基于压缩域的自适应场景叠化检测算法

Adaptive detection of dissolve for MPEG videos based on sliding windows modeling

给出了一种压缩域中快速有效的自适应场景叠化检测算法,该算法综合利用了视频帧中宏块类型信息的场景叠化和非场景叠化特征,并利用中值滤波对形成的数据流进行滤波,再利用滑动窗口准确地检测出场景叠化的开始帧和结束帧。实验结果达到了较高的查全率、查准率和F度量,并和现存的场景叠化算法进行了对比实验。

The proposed method of dissolve detection exploits the MPEG motion estimation and compensation scheme by examining the character of dissolve and no-dissolve for each macro-block inside frames. A sliding windows mechanism and median filter are employed to detect the precision location of dissolve. The proposed algorithm achieves superiority measured by recall, precision and F measure rates compared with existing methods.