Behavior and event detection for annotation and surveillance

Visual surveillance and activity analysis is an active research field of computer vision. As a result, there are several different algorithms produced for this purpose. To obtain more robust systems it is desirable to integrate the different algorithms. To achieve this goal, the paper presents results in automatic event detection in surveillance videos, and a distributed application framework for supporting these methods. Results in motion analysis for static and moving cameras, automatic fight detection, shadow segmentation, discovery of unusual motion patterns, indexing and retrieval will be presented. These applications perform real time, and are suitable for real life applications.