Position: Home > Articles > Vehicle Illumination Denoising Algorithm Based on Improved Non-local Mean Filtering under Low Illumination
Journal of Jinling Institute of Technology
2018
(4)
15-20
基于改进非局部均值滤波的低照度下车牌去噪算法
作 者：
孙克雷;陈慷;田锦;刘静
单 位：
安徽理工大学计算机科学与工程学院;金陵科技学院网络与通信工程学院
关键词：
低照度;直方图均衡化;Shen-Castan;非局部均值;去噪处理
摘 要：
低照度下车牌图像可见度低且含高斯噪音,以传统非局部均值算法处理时,容易丢失图像边缘信息。提出了一种基于改进非局部均值的低照度下车牌去噪算法,以标准化欧氏距离代替简单欧氏距离度量块相似性,同时引入Shen-Castan边缘检测器对车牌边缘部分分离去噪。实验结果表明:提出的算法相较于传统去噪算法,处理结果有更好的信噪比值,且结构相似性值更接近1,说明该去噪算法有更好的去噪效果,同时也能保护边缘信息的完整性。
译 名：
Vehicle Illumination Denoising Algorithm Based on Improved Non-local Mean Filtering under Low Illumination
作 者：
SUN Ke-lei;CHEN Kang;TIAN Jin;LIU Jing;Anhui University of Science and Technology;Jinling Institute of Technology;
单 位：
SUN Ke-lei%CHEN Kang%TIAN Jin%LIU Jing%Anhui University of Science and Technology%Jinling Institute of Technology
关键词：
low illumination;;histogram equalization;;Shen-Castan;;non-local mean;;denoising processing
摘 要：
Under low illumination,the license plate image has low visibility and contains noise.When processed by the traditional non-local mean denoising algorithm,image edge information is easily lost.This paper proposes a method of denoising the license plate under low illumination based on improved non-local mean.The Euclidean distance is used to replace the simple Euclidean distance metric block similarity.Meanwhile,the ISEF edge detector is introduced to separate and denoise the edge part of the license plate.The experimental results show that the proposed algorithm has better signal-to-noise ratio than the traditional denoising algorithm,and the structural similarity value is closer to 1,indicating that the denoising algorithm has better denoising effect and can also protect the integrity of edge information.
相似文章
-
低照度间歇光照制对AA肉鸡仔鸡生产性能影响的研究 [王淑芝, 纪洪海] 黑龙江畜牧兽医 1991,000 (8) 7-9
