波谱学杂志 ›› 2007, Vol. 24 ›› Issue (4): 462-468.

• 祝贺邬学文先生八十华诞(专辑) • 上一篇    下一篇

结合小波域变换和空间域变换的图像增强方法

徐凌,刘薇,杨光*   

  1. 上海市功能磁共振成像重点实验室,华东师范大学 物理系,上海 200062
  • 收稿日期:2007-08-28 出版日期:2007-12-05 发布日期:2007-12-05
  • 作者简介:徐凌(1983-),女,上海人,硕士,无线电物理专业.*通信联系人:杨光,电话:021-62233293,E-mail:gyang@phy.ecnu.edu.cn.
  • 基金资助:

    上海市科委资助项目(06DZ22922), (06DZ22014).

An Image Enhancement Algorithm Combining Wavelet Domain and Image Domain Processing

XU Ling, LIU Wei,YANG Guang*   

  1. Shanghai Key Laboratory Functional Magnetic Resonance Imaging, Department of Physics, East China Normal University,Shanghai 200062, China
  • Received:2007-08-28 Online:2007-12-05 Published:2007-12-05
  • About author:*Corresponding author: Yang Guang,Tel:021-62233293,E-mail:gyang@phy.ecnu.edu.cn.
  • Supported by:

    上海市科委资助项目(06DZ22922), (06DZ22014).

摘要:

提出了一种结合了小波域和空间域处理方法的图像增强算法. 该算法首先对小波域中的高频系数进行修正,使图像具有更好的局部对比度和更丰富的细节,由于双树复小波变换(Dual-tree Complex Wavelet Transform,DT CWT)具有更好的方向选择性,在小波变换的过程中选用了这一方法;然后,通过空间域中的非线性变换,调整图像的整体对比度. 该算法可根据图像本身的特性实现参数的自动选择. 经过本文方法的处理所得的图像,无论在视觉效果上还是在统计上,都优于前人工作的结果.

关键词: 图像增强, 小波变换, 双树复小波

Abstract:

A new image enhancement algorithm combining wavelet domain and image domain processing was proposed. Dual-tree complex wavelet transform (DTCWT) was used in the wavelet transform due to its good directional selectivity. To enhance the images, high frequency wavelet coefficients were firstly modified to enhance the local contrast and the details of image. The overall contrast of the image was then adjusted with non-linear transforms in the image domain. The algorithm allows automatic selection of processing parameters according to the characteristics of the raw images. Compared with results reported previously, images processed by this algorithm achieved better results both visually and statistically.

Key words: image enhancement, wavelet transform, dual-tree complex wavelet transform

中图分类号: