基于非线性调整的伽马校正图像增强算法  

Gamma-corrected image enhancement algorithm based on non-linear adjustment

作　　者：朱铮涛 萧达安 ZHU Zheng-tao;XIAO Da-an(Sehool of Mechanical and Electrical Engineering,Guangdong University of Technology,Guangzhou 510006,China)

机构地区：[1]广东工业大学机电工程学院,广东广州510006

出　　处：《计算机工程与设计》2018年第9期2822-2826,2866,共6页Computer Engineering and Design

基　　金：国家自然科学基金面上基金项目(61471134)

摘　　要：针对由多种降质因素导致图像增强效果不理想的情况,提出一种参数化的伽马校正算法实现图像的增强,即整幅图像采用变化的伽马值进行幂律变换。图像处理时,将图像中的每个像素点与平均亮度比较划分为亮点、暗点和平滑点,将像素分类有关的非线性变换函数作为传统伽马校正中幂律值计算的参数,获得参数化的伽马校正增强函数。实验结果表明,对过曝光、曝光不足以及混合多种性质的复杂图像,该方法均有较好的增强效果,鲁棒性较好,在保持输入图像亮度的情况下,能够自适应处理各种质量较低的图像。Aiming at the situation that the image enhancement effect is not ideal due to various degradation factors,a parametric gamma correction algorithm was proposed to enhance the image,namely,the whole image was transformed using the gamma value.When the image was processed,the image of each pixel and the average brightness comparison were divided into bright spots,dark spots and smooth points,the nonlinear transformation function related to the pixel classification was taken as the parameters of power law value in traditional gamma correction,and the parametric gamma enhancement function was determined.Experimental results show that the proposed method brings about better enhancement effects on the over-exposure,underexposed and mixed complex images,and its robustness is better.In the case of maintaining the brightness of the input image,it can adaptively deal with a variety of low quality images.

关 键 词：降质图像 伽马校正 像素分类 非线性变换 图像增强 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]