基于包含度理论的双阈值视频镜头边界检测算法

A dual threshold shot boundary detection method based on subset-hood theory

在视频镜头边界检测中,由于人眼视觉具备的主观性和视频帧图像结构与边界的不确定性,比较适合采用模糊理论,本文基于模糊集中的包含度理论提出双阈值化的视频镜头边界检测方法:首先基于包含度公式定义视频镜头边界检测函数,然后利用包含度理论提出视频边界镜头渐变和突变的通用检测算法,最后提出双阈值优化选取办法,实验结果表明本文方法具有较好的检测渐变和突变镜头能力。

As the subjective of human vision and the uncertainty of video image structure,fuzzy techniques are used in shot boundary detection.Based on the subset-hood theory,a dual threshold shot boundary detection method is proposed in this paper.Firstly,the new detection function of video shot boundary is defined.A new algorithm for gradual and cut transition detection is then proposed on the basis of subset-hood theory.Finally,an optimal method for choosing dual threshold is presented.Experiments show that the algorithm is effective in detecting gradual and cut transition.