人眼可感知最多相邻像素灰度差的全局图像优化方法  

Global Image Optimization with Gray Difference Perceptible by the Most Neighboring Pixels

作　　者：蔡铁峰 CAI Tiefeng(Educational Technology and Information Center,Shenzhen Polytechnic,Shenzhen,Guangdong 518055,China)

机构地区：[1]深圳职业技术学院教育技术与信息中心,广东深圳518055

出　　处：《深圳职业技术学院学报》2020年第3期3-7,28,共6页Journal of Shenzhen Polytechnic

基　　金：深圳职业技术学院校级青年重点资助项目(6018-22k370199991)。

摘　　要：图像优化可以使图像内更多面向探测识别的信息被人眼感知.现有的图像优化方法缺少合理的量化指标,图像没有做到面向人眼探测识别最优,甚至优化后图像人眼可探测识别的景物信息会减少.本文基于图像中景物轮廓与纹理信息主要体现为相邻像素灰度差的认识,以图像中最多相邻像素对灰度差人眼可感知为量化指标,设计全局图像优化快速算法.实验表明,与现有的全局图像优化方法比较,本方法优化后图像有最多的相邻像素灰度差人眼可感知,某种意义上使图像面向人眼探测识别最优.Image contrast optimization can make more detection and recognition information in the image perceived by human eyes.The existing image optimization methods lack reasonable quantitative indicators,and the optimized image does not achieve the optimal detection and recognition for human eyes.Even after optimization,the information of the scene that can be detected and recognized by human eyes could be reduced.In this paper,based on the recognition that the contour and texture information of the scene in the image is mainly reflected in the gray difference of neighboring pixels,the paper designs a fast algorithm for global image optimization with the gray difference perceptible by the most neighboring pixel pairs as the quantitative index.Experimental results show that compared with the existing global image optimization methods,the image optimized by the proposed method has the most neighboring pixel pairs perceptible,which makes the image detection and recognition for human eyes optimal in a sense.

关 键 词：探测识别 图像优化 对比度增强 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]