提出了一种改进的保持形状的图像对比度增强算法.改进包含两个方面：1) 修改并推广了原有的基于PDE的灰度增强方法,使之可适用于任何设定的灰度拉伸函数;2)给出了一种通用的分段线性拉伸函数设计方法,它较采用累积直方图作为灰度变换函数可达到更好的直方图均衡化效果.还提出了一种松弛阈值化方法用以消除局部反差增强时可能出现的“斑点效应”.实验结果表明,本文提出的改进算法适用于多种自然图像,增强效果良好.

An improved algorithm for shape preserving contrast enhancement was presented. The improvements consist of two parts: 1) generalizing the PDE based approach to contrast enhancement to adapt to any specified stretch function for global and local contrast enhancement; 2) proposing a piecewise linear stretch function design method to achieve better histogram equalization. A relaxed thresholding scheme was proposed to reduce “spot” effects. Experiments show that the algorithm is adapted to a variety of real images and yields satisfactory results.