Camera Operation Estimation Algorithm using Low Level Vision in Digital Video

H.C. Lee, Y.W. Chang, J.H. Lee, and O.B. Jang (Korea)


HCI, manufacturing, optical flow, video Index, visual interface Video database


Digital video computing and organization are two important issues in multimedia systems, signal compression, and databases. The segmentation of the video into shots requires identification and indexing. This approach requires a suitable method for the automatic location of cut points in order to separate the shots in a video. Various changes may happen within a single shot such as those related to camera operations, object movements and flashes. Therefore, we propose a new cut detection technique, based on linear correlation coefficients of inter frame motion vectors, and a new video indexing method that uses motion vectors to identify video sequences. The index is based on the identification of discrete cut points and camera operations made possible by analyzing motion vectors. Experimental results applied to commercial video are presented and evaluated.

Important Links:

Go Back