Scene cut detection in video by using combination of spatial-temporal video characteristics

For the mainstream hard cut scene detection algorithm based on the color-histogram difference new criteria are added to increase precision and improve computational efficiency. This approach utilizes additional detection criterion based on spatial distribution of luminance blocks difference during scene cut. Also, in order to improve computational efficiency, time sampling at larger time resolutions and detection at smaller time resolutions is used. On the different video examples the results in precision and reliability are given.