摘要
局部直方图均衡方法是目前最为普遍有效的图像增强方法,但其不可避免的会带来块效应以及过度增强的效果.而传统的消除块效应方法一般耗时巨大或者效果不佳,针对这种情况,本文提出使用子块不重叠线性/双线性插值算法来消除块效应,并提出使用对比度受限直方图算法来抑制过度增强.本文详细阐述了插值法和对比度受限算法的具体过程;并进行Matlab仿真实验,通过对比分析各种均衡算法的结果和运行时间,证明本文所提出的基于插值和对比度受限算法的局部直方图均衡方法在图像增强的精度和速度方面确实有效可行.
local histogram equalization is themost common and effectivemethod for image enhancement at present, but it will inevitably bring about block effect and over-enhanced effect. However, the traditionalmethods of eliminating block effect are usually time-consuming or not that good enough. Forsolving this problem, this paper proposes to use unoverlappedsubblock linear/bilinear interpolation algorithm to eliminate the block effect, and proposes to apply contrast limited algorithm tosuppress excessive enhancement. This paper describes thespecific process of interpolation and contrast limited algorithm, and alsoshowsmatlabsimulation results. By analyzing and comparing the results and running time of various equalizations, the proposed local histogram equalizationmethod based on interpolation and contrast limited algorithm is proved to be effective and feasible in both terms of the accuracy andspeed of image enhancement.
作者
张鲁筠
韩锴
徐永贵
黄小东
ZHANG Lu-jun;HAN Kai;XU Yong-gui;HUANG Xiao-dong(Weifang University,Weifang,261061,China)
出处
《潍坊学院学报》
2019年第2期1-5,15,共6页
Journal of Weifang University
关键词
图像增强
局部直方图均衡
块效应
子块不重叠
线性/双线性插值
对比度受限算法
image enhancement
local histogram equalization
block effect
unoveriappedsub-block
linear/bilinear interpolation
contrast limited algorithm
