, 20nm and 40nm bandwidth.
Setp size configuration: minimum of 1nm
Field of view at 50 cm: SWIR 6x6cm2.
Optimal working distance: nominal 50 cm.
Communications and signals (AOTF (USB), Light (DC signal).
Power needs: 220AC (converted to 24V DC internally).
Specific casing and output optical configuration for stand-alone operation not developed in the project.
THE PROCESSING APPLICATION
The processing application is a stand-alone piece of software which can be run independently from the SYDDARTA prototype. The application has been developed to work with standard image formats (such as jpg). The application implements a number of pre-treatment features, particularised for the SYDDARTA control application raw data. This functionality can be bypassed and the application is capable of loading standard hyperspectral imaging format .bsp. Likewise, embedded pretreatment can be either enabled or disabled, in relation to white reference compensation, normalisation and dark current removal. The application structure is configured in such a way that it can either load the fringe pattern images - and generate the 3D point cloud - or load a vrml and XYZ 3D data file formats. The application is conceived as a working bench to manually inspect 4D data representations of objects (by displaying non processed 3D point clouds or hypercubes) or apply processing techniques for identification of both 3D patterns and hyperspectral trends. The features of the processing motor are the usage of SVM modeling techniques with different kernels (linear, polynomial and radial basis functions) to produce classifiers for automatic spectral fingerprint identification and local deviations of a 3D surface form a planar standard.
FOREGROUND CONTENTS:
The two main sets of processing algorithms are implemented in this application:
3D data processing algorithms for the identification of non-planar surface areas at different scales
Hyperspectral automatic modeling for substance, mixture and alterations identification.
The application consists of a single GUI interface which allows the user: i. to manage 4D data (load, save, select) form a pane on the left part of the window, ii. visualise and display data (3D point cloud and hyperspectral imaging) on a central image control, iii. visualise spectral curves from the hyperspectral image in real time on a central lower pane, iv. spectral digital sample, model and classify with built in functionality from a left positioned panel, and v. apply the degradation estimation functionality on a left bottom panel. Displaying of 3D degradation maps, classification maps and confidence maps is one in the main image display control (central to the GUI).
Technical features
Windows OS application
Programming language C++
File acceptance: JPG, TIFF, .bsq, .txt, .vrml
File maximum size: 1GB
SECONDARY FOREGROUND
As part of the development of the prototype, different components were produced and adjusted by the technical partners, which resulted in variations of already existing commercial products. This is the case of the following optical components, integrated as part of the prototype:
Second foreground - DLP projector:
Based on the DLP technology by Texas Instruments, the DLP produced in the project was improved with respect to the already existing version by the implementation of a HD DMD chip (1080x1920 pixels) with a high power LED in a luminescent reflector which produces a constant continuous emmittance between 420 to beyond 700nm. Implementation of electronics, projecting optics and encapsulation where considered a novelty and hence a secondary foreground.
Second foreground - SWIR camera:
Based on previous camera models commercialized, the MCT SWIR camera unit integrated in the prototype was specially design to optmise quantum efficiency. In fact, this parameter was almost kept constant throughput the range of interest, according to manufacturing measurements. The camera incorporated also an improved procedure for image adjustment (compensation and removal of dark and white pixels, typical in MCT sensor after manufacturing).
Potential Impact:
SYDDARTA POTENTIAL MARKET
The main output of the project is the achievement of a pre-industrial prototype with highly innovative features (product family 1) that will have a modular character. This means that the equipment may be understood as a whole system but also the features may be separate into specific devices (product family 2 and 3) and a complete spectral database (product family 4) in order to estimate market sales.
As a consequence of the previous analysis, the consortium has defined two markets: the direct one composed by cultural heritage end-users and the secondary one composed by other sectors to explore.
I. Principal market: cultural heritage end-users
According to the EGMUS - European Group on Museum Statistics there are, approximately, 21,000 institutions across Europe that exhibit permanent or temporal art collections and act as a key role for the heritage sustainability by joining the importance of culture dissemination and the tourism as an exploitation market.
