一种利用二次筛法分析画面异常的方法

A Method of Analyzing Image Abnormality by Quadratic Sieve

视频内容检测可以保证数字视频的质量,数字化时代的视频内容自动检查系统是一种自动化的代替人工检测的视频内容检测系统.该系统使用的检测算法还存在检测效率低、CPU占用率高的缺点.本文针对这一问题提出一种综合使用二次筛法、提起退出机制、多样化采样技术等技术的检测方法.经试验测试,该方法的检测效率可达传统检测效率的4倍左右,且检测准确度比传统检测算法高7个百分点左右.本文提出的高效检测方法可用于黑场检测、白场检测、蓝底检测、绿底检测、彩帧检测、画面颜色丢失检测等常用检测项.

Video content detection can guarantee the quality of digital television.Video content automatic detection system is used to replace the manual detection.Their algorithms have low detection efficiency and high CPU occupancy rate.Aiming at this problem,we propose a new efficient video content detection algorithm in this paper.The detection efficiency of the algorithm is much higher than that of traditional algorithms,and the detection accuracy is higher.The algorithm proposed in this paper can be used for the detection of black fields,white fields,green frames,blue frames,single color frames,color loss frames,etc.