SMART aims to provide the deminers (end-user) with safe, user-friendly, cost-effective, efficient and innovative tools for the monitoring of the environment and for the assistance to people in countries afflicted by landmines in order to achieve a higher quality of the service, by efficiently improving level 1 minefield surveys. For that, SMART will collect data with an airship multi-sensor survey system and during ground truth data collection sessions, process all data using data fusion techniques, land-cover classification tools and anomaly detection algorithms, integrate the tools and the data into the minefield survey system of the end-user and validate the results on validation test sites. SMART aims not at solving automatically the problem of mine suspected area reduction, but at helping the human analyst in their interpretation tasks.
The main objectives of SMART are:
1. to develop a generic methodology to systematically collect significant information on mine polluted regions with suspected area reduction as main topic
2. to develop efficient tools for mine infected area investigation and reduction using active and passive airborne high resolution sensors in order to assess the potential use of next generation high resolution satellite data for future survey operations for humanitarian demining
3. to validate the use of these tools and to validate their efficiency and reliability for improving the level 1 minefield survey
4. to provide the end-user with these tools in order to help them in improving efficiency and reliability of level 1 minefield survey
Mine clearing is a long and dangerous task. A lot of time is lost inspecting areas that turn out to be mine free. Deminers are looking forward to methods that can reduce the suspected areas. An important lesson of some EC projects is that it is impossible to detect AP-mines with very high aerial photography. For that reason, SMART aims to reduce the suspected areas indirectly using land-cover classification and anomaly detection such as change detection. The data collection module of SMART will consist of the organisation of an airborne campaign (active full polarimetric SAR and passive multi-spectral sensors), the gathering of existing spaceborne data and the collection of ground-truth data, expert knowledge and context information. SMART aims to transform the raw data into meta-data that will be useful for the end-user survey expert. The transformation of the raw data into the meta-data will be based upon land-cover classification and change detection. In order to use all available information, data fusion will be considered. Data fusion techniques are needed to fuse the spectral bands of one sensor, to combine the airborne data with the spaceborne data, to merge the active sensors with the passive sensors and to integrate the expert knowledge and context information. A pre-processing module will transform the data so that fusion can be applied. In this module, the effort will be putted on geo-coding and registration of the airborne data. The SMART tools will be integrated into the minefield survey management system of the end-user. Besides the development of the processing tools, the accent will be putted on their validation. The end-user survey expert will for that reason actively be involved from the beginning of the project.
T0 : start of the project
T1 : T0+18expected results: collected data and multispectral measurements, geocoded and registered data, validation protocols, architecture design management system
T2 : T0+36expected results: management system, tools (fusion, landcover, change detection,...) for key parameter detection for the reduction of suspected areas.
Funding SchemeCSC - Cost-sharing contracts
75634 Paris 13