Skip to main content

ORIC

Activity Recognition & Anomaly Detection

CCTV surveillance systems aren’t as accurate as needed. We gather extensive information on object’s activity based on motion trajectories, e.g., tracking of players and ball in the playing field. To extract motion sequences from video, tracking algorithm is used. This technique classifies new trajectories and anomalies by checking closeness of object trajectories. The system has potential application in CCTV surveillance systems, GPS, SATNAV and phone cameras. A major challenge in modern day surveillance systems is the difficulty in separating primary target from the background. The proposed system uses thermal radiation information to separate the target on the basis of heat signature and detailed texture information from visible images to create a more clarified image.