蓝章礼, 帅丹, 李益才. 基于帧间相关性的道路监控视频关键帧提取[J]. 微电子学与计算机, 2015, 32(5): 51-54. DOI: 10.19304/j.cnki.issn1000-7180.2015.05.011
引用本文: 蓝章礼, 帅丹, 李益才. 基于帧间相关性的道路监控视频关键帧提取[J]. 微电子学与计算机, 2015, 32(5): 51-54. DOI: 10.19304/j.cnki.issn1000-7180.2015.05.011
LAN Zhang-li, SHUAI Dan, LI Yi-cai. Key Frame Extraction of Road Monitoring Video Based on Inter-Frame Correlation[J]. Microelectronics & Computer, 2015, 32(5): 51-54. DOI: 10.19304/j.cnki.issn1000-7180.2015.05.011
Citation: LAN Zhang-li, SHUAI Dan, LI Yi-cai. Key Frame Extraction of Road Monitoring Video Based on Inter-Frame Correlation[J]. Microelectronics & Computer, 2015, 32(5): 51-54. DOI: 10.19304/j.cnki.issn1000-7180.2015.05.011

基于帧间相关性的道路监控视频关键帧提取

Key Frame Extraction of Road Monitoring Video Based on Inter-Frame Correlation

  • 摘要: 为提取道路固定摄像头监控视频的关键帧,提出了基于帧间相关性的关键帧提取算法.算法将监控视频中的彩色图像帧预处理为灰度图像帧,结合滑动窗口机制,通过帧间相关性与帧自相关性的相对偏差与阈值θ的比较判定关键帧.实验结果表明,算法流程清晰,能够有效提取与目测一致的关键帧,同时关键帧的灵敏程度可以通过阈值θ调节.

     

    Abstract: An algorithm of key frame extraction is proposed by using the inter-frame correlation, which is used to extract key frames for road monitoring video photographed by fixed cameras. Firstly, the pre-processing is utilized to operate frames, and then the color frames change to gray frames. Secondly, the sliding window is combined with the gray frames to determine key frames through comparing the threshold with the relative deviation of the inter-frame correlation and autocorrelation. The experimental results show that the process of the algorithm is clear, and the key frame consistent with the visual can be extracted effectively and the level of sensitivity adjusts to the threshold well.

     

/

返回文章
返回