摘要
视频中犯罪嫌疑目标色彩的正确标定对于锁定、追踪犯罪嫌疑目标十分重要。针对光照变换条件下视频中目标与真实目标之间所存在的色彩偏离现象,首先分析了待检验视频相邻帧间整体色彩分布范围的相关性,然后随机选取包含犯罪嫌疑目标的视频关键帧作为参考帧,最后使用基于Retinex的视频目标色彩多重迭代循环校正方法对偏色帧进行校正。实验表明,该方法能够对视频中犯罪嫌疑目标的色彩偏离进行有效的校正。
引文
[1]周尚波,胡鹏,柳玉炯.基于改进Mean-Shift与自适应Kalman滤波的视频目标跟踪[J].计算机应用,2010(6):1573-1576.
[2]孙鹏,刘磊,李海,等.改进的完美反射法及其在视频偏色校正中的应用[J].中国刑警学院学报,2018(3):125-128.
[3]Guo D W,Qian D,Zhen K P,et al.Retinex theory based active contour model for segmentation of inhomogeneous images[J].Digital Signal Processing,2016(50):43-50.
[4]Forsyth D A.A novel algorithm for color constancy[J].International Journal of Computer Vision,1990(1):5-35.
[5]William T F,David H B.Brainard color constancy[J].Journal of the Optical Society of America A Optics image science and Vision,1997(7/8):1393-1411.
[6]任航,张涛.基于灰度投影法运动估计的成像CCD平移补偿法[J].应用光学,2009(3):417-422.
[7]孙鹏,周纯冰,张洪君,等.面向视频侦查应用的监控视频关键帧检测软件设计[J].中国刑警学院学报,2014(4):39-41.
[8]Land,Edwin H.The Retinex Theory of Color Vision[J].Scientific American,1977(6):108-128.
[9]Kimmel R,Elad M,Shaked D,et al.A Variational Framework for Retinex[J].International Journal of Computer Vision,2003(1):7-23.
[10]Gijsenij A,Gevers T,Lucassen M P.Perceptual analysis of distance measures for color constancy algorithms.[J].Journal of the Optical Society of America A Optics image science and Vision,2009(10):2243-2256.
[11]Yang J,Wan L,Qu C.Illumination Processing Recognition of Face Images Based on Improved RetinexAlgorithm[J].Journal of Multimedia,2013(5):541-547.