Video segmentation algorithm using threshold and weighting based on moving sliding window
Wonhee Kim, Tae-Il Jeong, Jong‐Nam Kim · International Conference on Advanced Communication Technology · 2009
The video segmentation is a main technique for automatic management of video data. We propose the video segmentation algorithm using thresholds and weighting based on moving sliding window. The proposed method generates threshold and weighting factor using feature value of frame on moving sliding window. And we improve our algorithms to apply adaptive for various video data. We apply variable weighting according to the number of fade-out frame to remove missed-detection by fade-our transition. We obtained better the detection rate than conventional methods based on histogram about 2.5% in F1.