一种基于引导滤波的图像分层增强方法  

A Hierarchical Image Enhancement Method Based on Guided Filter

作　　者：赵柏山[1] 丛喜龙 张帆 ZHAO Baishan;CONG Xilong;ZHANG Fan(School of Information Science and Engineering,Shenyang University of Technology,Shenyang 110870,China)

机构地区：[1]沈阳工业大学信息科学与工程学院,沈阳110870

出　　处：《微处理机》2021年第5期26-28,共3页Microprocessors

基　　金：辽宁省自然科学基金(No.2019-ZD-0213)。

摘　　要：针对红外图像普遍存在对比度低、噪声大、细节信息微弱等问题,为满足红外图像兼顾抑制噪声、提高图像对比度和突出边缘的需要,提出一种基于引导滤波的图像分层增强算法。算法对图像中的像素点进行判定,区分出噪声点、背景点和边缘点,进行有选择的模板滤波;利用引导滤波将图像分层,并采用双向直方图均衡化提升背景层图像的对比度;利用Gamma变换增强图像细节层,得到一幅经由加权合成的图像。经与其他图像增强算法综合对比,该算法能够有效提升图像对比度、突出边缘细节,获得良好的图像增强效果。Aiming at the problems of low contrast,high noise and weak detail information in infrared images,a hierarchical image enhancement algorithm based on guided filtering is proposed to meet the needs of suppressing noise,improving image contrast and highlighting edges in infrared images.The algorithm judges the pixels in images,distinguishes noise points,background points and edge points,and carries out selective template filtering.Images are layered by guided filtering,and the contrast of the background layer images are improved by bidirectional histogram equalization.Using Gamma transform to enhance the detail layer of images,a weighted composite image is obtained.Compared with other image enhancement algorithms,this algorithm can effectively improve image contrast,highlight edge details,and obtain good image enhancement effect.

关 键 词：红外图像 引导滤波 图像分层增强 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]