广东石油化工学院学报2025，Vol.35Issue(6)：74-79,85,7.DOI:10.26962/j.cnki.1991.2025.0055
基于广义规范化稀疏模型的监控图像去雾算法
Algorithm of Monitoring Image Defogging Based on Generalized Normalized Sparse Model
摘要
Abstract
In response to the problems of image distortion and long processing time in traditional image dehazing algorithms,a surveillance image dehazing algorithm based on the generalized normalized sparse model is proposed.The constraint function of haze-free surveillance images is used as the processing target for hazy images.The Tikhonov regularization generalized normalized sparse surveillance image prior blind dehazing algorithm is introduced.This algorithm serves as the prior constraint for the blur kernel and the haze-free surveillance image,and the constraint function of the haze-free image is calculated.The dark channel rule of the hazy surveillance image is obtained through the dark channel prior.The sparse features of the hazy sample images are extracted,and a medium transmission map energy model with the same features is established.Combining with the model,the objective function of the blur kernel is solved according to the objective function and rules of the haze-free surveillance image.The dehazing processing of surveillance images of various sizes is realized through detail supplementation.Ten groups(10 images per group)of surveillance images from the standard surveillance image dataset are randomly selected for experiments.The results show that after being processed by this algorithm,the edges of the images are clear and have a strong sense of hierarchy.The dehazing effect is relatively ideal,and the dehazing time is short.关键词
广义规范化/稀疏模型/监控图像/图像去雾/介质传输图Key words
generalized normalization/sparse model/monitor image/image defog/dielectric transmission diagram分类
信息技术与安全科学引用本文复制引用
鲍绍武,王佳佳..基于广义规范化稀疏模型的监控图像去雾算法[J].广东石油化工学院学报,2025,35(6):74-79,85,7.基金项目
安徽省科学研究项目重点项目(2023AH052663) （2023AH052663）
安徽工商职业学院科学研究重点项目(SK2023B003) （SK2023B003）
