基于小波变换及熵的视频镜头分割检测方法  

Video Shot Detection Method Based on Wavelet Transform and Wavelet Entropy

作　　者：戴小文[1] 蔡志平[2] 钟桂英[1] 

机构地区：[1]西南交通大学电气工程学院,四川成都610031 [2]厦门大学计算机科学技术学院,福建厦门361000

出　　处：《西南交通大学学报》2008年第3期314-318,共5页Journal of Southwest Jiaotong University

基　　金：信息产业部电子发展基金资助项目((2005)No.555)

摘　　要：提出了基于小波变换及熵的视频镜头分割检测方法.用图像低频信号的小波系数的均值和标准差检测剪切镜头转换边界.分别计算图像小波分解后水平、垂直和对角方向细节信号的小波熵分量,以此3个小波熵分量作为特征量,计算相邻视频帧间特征量的欧氏距离.在检测窗内欧氏距离两次以上大于设定的阈值时即可判定镜头发生渐变转换.实验结果表明,剪切镜头查准率为95%,查全率为96%;渐变镜头查准率为87.5%,查全率83.8%.A detection method for video shot transitions based on wavelet transform and wavelet entropy was presented. The mean and standard deviation of the low frequency components of an image is used to detect cut transitions. To detect gradual transitions, the wavelet entropies of the image in horizontal, vertical and diagonal directions are used as characteristic values, and these values are taken as inputs to calculate the Euclidean distances between neighboring images. A gradual transition is determined if at least two Euclidean distances are larger than a preset threshold in the detection window. Experiment results indicate that using the proposed method the ratio of correct detection is 95% and the detected ratio is 96% respectively for cut transitions, and 87. 5% and 83. 8% respectively for gradual transitions.

关 键 词：视频镜头 检测 小波变换 小波熵 直方图 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]