基于NSCT变换的红外图像增强新算法 |
| |
引用本文: | 严红丽,石永华,孙凯传.基于NSCT变换的红外图像增强新算法[J].重庆科技学院学报(自然科学版),2017(6):120-124. |
| |
作者姓名: | 严红丽 石永华 孙凯传 |
| |
作者单位: | 1. 滁州学院电子与电气工程学院,安徽 滁州,239000;2. 滁州学院计算机与信息工程学院,安徽 滁州,239000 |
| |
基金项目: | 安徽省高校优秀拔尖人才培育资助项目 —— 高校优秀中青年骨干人才国内外访学研修重点项目 |
| |
摘 要: | 针对夜间红外图像噪声大、对比度低的问题,探讨一种基于正态分布特性和NSCT变换相结合的夜间红外图像增强算法。充分利用了高频子带的系数标准差、系数均值和每个子带系数最大值的构造自适应阈值,保护图像的边缘细节,抑制图像噪声;提出一种新的弱边缘系数增益函数,有效地提升了图像的弱边缘信息。低频系数采用了基于正态分布特性的对比度拉伸增强算法。
|
关 键 词: | 红外图像增强 正态分布 图像边缘 自适应阈值 |
Infrared Image Enhancement Algorithm Based on the NSCT Transform |
| |
Abstract: | Because of the low contrast and loud noise of night infrared image,the infrared image enhancement meth-od is presented in this paper,which is based on normal distribution characteristics and NSCT transform. By making full use of the high frequency sub-band coefficient of standard deviation,coefficient of mean and each sub-band coefficient maximum value for the construction of adaptive threshold,the image edge details are protected to better suppress the image noise. It also puts forward a new weak edge gain coefficient function,so weak edge information of the image can be effectively enhanced. The low frequency coefficient is based on the normal distribution charac-teristics and the contrast is enhanced. |
| |
Keywords: | infrared image enhancement normal distribution edge of image adaptive threshold |
本文献已被 CNKI 万方数据 等数据库收录! |
|