基于NSCT与模糊对比度的图像增强算法  

Image enhancement algorithm based on NSCT and fuzzy contrast

作　　者：王静静[1] 贾振红[1] 覃锡忠[1] 杨杰[2] Nikola Kasabov 

机构地区：[1]新疆大学信息科学与工程学院,新疆乌鲁木齐830046 [2]上海交通大学图像处理与模式识别研究所,上海200240 [3]新西兰奥克兰理工大学知识工程与发现研究所

出　　处：《计算机工程与设计》2015年第10期2754-2758,共5页Computer Engineering and Design

基　　金：教育部促进与美大地区科研合作与高层次人才培养基金项目(20101595)

摘　　要：针对图像在增强的过程中容易放大噪声,造成图像的失真和细节不明显等问题,结合NSCT(non-subsampled contourlet transform)和模糊集理论,提出一种改进算法。利用NSCT的平移不变性和多分辨率分析特点对图像进行非下采样Contourlet变换,对低频系数采用线性增强,用改进的自适应阈值函数来处理高频系数,利用改进的模糊对比度来增强图像的边缘和纹理,使用伽马校正对增强后的图像进行校正,提高全局对比度。实验结果表明,该算法能够明显改善图像的视觉效果,增强图像的细节信息。To solve the problem of enlarging noise and image distortion in the process of image enhancement,a threshold algorithm was proposed combining with NSCT and improved fuzzy set.Translation invariance of the NSCT and multi resolution analysis were used,and the image was decomposed by nonsubsampled contourlet transform.Linear enhancement was used in low frequency sub-band,and the improved adaptive threshold function was used to deal with the high frequency coefficient.The improved fuzzy contrast was used to enhance the image edge details and the gamma correction was used to enhance the global contrast.Experimental results show that the proposed algorithm can significantly improve the visual effect of images and enhance image details.

关 键 词：NSCT变换 线性增强 自适应阈值函数 模糊对比度 图像细节 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]