MetadataShow full item record
Surveillance video is practically never used because there are too many hours of video for people to watch. Most attempts to deal with the overflow of surveillance video involve the development of automatic video understanding, including object recognition and activity understanding. Video synopsis is complementary to video understanding. After objects are detected, video synopsis changes the time of display of each object so that more objects are “packed” into a shorter time. The resulting video is a shorter summary of the original, where the objects are shown more densely. While video synopsis can reduce, on average, an hour of video into a minute, the synopsis loses causality, resulting in the possibility of objects that appear together in the original video appearing at different times in the synopsis, and vice versa. Video synopsis is an excellent tool for browsing video, and when used in conjunction with the results of video, an excellent method for understanding tools.
- IRIM Seminar Series