Efficient algorithm on video shot segmentation of the adaptive threshold

Abstract

This paper presents an efficient algorithm on video shot segmentation of the adaptive threshold. First of all, the features of the images color and texture will be extracted by using the wavelet transform, then the mutation and gradual shot boundary detection will be achieved by utilizing the information entropy. According to the difference distribution of the sliding window, the local threshold will be calculated dynamically. And the degree of accuracy on the video shot segmentation algorithm is increased. This method can not only detect the mutation shots but also the gradual shots. The experimental results show that this algorithm method can efficiently deal with the video shot segmentation.