多尺度域视频监控图像光照分量智能校正方法  

Intelligent Correction Method for Illumination Component of Video Surveillance Image in Multi-Scale Domain

作　　者：刘文春[1] LIU Wenchun(Department of Computer and Information Engineering,Anhui Polytechnic of Industry and Trade,Huainan Auhui 232001,China)

机构地区：[1]安徽工贸职业技术学院计算机信息工程系,安徽淮南232001

出　　处：《重庆科技学院学报（自然科学版）》2024年第1期64-69,共6页Journal of Chongqing University of Science and Technology：Natural Sciences Edition

基　　金：2022年度安徽省科研编制计划项目“高标准农田建设智慧监测管理平台应用研究”(2022AH053024)。

摘　　要：针对视频监控图像中存在的清晰度不足和光照不均匀等问题,提出了多尺度域视频监控图像光照分量智能校正方法。应用背景差分法与差值图二值化法对视频监控图像进行预处理,以减少噪声干扰。通过融合TH变换和BH变换来加强图像多尺度细节特征;利用控制因子控制锐化强度,以凸显细节特征,避免出现图像伪影情况。对图像光照分量进行自适应调节,利用二维伽马函数计算画面亮度平均值,以防止出现过度校正的情况,从而实现智能校正。测试结果表明,采用该方法校正的图像更清晰,画面光照分量均匀,图像失真较少,视觉效果较好。An intelligent correction method for illumination components of video surveillance images multi-scale domain is proposed to address issues such as insufficient clarity and uneven lighting in video surveillance images.Background difference method and difference graph binarization method is applied to preprocess video surveillance images to reduce noise interference and multi-scale detail features of images is strengthened by fusing TH transform and BH transform.The control factor is used to control the sharpening intensity,and the image artifacts are avoided while the details are highlighted.The image light component is adjusted adaptively,and the two-dimensional gamma function is used to calculate the average brightness of the image so as to prevent over-correction so that the intelligent correction is achieved.The test results show that the image corrected by this method is clearer,the light component is uniform,the image is less distorted and the visual effect is better.

关 键 词：多尺度域 视频监控图像 光照分量 光照智能校正 图像预处理 

分 类 号：TN919.8[电子电信—通信与信息系统]