Abstract: In order to effectively extract the key frame from the road monitoring video photographed by fixed cameras, the algorithm of extracting key frame based on correlation coefficient was proposed. Firstly, the RGB values of the video monitoring video were translated into gray values by the proposed algorithm. Secondly, combining with the sliding window, whether the current frame was the key frame was determined by comparing the threshold and absolute value between the correlation coefficient of the current frame and the current key frame. Experimental results show that the algorithm can not only extract the key frame effectively, but also has the clear process and can avoid redundant key frame. Finally, the extracted key frame was evaluated by the frame difference based on Euclidean distance, which verified the validity of the algorithm and the correctness of the key frame.

Key words: traffic and transportation engineering, key frame, correlation coefficient, sliding window, road monitoring

摘要： 为从固定摄像头的道路监控视频中有效地提取出关键帧，提出了基于相关系数的关键帧提取算法。算法将监控视频中图像帧的RGB值处理为灰度值，结合滑动窗口机制，通过当前帧和当前关键帧之间相关系数的绝对值与阈值比较判断当前帧是否为关键帧。实验结果表明：算法在有效提取关键帧的同时还具有流程清晰、避免了冗余关键帧的特性。最后利用基于欧式距离的帧间差异性对提取的关键帧进行评价，验证了算法的有效性和合理性。

关键词: 交通运输工程, 关键帧, 相关系数, 滑动窗口, 道路监控