基于视觉传达机理的低光照图像增强方法  

Low light image enhancement method based on visual communication mechanism

作　　者：郑哲孚 钟健 ZHENG Zhefu;ZHONG Jian(School of Computer Science and Engineering,Guangzhou Institute of Science and Technology,Guangzhou 510540,China;Information&Network Center,Guangzhou Institute of Science and Technology,Guangzhou 510540,China)

机构地区：[1]广州理工学院计算机科学与工程学院,广州510540 [2]广州理工学院信息与网络中心,广州510540

出　　处：《激光杂志》2025年第4期103-108,共6页Laser Journal

基　　金：广东省普通高校特色创新类基金资助项目(No.2020KTSCX292);广州市科技计划项目(No.202102080235);广州市高等教育教学质量与教学改革工程名师工作室项目(No.2022MSGZS017)。

摘　　要：低光照影响图像的质量,针对当前低光照图像增强方法无法获得理想结果的问题,提出基于视觉传达机理的低光照图像增强方法。该方法利用平滑滤波消除原始低光照图像噪声,根据人眼视觉特性对低光照图像进行增强处理,并通过脉冲耦合神经网络(Pulse Coupled Neural Network,PCNN)模型优化图像细节。仿真结果表明,明显提高了低光照图像视觉效果,低光照图像增强后的G梯度均值为0.0078,PSNR为23.94 dB,SSIM达到0.852,均优于对比方法。该方法有效保留了图像边缘细节,降低了噪声和失真,保持了图像结构信息的完整性,从而提升了低光照图像的整体质量和视觉效果。Low light affects the quality of images.In response to the problem that current low light image enhancement methods cannot achieve ideal results,a low light image enhancement method based on visual communication mechanism is proposed.This method utilizes smooth filtering to eliminate noise in the original low light image,enhances the low light image based on human visual characteristics,and optimizes image details through a Pulse Coupled Neural Network(PCNN)model.The simulation results show that this paper significantly improves the visual effect of low light images.The average G gradient after low light image enhancement is 0.0078,the PSNR is 23.94 dB,and the SSIM reaches 0.852,all of which are better than the comparison methods.This method effectively preserves image edge details,reduces noise and distortion,maintains the integrity of image structural information,and thus improves the overall quality and visual effect of low light images.

关 键 词：视觉传达 低光照图像 增强方法 平滑滤波 

分 类 号：TN279[电子电信—物理电子学]