Skip to main content

INTELLIGENT VISION SYSTEM FOR AUTOMATIC INSPECTION OF ALIMENTARY PRODUCTS

Objective


At the end of the Project, the following results were reached:

1. Techniques have been developed for the on-line examination of meat using multiple, existing technologies. Studies of the most appropriate sensors have been carried out too. Though, the X-Rays and Infrared sensors chosen were for chickens and pig leg examination, their application in other areas within the food industry were considered, especially within the meat industry. Safety issues, the effect to production environment on the sensors, and their costs were also considered

2. Image processing algorithms have been developed for the recognition of most of the chicken meat features, including sub-surface features using infrared sensors

3. A chicken test bed was constructed, so to demonstrate the ALINSPEC system in terms of image acquisition capabilities

4. Although a pig leg demonstrator was not built, several stand-alone software modules, to be demonstrated on an off-line mode and covering almost all the different industrial requirements, have been developed.

4. For pig legs, the priority is not in the recognition of meat features required for on-line examination, but in the recognition of features which will be beneficial to those developing systems for automating the pig butchery process. Hence classical and neural algorithms have been developed for identifying bones and muscle groups on the cut surface of a pig leg. In addition, algorithms of meat features and the measurement of the fat thickness have been developed. Such features are essential for assessing the suitability of a pig leg for spanish dry cured hams.
An Intelligent Vision Sustem For Real Time Inspection of Alimentary Products (ALINSPEC) especially suited for "on line" quality control of chicken and pig meat is proposed.

Multiple sensor data are processed by means of fusion techniques to obtain the most amount of significant information of the observed "scenario". The system employs Artificial Intelligence techniques to produce fast decision for actions to be performed, in order to "react" and to govern the production process according to production flexible rules. Working demonstration prototypes are planned to be implemented for quality control and government of two distinct production lines.

The main goals of the system will be the following ones :

- Reduce level of incertainties due to visual inspection;

- Increase yield;

- Continuous and exaustive control of the overall quality level of the production;

- An Automatic Management of the production process;

- An immediate analysis of the effects due to of a poor production strategy or to defects on production process, allowing a quick feedback towards previous stages of the production chain.

Funding Scheme

CSC - Cost-sharing contracts

Coordinator

Consorzio Ceo Centro di Eccellenza Optronica
Address
Largo Enrico Fermi 6
50125 Firenze
Italy

Participants (7)

BULL SA
France
Address
68 Route De Versailles
78430 Louveciennes
COREN FRIGOLOURO
Spain
Address
Puente Del Valo
36400 Porrino
Campden and Chorleywood Food Research Association (CCFRA)
United Kingdom
Address
Station Road
GL55 6LD Chipping Campden
Officine Galileo SpA
Italy
Address
Via Albert Einstein 35
50013 Campi Bisenzio (Fi)
UNIV VIGO
Spain
Address
Campus Univ.-lagoas-marcosende
36200 Vigo
UNIVERSITÀ DEGLI STUDI DI FIRENZE
Italy
Address

50125 Firenze
University of Strathclyde
United Kingdom
Address
Royal College 204 George Street
G1 1XW Glasgow