视频关键帧提取技术是视频数据处理研究领域的热点研究问题。该文针对现有的镜头边界检测技术不能有效提取关键帧的不足，提出一种基于小波边缘检测算子的自适应分块视频镜头边界检测算法。通过检测视频镜头变化，得到分割的镜头，然后对视频帧提取图像特征，并利用基因表达式编程（GEP）的自动聚类功能对视频帧进行聚类，提出并实现了基于GEP自动聚类的视频关键帧提取算法（KFC‐GEP）。实验证明该方法能较好的提取视频序列的关键帧。%The technology of key frame extraction is a research focus in video data processing do‐main .A video shot boundary detection algorithm with adaptive division based on wavelet edge detec‐tion is presented to overcome the drawbacks of the available algorithms for shot detection technology in this paper .First ,we obtain the video shot segmentation by detection of video shot change .Fur‐thermore ,we extract the image feature from video ,which cluster by autoclustering based on Gene Expression Programming ,propose and implement the video key frame extraction using an autocluster‐ing algorithm based on Gene Expression Programming (KFC‐GEP) .The proposed method is demon‐strated efficiently and effectively for extracting the key frame in video experimental results .
展开▼
