基于噪声特征和矢量中值滤波的彩色图像去噪算法

Color image denoising based on noise characteristic and vector median filtering

为了有效去除彩色图像噪声,同时保留图像细节信息,并提高处理速度,在分析彩色图像噪声特征的基础上,探索了基于噪声点的距离测度特性的判别算法,并结合改进的矢量中值滤波(VMF)算法对检测出的噪声点予以降噪处理。试验结果表明:该方法能够较好地抑制噪声,有效地保留了图像细节信息,且所用时间仅为传统矢量中值滤波算法的1/2左右。

In order to effectively denoise the color image, while preserving image detail information, and increase processing speed, a new noise pinpointing method based on noise characteristic was proposed according to the analysis of noise characteristic of color images. Combined with the modified Vector Median Filtering （VMF） algorithm, the detected noise was denoised. Experimental results show that the algorithm can preserve the detail of image very well when restraining noise effectively, and its computation time is only half of the classical vector median filtering algorithm.