红外图像 WBC 变换方向内和尺度间降噪

摘要

As traditional denoising algorithm can’t deal with the infrared image with noise signal well,an infrared im-age denoising algorithm of neighborhood coefficient correlation and inter-scale dependency based on WBC transform is proposed by analyzing infrared image’s correlation of the neighborhood coefficient after the WBC transform.Firstly, the generalized CB morphology is applied to neighborhood coefficient denoising.Then,infrared images are denoised combining with the self-adaptive threshold based on inter-scale dependency.The results show the proposed algorithm can get higher SNR compared with Bayes estimation denoising method,WBC inter-scale hard-threshold denoising method and WBC inter-scale adaptive threshold denoising method.Its SSIMis also closer to 1 .%针对传统图像降噪算法不能很好处理含噪红外图像的问题，本文分析红外图像在WBC 变换后邻域系数相关性将广义 CB 形态学应用于邻域系数降噪，并结合基于尺度间相关性的自适应阈值降噪，提出一种基于 WBC 变换邻域系数相关性和尺度间相关性的红外图像降噪算法。仿真结果表明，该方法相比 Bayes 估计降噪、WBC 尺度间硬阈值降噪和 WBC 尺度间自适应降噪获得了更好的 SNR 提升，其 SSIM值也更接近于1。