基于频域的图像增强技术的改进算法  

An Improved Image Enhancement Technology Algorithm Based on Frequency Domain

作　　者：袁小平 张侠 张毅 崔棋纹 闫泽宇 YUAN Xiao-ping;ZHANG Xia;ZHANG Yi;CUI Qi-wen;YAN Ze-yu(School of Information and Control Engineering,China University of Mining and Technology,Xuzhou 221116,China)

机构地区：[1]中国矿业大学信息与控制工程学院

出　　处：《科学技术与工程》2019年第31期246-250,共5页Science Technology and Engineering

基　　金：科技部科技支撑项目(2013BAK06B08)资助

摘　　要：在传输有限的情况下完成图像压缩传输,针对整体偏暗,低对比度图像增强后的主观视觉质量较差问题,提出了一种改进的基于离散余弦变换(discrete cosine transform,DCT)域块分类的图像增强算法。改进后的图像增强算法基于对图像块分类,对每个图像块判定类别,针对平滑块、纹理块、边沿块,分析图像块交界处块效应产生的原因并根据图像块背景特点,结合原CES(color image enhancement by scaling)算法并改进,对不同类型图像块采取合适的改进的图像增强算法,达到图像增强的目的,提高图像的主观视觉质量。仿真结果表明,改进的算法,以牺牲一些算法计算复杂度,针对纹理块和边沿块获得很好的图像增强效果,同时,能够最大限度地抑制块效应,大大地提高了图片的视觉效果。To complete image compression transmission with limited transmission,an improved image enhancement algorithm based on discrete cosine transform(DCT)domain block classification was proposed to solve the problem of poor subjective visual quality after image enhancement with overall dark and low contrast.The improved image enhancement algorithm based on classification of image block,for each image block to determine category,in view of the smooth,texture and edge blocks,to analyze the causes of image block junction block effect based on image block background characteristics,combined with the original color image enhancement by scaling(CES)algorithm and improved,and the different types of image block take appropriate improved image enhancement algorithm,achieve the goal of image enhancement,improve the image of the subjective visual quality.Simulation results show that the improved algorithm,at the expense of some algorithm computational complexity,can achieve a good image enhancement effect for texture blocks and edge blocks at the same time,can minimize the block effect,greatly improve the image visual effect.

关 键 词：压缩传输 图像增强 离散余弦变换 块效应 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]