Digital library creation based on wavelet coefficients for video stream indexing and retrieving

The increased availability and usage of on-line digital video has created a need for automated video content analysis techniques including indexing and retrieving. Most research on video content involves automatically detecting the boundaries between camera shots. After shot detection there is a need for shots indexing in a library which enables retrieving for the required stream based on shot. There is a need to build an efficient digital library for those shots, saving the proper features for each shot/frames, which will enable efficient retrieval for user´s requirement. This paper presents an automation technique for video indexing and creation of a digital library. We will present how to detect shots cuts and features, which will be saved for the video shots/frames. Also build a video digital library composed of the stream shots/frames plus wavelet coefficients for video stream shot´s key or all frames (for full search function in all frames of the indexed video). The digital library represents full video streams frames indexed into shots with overhead in the stored size just 16.2 % as wavelet coefficients. The retrieving efficiency for a shot while having an input frame is ranged from 100%, if the input frame is saved keyframe, down to 94% if the input frame is any other frame. Our digital library is valid for any number of video streams, any number of shots/ streams and any number of frames/shot.