摘要： 针对彩色图像脉冲噪声的分布特性，提出噪声检测和滤波复原的2步算法。该算法通过长针统计检测技术判断脉冲噪声位置，用改进适应性中值滤波技术复原图像。实验表明，与已有算法相比，该算法能有效去除彩色图像中的高水平脉冲噪声，噪声密度大于80%时仍有良好性能，且算法实时性好，适宜于在线处理。

关键词: 脉冲噪声, 长针统计检测, 适应性中值滤波

Abstract: Considering the distribution of impulse noises of color images, a two steps algorithm is proposed. The algorithm uses spike statistical detection to detect noise pixels, and rebuilds image using adaptive median filter. Experimental results show that the proposed algorithm works effectively in denosing high level impulse noises although the noise density is higher than 80%, compared with other existing algorithms. It is suitable for online processing tasks, as its lower time-cost.

Key words: impulse noise, spike statistical detection, adaptive median filter

中图分类号: 

• TP391.41