Semi-automated Reviewing Station for IAEA Optical Surveillance Data

A study is underway on the use of computer vision technology to assist in visual inspection of optical surveillance data. The IAEA currently uses optical surveillance as one of its principle Containment and Surveillance (C/S) measures. The review process is a very time-consuming and tedious task, due to the large amount of optical surveillance data to be reviewed. For some time, the IAEA has identified as one of its principle needs an automated optical surveillance data reviewing station that assists the reviewer in identifying activities of safeguards interest, such as the movement of a very large spent fuel cask. The present development reviewing station consists of commercially available digital image processing hardware controlled by a personal computer. The areas under study include change detection, target discrimination, tracking, and classification. Several algorithms are being evaluated in each of these areas using recorded video tape of safeguards relevant scenes. The computer vision techniques and current status of the studies are discussed.