一种彩色图像的同步去噪增强算法  

Synchronization Algorithm of Color Image Denoising and Enhancement

作　　者：贾迪[1] 孟琭[2] 张一飞[2] 贺学平[1] 方金凤[1] 

机构地区：[1]辽宁工程技术大学电子与信息工程学院,辽宁葫芦岛125105 [2]东北大学信息科学与工程学院,沈阳110819

出　　处：《小型微型计算机系统》2014年第3期659-662,共4页Journal of Chinese Computer Systems

基　　金：国家自然科学基金项目(61101057)资助

摘　　要：在彩色图像获取过程中,由于受到环境、设备等客观因素的制约,常得到含有噪声、对比度差的图像,这些对于后续图像分割与配准的准确性方面都产生较大影响,为此提出一种彩色图像的同步去噪增强模型.首先引入RGB与Ycbcr的变换及逆变换,构造了适用于彩色图像的自适应直方图均衡化偏微分方程.其次,通过计算本征矢与x轴的夹角统一梯度变化率、消除法相扩散,给出了彩色图像的PM模型.最后将上述两种模型通过增加系数调节项的方式进行融合,达到同步去噪增强的目的,实验结果验证了本文算法的有效性与实用性.In the process of color image acquisition, due to the objective factors of the environment and equipment etc, the image which contains noise or poor contrast is often obtained. These have a greater impact for the subsequent image segmentation and regis- tration accuracy. For this reason, the paper proposes a synchronization model of color image denoising and enhancement. Firstly, RGB and Ycbcr is introduced to construct adaptive histogram equalization partial differential equations by transform and inverse trans- form. Secondly, the rate of changing gradient is unified by the calculation of the eigenvectors with the x axis to improve color PM model. Finally, the purpose of denoising and enhancement is achieved by combining the two items and increasing the coefficient of adjustment parameters. The results of experiment show that algorithm has validity and practicality.

关 键 词：图像去噪 图像增强 偏微分方程 直方图均衡 方向扩散 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]