基于Curvelet变换的图像自适应增强算法  

Image Enhancement Algorithm Using Adaptive Threshold Based on Curvelet Transform

作　　者：蒋婷婷[1] 

机构地区：[1]中国电子科技集团公司第二十八研究所,南京210007

出　　处：《指挥信息系统与技术》2011年第2期14-18,共5页Command Information System and Technology

摘　　要：Curvelet变换具有各向异性和多方向性等良好特性,是表示分段平滑曲线边缘的最优基。提出了一种基于Curvelet变换的图像自适应增强算法,该算法将图像进行多尺度、多方向的Curvelet变换,引入分段非线性函数的思想,调整系数,提高图像整体的对比度,增强图像细节,并进行阈值降噪。实验表明,该方法优于常用的单阈值增强法和双阈值增强法,能提高图像的对比度、降低噪声,并且较好地保留边缘信息,具有良好的视觉效果。Curvelet transform has anisotropic and multi-directional characteristics. It is the best basis for representing the edge of a piecewise smooth curve. An adaptive image enhancement algorithm based on the Curvelet transform is proposed. By the algorithm, the Curvelet transform is implemented on the image in multiple dimensions and directions. The conception of piecewise non- linear function is introduced. As a result of the transform, the coefficients are adjusted, the overall contrast and the details of the image are enhanced, and the image is denoised by thresholding. Experiments show that the algorithm is better than common single-threshold enhancement and dual-threshold enhancement algorithms. It can effectively improve the image contrast, denoise the image and well retain the edge information, thus it has a good visual effect.

关 键 词：CURVELET变换 小波变换 非线性函数 图像自适应增强 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]