基于提升小波变换的红外图像双重滤波算法

为了有效抑制红外图像中的随机噪声,采用一种基于提升小波变换的双重滤波算法来进行处理。该算法对含有噪声的红外图像实现第1次提升小波分解,然后对获得的低频和高频分解系数再次实现提升小波变换,舍弃由低频系数经过第2次提升小波变换后获得的低频系数以及由高频系数经过第2次提升小波变换后获得的高频系数。对剩余的高频系数和低频系数分别采用改进阈值函数模型以及改进非局部均值滤波算法进行处理,在此基础上实现小波系数重构。为了改善滤波后图像视觉效果,再引入直方图均衡化算法进行处理。通过理论分析和实验验证,获得了相关的标准测试图像和红外图像测试结果以及峰值信噪比和结构相似度测试数据。结果表明,该滤算法对于高质量地去除红外图像中的噪声是有帮助的。

Double filtering algorithm of infrared images based on lifting wavelet transform

In order to filter the random noise in infrared images effectively, new double filtering algorithm was proposed based on lifting wavelet transform. Firstly, the noise infrared image was decomposed with lifting wavelet at first time. And then, the obtained high-frequency and low-frequency wavelet coefficients were decomposed with lifting wavelet transformation again. The improved threshold function model and nonlocal mean filter algorithm were used to filter the noise of lifting wavelet coefficients. Finally, histogram equalization algorithm was introduced to improve the visual effect of the filtering image. The standard test images, the experimental infrared images, peak signal-noise-ratio (PSNR) and structural similarity (SSIM) were obtained. The results show that, the performance of the algorithm in this paper is good to deal with noise infrared images.