首页 | 官方网站   微博 | 高级检索  
     

基于非下采样Contourlet系数尺度相关性的图像增强算法
引用本文:何力,曲仕茹,张大奇.基于非下采样Contourlet系数尺度相关性的图像增强算法[J].西北工业大学学报,2010,28(1).
作者姓名:何力  曲仕茹  张大奇
作者单位:西北工业大学,自动化学院,陕西,西安,710072
基金项目:博士点基金,陕西省科学技术研究发展计划 
摘    要:针对传统图像增强算法对边缘类型划分过粗的问题,文章提出一种基于非下采样Contourlet系数尺度间相关性的图像增强算法。该算法由非下采样Contourlet单尺度系数得到噪音阈值、增强阈值和强边缘阈值,再基于对非下采样Contourlet系数尺度间相关性的分析,以归一化尺度积与单尺度系数的比值作为新的参数进一步划分边缘类型,并结合单尺度系数得出相应增益。对比测试结果表明算法能有效地提高图像对比度、去除噪音,在避免过增强的同时,对边缘视觉效果的改善更明显。

关 键 词:图像增强  非下采样Contourlet  尺度间相关性

Image Enhancement Based on Inter-Scale Correlations of Nonsubsampled Contourlet Coefficients
He Li,Qu Shiru,Zhang Daqi.Image Enhancement Based on Inter-Scale Correlations of Nonsubsampled Contourlet Coefficients[J].Journal of Northwestern Polytechnical University,2010,28(1).
Authors:He Li  Qu Shiru  Zhang Daqi
Abstract:We represent a novel image enhancement algorithm which is based on nonsubsampled contourlet transform(NSCT).We first use normalized multiscale products and NSCT coefficients to represent the difference between edges in inter-scale correlations of NSCT,and use both the products and the coefficients to distinguish different types of edges and to give different gains respectively.There are three steps in our algorithm:(1) we estimate the noise to determine the noise threshold,the enhancement threshold and the ...
Keywords:image enhancement  nonsubsampled contourlet transform  inter-scale correlation
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司    京ICP备09084417号-23

京公网安备 11010802026262号