基于最优直线段边缘方向的视频清晰度评估

摘要

To test the problem of quality degradation of security monitoring video caused by improper camera parameter setting,a video sharpness testing method based on the edge direction of the optimal straight-line segment was proposed.In view of the characteristics of overall video quality degradation caused by the blurring of the surveillance video image and the totally different performances in local areas,accurate video image sharpness metrics were obtained through the extraction of the optimal straightline segment and the optimal edge direction,combining with the strong edges of the image.The algorithm was tested in the LIVE database and security monitoring video database,and the correlation analysis was also conducted.Experimental results show that the video evaluation data obtained through the proposed method has good consistency with the subjective assessment,meeting the actual needs of the security monitoring.%为检测由摄像机参数设置不当引起的安防监控视频质量退化问题,提出基于最优直线段边缘方向的视频清晰度检测方法.基于监控视频图像的模糊造成全局的图像质量下降,但在局部区域的表现各不相同的特性,通过最优直线段和最优边缘方向的提取,结合图像强边缘宽度图,得到精准的视频图像清晰度度量指标.算法在LIVE库和安防监控视频库上进行实验,并进行相关性分析,实验结果表明,该方法得到的视频评估数据与主观评价具有良好的一致性,满足安防监控的实际需求.