自适应彩色图像SCM去噪

Self-Adaptive Method Using SCM for Noise Removal in Color Images

提出一种彩色图像去噪方法。该方法首先利用脉冲发放皮层模型(spiking cortical model,SCM)的同步脉冲发放特性和脉冲噪声明显不同于周围非噪声像素的特点定位彩色图像各通道中脉冲噪声点的位置;然后根据噪声检测情况自适应地选择窗口的大小及对应的滤波方法仅滤除噪声像素,保持非噪声像素不变。实验数据表明该方法能够充分去除噪声的同时很好地保持图像细节。对污染严重的图像,用该方法去噪后的主观视觉和客观评价明显有所改善。

A new algorithm for reducing impulsive synchronous pulse burst property of spiking cortical model noise in color images is presented. Firstly, the （SCM） and the characteristic that impulse noise is significantly different from surrounding image pixels are introduced to locate noises in each channel of color images. Secondly, according to the number and position of detected noises, just the noises are filtered by the properly selected window and related filtering methods, and free-noise pixels are kept unchanged. Experimental data show that the proposed method can preserve more image details while sufficiently remove noises sufficiently. The good visual effect and objective results are achieved for color images corrupted at heavy level.