The European Fine Art Foundation (TEFAF) has released its annual report on the international art and antiques market. The study for 2013, carried out by Arts Economics , used data gathered from a range of sources: art dealers, auction houses, collectors, fairs, financial databases, industry experts and many others involved in the international trade of art. Although the report focused its attention on the US and Chinese art markets in particular, this year's headline finding was that the global art market reached €47.4bn (£39.7bn) in total sales of art and antiques in 2013, close to its highest ever recorded total and advancing 8% year-on-year.
The project has applicability for different types of entities, all in relation to the arts and heritage in two principal ways.
- Conservation
- Authentication
The first impact of SYDDARTA is to contribute to overall preventive strategy to maintain and secure the conservation and management of movable assets in public and private cultural heritage collections, regarding "natural" deterioration processes and human interventions.
CONSERVATION
Developed around two parallel lines of work:
1. Preventive Conservation: risk analysis and implementation of appropriate measures for storage, handling, transport and exhibition of art works, to ensure their proper preservation. Preventive conservation is now an essential element of any policy on heritage conservation, including;
-Preventive measures: optimal storage conditions, storage, handling, packaging, transport, exposure and maintenance works, taking into account parameters such as room temperature, relative humidity, radiation, visible light, ultraviolet radiation, vibration, shocks, disasters and plagues, etc..
-Action protocols in case the risk materializes.
2. Study of techniques and materials of construction, as well as causes and degradation of the latter, in order to establish the necessary conservation measures and, where appropriate, consider the most appropriate treatment or interventions.
Each work must be comprehensively analyzed to identify technical and physical characteristics, to establish its state of conservation and develop the best strategy capable of ensuring its preservation. In parallel, some have set a specific program storage, handling, transport, exposure and maintenance for each work.
In SYDDARTA, IPCHS and RABASF are end-users representing heritage and collection owners which have the need to incorporate better cost-benefit conservation equipment and ensure the sustainability exploitation of the artworks.
Moreover, the business plan contemplates the option of renting the prototype in terms of providing a service to the cultural heritage sector in Europe. Due to the fact that some end-users will require the use of SYDDARTA in a concrete period or for a specific study of an art work, the consortium has contemplated the possibility of renting the SYDDARTA prototype for different institutions and entities. This option will give an opportunity to avail from the SYDDARTA technology without having to invest in equipment. This option is particularly recommended for medium and small preservation centers, museums, and research institutions.
Taking into account the specific characteristics of the system and the importance of the objects to be inspected by SYDDARTA, the consortium will create a group of technicians formed by representatives of project participants that will elaborate a prototype manual for technical use.
AUTHENTICATION
The expertise and authentication play an increasingly important role in the art market. The evaluation of the price of an art object (or collection) depends primarily on the degree of authenticity. There are still many identification and allocation tables errors, with financial and legal consequences. Scientific techniques are becoming increasingly sophisticated, and go further in the exploration of materials.
Many players can make the cost among them: sellers, buyers, companies Sale Volunteers (SVV), research laboratories, restaurant, insurance companies, lawyers and judges, historians of and finally the art experts.
Determining authenticity is the mission of the expert. The art expert, with experience gained over many years, is a specialist capable of determining the nature, origin and time of manufacture of the object subject to judgment. He is asked to authenticate, but also to detect alterations, conversions and repairs that may be incurred by this and indicate a range of values to be used for trading on a sale or estimate, for example, an insurance company.
The expert operates in two business areas:
1. Activity authentication and estimate for the certificates insurance
2. Activity expertise and estimation in the context of public sales
It is often faced with many difficulties: firstly, it is technically easy to produce fake. The result of expertise then depends on the importance of research that the expert will complete. Finally, the state of the art and knowledge are constantly changing, challenging entrenched beliefs.
