微光图像麻点噪声滤波理论与实验研究

Theory and Experiment on Low Light Level Image by Snow Noise Filter

均值滤波法处理微光图像时会造成严重的图像模糊。该文对此做了进一步的研究与实验，采用一种新的改进性处理措施——麻点噪声滤波法抑制微光图像噪声。该方法是将滤波窗口内中间像素点灰度值与所有像素点灰度平均值的绝对差值同阈值进行比较，并根据比较结果决定滤波输出。实验结果表明，麻点噪声滤波能有效地抑制微光图像噪声，对图像边缘保护特性比均伍滤波好。

When low light level(LLL)image is processed by average filter, indistinctness may be caused seriousiy.In this paper, further studies and experiments are carriedout. And a new improved method,snow noise filter,is put forward and taken to suppressthe LLL image noise。This method is comparing the absolute difference between the greyscale value of the central pixel and the grey scale mean value of all pixels inside the fiItering wiridow with the threshold valtie,and determining the filtering output according to thecomparing result. The experimental results show that snow noise filter can suppress theLLL image noise effectively; And as,for the protection to the image edge, the property ofsnow noise filter is better than that of average filler.