摘要
对现有传统分割算法进行分析,提出了一种改进的自动分割镜头的算法.通过计算视频序列相邻帧之间的灰度差值及色彩差异,自适应进行视频镜头的边界检测和闪光检测.这种算法将帧差法和直方图法这两种传统算法相结合,有效地提高了检出率及精确率,是一种简便、快捷的算法.
Through the analysis of traditional ways, an improved automatic method is put forward for shot segmentation. Based on the differences of gray and color between the two adjacent frames, the adaptive detection for the shot boundary and flashlight is performed. This algorithm combines the frame difference method and his togram method.Hence the detection is more accurate, simpler and quicker than traditional ways.
出处
《应用科技》
CAS
2006年第6期90-92,共3页
Applied Science and Technology
关键词
镜头分割
边界检测
闪光检测
直方图
shot segmentation
boundary detection
flashlight detection
histogram
