Video shot boundary detection method using histogram differences and local image descriptor

Video shot boundary detection is a fundamental component in automatic video indexing, archiving and editing. Different techniques have been developed in the past with reliable performances in the video processing, especially for the key frame extraction and video summarization. In this paper, we propose a new method for video cut detection in two steps. First, we calculate the histogram differences between each two consecutive frames. Afterwards, we define a thresholding selection that give a set of images candidates for which we adapt the surf algorithm to eliminate false detections. Experimental results are presented to illustrate the good performance of the proposed method with a pourcentage of 100% for the recall criterion.