Shot Boundary Detection Based on HSV Color Model

In this paper, the traditional algorithm of shot boundary detection based on histogram difference is introduced briefly. Based on this, a video shot boundary detection algorithm based on HSV color model is proposed. The method firstly detects the shot hard cut based on the weighted color histogram difference curve, then performs the shot gradual transition detection according to the image local average pixel value characteristic curve. Finally, two partial detection results are fused, and all the shot transitions of the video can be obtained. Experimental result shows that the method has a higher recall and precision.