| 多分辨率域HMT图像去噪增强算法 |
| |
| 引用本文： | 尚政国,赵春晖,刘金梅.多分辨率域HMT图像去噪增强算法[J].哈尔滨工程大学学报,2008,29(9). |
| |
| 作者姓名： | 尚政国 赵春晖 刘金梅 |
| |
| 作者单位： | 哈尔滨工程大学,信息与通信工程学院,黑龙江,哈尔滨,150001 |
| |
| 基金项目： | 国家自然科学摹金资助项目，高等学校博士学科点专项科研项目，黑龙江省自然科学基金 |
| |
| 摘 要： | 为了同时削弱wavelet的伪吉布斯现象以及Contourlet的划痕效果,该文根据多分辨率分析原理.在Wavelett域与Contourlet域建立统一的隐马尔可夫树(HMT)去噪模型,实现了对图像的有效去噪与细节增强.该方法具有多向选择性、图像信息并行处理、信息利用率高、多频率图像融合增强等特点.通过仿真实验与Wavelet、Contourlet和Wavelet HMT等去噪算法进行比较,验证了该方法的有效性和优越性.
|
| 关 键 词： | 多分辨率分析 隐马尔可夫树 小波变换 轮廓波变换 图像去噪 |
An image denoising and enhancement method based on multi-resolution hidden Markov tree |
| |
| SHANG Zheng-guo,ZHAO Chun-hui,LIU Jin-mei.An image denoising and enhancement method based on multi-resolution hidden Markov tree[J].Journal of Harbin Engineering University,2008,29(9). |
| |
| Authors: | SHANG Zheng-guo ZHAO Chun-hui LIU Jin-mei |
| |
| Abstract: | A unified hidden markov tree(HMT) denoising model in the Wavelet and Contourlet domains has been developed by using multi-resolution principle.It effectively reduces the pseudo-Gibbs effect found in Wavelet transforms and the scratching effect in Contourlet transforms,enhancing the image details and denoising effect.The method possesses multidirectional options,parallel processing ability for image information,high utility rate of information and enhanced image fusion capability in different frequencies.Through a series of simulations,we proved the effectiveness and superiority of the new method after comparing it with Wavelet,Contourlet and Wavelet-HTM methods. |
| |
| Keywords: | multi-resolution analysis hidden Markov tree Wavelet transform Contourlet transform image denoising |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