The European Fine Art Foundation (TEFAF) has released its annual report on the international art and antiques market. The study for 2013, carried out by Arts Economics , used data gathered from a range of sources: art dealers, auction houses, collectors, fairs, financial databases, industry experts and many others involved in the international trade of art.
II. Secondary markets
SYDDARTA system can be applied into different sectors other than conservation and restoration of cultural heritage. Depending on the specific field where it is going to be applied the system will need a specific adaptation to be implemented in a concrete market, due to the fact that each industry requires different functionalities and operates with special compounds and substances.
Therefore the system must implement a specific database for the concrete sector where it is going to be applied; this means that following the market sector where project will be used, it will be created a database for the characterisation and composition of the substances that compounds the object to be inspected. In addition it will be elaborated different algorithms specifics for the software development in order to better detect and identify the presence of the interesting components and provide a quantitative estimation of them.
SYDDARTA system will obtain the physic-chemical and 3D information of the objects and will inspect them. The system will be implemented in the following sectors:
• Agro-food – Under this market, the system will be able to detect the presence of harmful chemicals components, bacteria, virus, organic and inorganic contamination, food deformities, food good conditions (humidity, temperature, ripening, decay, etc.) and to distinguish the composition and volumetric estimation of food.
• Biomedical – Under this market, the system will be able to detect the characterisation and study of human body movements by analyzing the mechanics involved on them. The system will help on the detection of 3D and spectral information related to the human body behavior in a non-invasive way. The system can detect and provide: biometric estimations, size parameters, presence of oxygen, hemoglobin, etc.
• Pharmaceutical – Under this market, the system will be able to help on the process for medicine manufacturing, estimation of active components and excipients.
• Part manufacturing industry – Under this market, the system will be used for the inspection of part dimensionality as well as overall structural and chemical quality.
Based on the above considerations, the following includes an estimation of the Syddarta foreseen source of income per product family (see Business Plan D7.7):
Commercialization product family 1: Prototype Full Equipped
- Expected sales: First 2 years 15 units; First 5 years 100 units
Commercialization product family 2 : Extended spectral range imaging and 3D digitization applications for art heritage assets and other manufacturing industries
- Expected sales: First 2 years 15 units; First 5 years 80 units
Commercialization product family 3 : Hyperspectral and 3D digitization applications for art heritage management industry and research
- Expected sales: First 2 years 15 units; First 5 years 80 units
Commercialization product family 4 : Spectral database for the characterization of art assets
- Expected sales: First 2 years 30 units; First 5 years 120 units
EXPLOITATION OF RESULTS
Depending on the nature of the entity, the way of exploiting results can be quite different, oriented more or less directly to immediate profitability.
Non-profit entities
The mission of a non-profit entity is to investigate, to transfer knowledge in education, to transfer knowledge to economy through spin-offs and to disseminate knowledge through publications and congress contributions.
It is committed to scientific research and the highest international academic standards. Due to its non-profit nature, its potential exploitation activities mainly fall into the spheres of education, research and knowledge transfer.
In a particular way, the orientation for each entity will be the following:
• AIDO - Interest in patenting and exploitation of results through licensing to third parties.
• FORTH - Interest in patenting and exploitation of results through licensing to third parties.
• TUD - Interest in patenting and exploitation of results through licensing to third parties.
• ISAC - Interest in patenting and exploitation of results through licensing to third parties.
• IPCHS - Interest in patenting and exploitation of results through licensing to third parties.
• RABASF – Interest in patenting and exploitation of results through licensing to third parties.
For all of them, there is an interest in using the system once it has passed the end of the project, to study and produce scientific and technical documentation in order to make more publications and disseminate the
Companies and profit entities
After completion of the project, the business interest lies in the rapid exploitation of results, for maximum profitability. There are two main ways to achieve this profitability: intellectual or industrial protection, lease, or sale of property; immediate sale or license, without protection.
In a particular way, the orientation for each entity will be the following:
• SIGNINUM – Interest in patenting and use of the resulting product for generating inspection services business.
• AVANTES – Interest in patenting and production of lighting components (reflector and light source SWIR) on the line developed in the project, either as part of the prototype SYDDARTA either as standalone product marketing.
• G&H – Interest in patenting and production of lighting components (ATOF developed SYDDARTA)) on the line developed in the project, either as part of the prototype SYDDARTA either as standalone product marketing.
• VIALUX – Interest in patenting and production of lighting components (DLP projector) on the line developed in the project, either as part of the prototype SYDDARTA either as standalone product marketing.
• XENICS - Interest in patenting and production of sensing components (SWIR camera developed for the project) on the line developed in the project, either as part of the prototype SYDDARTA either as standalone product marketing.
DISSEMINATION ACTIVITIES
The Project Consortium has developed different activities focused in the dissemination of results, reported in the Periodic Report 1 and 2, classified as follows;
Press Releases in Web sites/Applications:
AIDO; https://arquivo.pt/wayback/20191218154828/http://www.syddarta.eu/ AIDO'S Newsletter, Twitter of Otri grupo, Facebook of Fedit (Federation of Spanish Technological Centers), Web Page and Newsletter of Fedit (Federation of Spanish Technological Centers), Diario de Salamanca, Agencia SINC, http://www.zvkds.si/sl/varstvo-kulturne-dediscine/raziskave-in-projekti/
Articles published in the popular press:
EL MUNDO Newspaper (Spain), LEVANTE Newspaper (Spain).
Presentations, Oral presentations to a scientific event and Exhibitions
TUD: NWO Getty Panel Paintings Initiative Meeting, Optical Metrology Techniques for Diagnostics of Artwork, Fotonica exhibition, International Conference of the Ageing of Materials & Structure
RABASF: Information to cultural institutions
SIGNINUM; Jornadas ARP; Universidade Católica Porto, Conference Heritage Colors: Faculdade Ciência e Tecnologia Lisboa, SPIE Photonics Europe; Syddarta Portuguese Tour
AIDO: Syddarta stand in the eighth edition of AR&PA Biennial, under the title "Innovation in Cultural Heritage"
VIALUX MESSTECHNIK + BILDVERARBEITUNG GMBH Symposium on Emerging and Industrial DLP Applications, SPIE Photonics West, Laser World of Photonics, 8th International Symposium on Emerging on Emerging and Industrial Texas Instruments DLP Technology Applications
CONSIGLIO NAZIONALE DELLE RICERCHE; Ambiente e patrimonio culturale: degrado, previsioni e strategie di conservazione. XIV CONGRESSO NAZIONALE DI CHIMICA DELL'AMBIENTE E DEI BENI CULTURALI; VIII Congresso Nazionale di Acheoetria. Participation in the round table “Prospettive e Strumenti per il finanziamento della ricerca applicata al Patrimonio Culturale”;
AIDO: VIII Congresso Nazionale di Archeometria. Poster Novel 3D hyperspectral methodology to estimate deterioration in paintings: Syddarta. Syddarta workshop: Novel 3D-hyperspectral methodology to estimate deterioration in paintings (Valencia, Spain)
FOUNDATION FOR RESEARCH AND TECHNOLOGY HELLAS: Syddarta: Development and Applications of novel technologies in Cultural Heritage
Publications;
SYDDARTA: new methodology for digitization of deterioration estimation in paintings. Proc. SPIE 8790, Optics for Arts, Architecture, and Archaeology IV, 879011
Deterioration estimation of paintings by means of combined 3D and hyperspectral data analysis. Proc. SPIE 8790, Optics for Arts, Architecture, and Archaeology IV, 879011
“Combined 3D-hyperspectral Analysis for art Asset Deterioration Assessment – SYDDARTA Project “
“A Aplicação do Sistema Syddarta na Conservação Preventiva”
List of Websites:
Project website address: https://arquivo.pt/wayback/20191218154828/http://www.syddarta.eu/
Contact details;
E-mail 1: ribes@aido.es
E-mail 2: eribes@aido.es