论文部分内容阅读
数字图象的噪声平滑是数字图象处理技术中非常重要的一个领域.本文从图象的统计特点出发,寻找出了一种局部灰度拟合的平滑方法.它不同于以往的一些噪声平滑方法,如中值滤波法,邻域平均法,梯度倒数法等等.本方法由于比较具体地分析了图象的特点,因此具有比较广泛的适应性和综合的图象噪声平滑及保持边缘能力.由于在本算法中存在一个可调节的参数k,所以只要略作修改,就可以实现噪声的自适应平滑.
The noise smoothing of digital image is a very important field of digital image processing technology.This paper starts from the statistical characteristics of image and finds out a smoothing method of local gray fitting, which is different from some previous noise smoothing Methods such as median filtering, neighborhood averaging, gradient reciprocal, etc. This method, due to a more specific analysis of image features, has a relatively wide range of adaptive and integrated image noise smoothing and edge-preserving capabilities Due to the existence of an adjustable parameter k in this algorithm, adaptive noise smoothing can be achieved with minor modifications.