多目标跟踪下交通监控失真视频增强算法研究  

Research on Traffic Monitoring Distorted Video Enhancement Algorithm under Multi Target Tracking

作　　者：孙雅娟 孙汉卿 冯志慧[2] SUN Ya-juan;SUN Han-qing;FENG Zhi-hui(College of Information Engineering,Henan University of Animal Husbandry and Economy,Zhengzhou Henan 450046,China;College of Information and Management Science,Henan Agricultural University,Zhengzhou Henan 450002,China)

机构地区：[1]河南牧业经济学院信息工程学院(软件学院),河南郑州450046 [2]河南农业大学信息与管理科学学院,河南郑州450002

出　　处：《计算机仿真》2025年第1期162-166,共5页Computer Simulation

基　　金：2021年度河南省重点研发与推广专项(科技攻关)(212 102210512);河南牧业经济学院重点科研创新基金(XKYCXJJ 2020010)。

摘　　要：在对交通监控失真视频增强过程中,由于受较多高斯噪声、色彩失衡、外界因素的干扰,导致出现增强效果差、效率低等问题。为此,提出多目标跟踪下交通监控失真视频增强算法研究。采用基于K-L变换算法校正视频色彩,采用基于粒子滤波的多目标跟踪方法获取目标的实时信息,避免增强过程忽略重要的汽车与行人。采用Retinex视频增强算法将交通监控失真视频分解为照度分量与反射分量,通过照度分量的降噪与反射分量的增强共同完成交通监控失真视频增强。仿真结果表明,所提算法的增强效果好,且效率高、均方根对比度较大。In the process of enhancing distorted traffic surveillance videos,the interference of Gaussian noise,color imbalance and external factors often lead to poor enhancement effect and low efficiency.To solve these problems,this article presented an algorithm of enhancing distorted traffic surveillance video under multi-objective tracking.Firstly,K-L transform algorithm was used to correct the video color,and then the multi-objective tracking method based on particle filter was used to obtain the real-time information of target,thus avoiding ignoring important cars and pedestrians during the enhancement process.Moreover,Retinex algorithm was adopted to decompose the distorted video into illumination component and reflection component.Finally,the enhancement of distorted video was completed by denoising the illumination component and enhancing the reflection component.The simulation results show that the proposed algorithm has good enhancement effect,high efficiency and large root-mean-square contrast.

关 键 词：视频色彩校正 基于粒子滤波的多目标跟踪 照度分量 反射 降噪 

分 类 号：TN918.81[电子电信—通信与信息系统]