摘要
滤除图像的噪声有很多方法,普遍使用的均值滤波滤除噪声效果不够理想,且易使图像变得模糊;而中值滤波滤除噪声效果较好,但易使图像的细节信息丢失.文章提出了一种基于梯度的滤波算法,该算法对含有噪声的图像取3×3领域,构造4个方向的梯度算子模板,通过比较4个方向梯度大小,对噪声、图像边缘、图像内部信息点分别处理.实验结果表明该算法在滤除噪声的同时,很好地保存了图像的原始信息且有较好的信噪比.
There are a variety of methods of filtering noises. And it is easy to make image become mistiness by using usual averaging filter, the effect of averaging filter is not ideal. Although it has good effect by using median filter,it is easy to make image become mistiness ad lose detail information. In this article,we proposed a method of image filter based on gradient. This new method selects a 3 ×3 neighborhood,constructs 4 templates of gradient,processes noises,edge and inside information pixel of image according to the magnitude of 4 templates of gradient. The re.sult of experiment proves that the new method not only saves the original image information and has a good Signal-to-Noise Ratio, but also removes noises simultaneously.
出处
《南通纺织职业技术学院学报》
2005年第3期17-20,共4页
Journal of Nantong Textile Vocational Technology college
关键词
图像处理
滤波算法
噪声
梯度
image processing
filter algorithm
noise
gradient
