关键词: 图像去噪, 椒盐噪声, 噪声检测, 中值滤波, 检测窗口

Abstract: To remove the salt and pepper noise in images effectively while preserving the detail, a salt and pepper noise detection algorithm based on directional median was proposed after analyzing merits and drawbacks of different slat and pepper noise detection algorithms. The algorithm divided the image into doubtful noise pixels and signal pixels in the first step. At the second step, the algorithm set nine directional detection areas in the 5×5 detection window centered on the doubtful pixels, and confirmed the noise pixels after comparing gray value of doubtful noise pixels and median gray value of the detection areas. The feasibility of the leak detection in the algorithm reduced load of subsequent processing. At the same time, the algorithm possessed low false detection rate which preserved the detail. The simulation results show the validity of the algorithm.

Key words: image denoising, salt and pepper noise, noise detection, median filter, detection window