多尺度卡尔曼滤波在图像去噪中的应用研究

The Application of Multiscale Kalman Filtering to Image Denoising

基于小波变换的卡尔曼滤波方法已成功应用于信号去噪。根据卡尔曼滤波应用方法和实现形式的不同、结合小波分解的结构分析.将现有的多尺度卡尔曼滤波方法归纳为三类,提出了基于分解层、重构层和系统层的卡尔曼滤波概念,并指出了各自的优缺点。仿真实验表明,基于系统层的多尺度卡尔曼滤波方法具有较好的性能。

Kalman filtering technology based on wavelet transform has been successfully applied in signal processing.Ac- cording to the implement ways of Kalman filters and analysis of wavelet transform,the current methods of Kalman filte- ring in wavelet domain are generalized and categorized into three kinds; the analysis-level,synthesis-level and system -level Kalman filter are defined.Simulation results indicate that Kalman filtering method based on system--level in wavelet domain has superior performance to the other two.