基于双向扩散机制的遥感图像去噪研究  

Remote Sensing Image Denoising Research Based on Bidirectional Diffusion Mechanism

作　　者：张龙龙[1] 李国辉[1] 吴成茂[1] 

机构地区：[1]西安邮电大学电子工程学院,西安710121

出　　处：《激光杂志》2016年第7期15-19,共5页Laser Journal

基　　金：国家自然科学基金重点资助项目(61136002);陕西省自然科学基金项目(2014JM8331;2014JQ5138)

摘　　要：为改善图像去噪和边缘保留效果,针对基于Demons算法的去噪模型中仅包含单向扩散的不足,将Active Demons算法和Symmetric Demons算法的双向扩散机制引入去噪模型,提出两种基于Active Demons算法和Symmetric Demons算法的改进去噪模型。实验结果表明,两种改进模型均能取得比Perona-Malik模型和基于Demons算法的模型更好的去噪效果,所得结果图像具有更高的峰值信噪比和清晰度。In order to improve the effects of image denoising and edge retention, in view of the problem that only the unidirectional diffusion is existed in the denoising model based on Demons algorithm, two modified denoising models based on the Active Demons algorithm and the Symmetric Demons algorithm are proposed by introducing the bidirectional diffusion mechanism of the Active Demons algorithm and the Symmetric Demons algorithm into the denoising model. Experimental results indicate that, compared with the Perona-Malik model and the model based on Demons algorithm, two modified denosing models proposed by this paper can achieve better denoising performance and the resulting images obtained by them have the higher Peak Signal to Noise Ratio and Definition.

关 键 词：图像处理 图像去噪 双向扩散机制 ACTIVE DEMONS算法 SYMMETRIC DEMONS算法 偏微分方程 

分 类 号：TN391[电子电信—物理电子学]