摘要
为解决传统滤波去噪算法使图像的边缘变得模糊的问题,提出一种能智能判断边缘细节和带噪声的平滑区域的综合改进算法。该算法的主要思想是使用方差来计算某一像素点相对于局部区域的导向,判断该点的偏离程度,区分该点是处于该区域的平滑区还是边缘纹理区,自适应地对图像中的像素点进行判断,将纹理边缘和平滑区区分开来。实验结果表明,该算法有效抑制了噪声,保持了图像边缘和纹理信息的清晰度,具有很好的滤波效果。
To solve the problem that the traditional filtering de-noising algorithm makes the edge of the image blur, an improved algorithm which intelligently judged edge details and noise smoothing region was proposed. The main idea of the algorithm was to calculate the variance of a pixel point relative to the local orientation, so as to decide the deviation degree of the point. Then whether this point was on the verge of smooth areas of the region or texture areas of it was determined, and the pixels in the image were judged adaptively, and the smooth area and texture edge area were judged. The experimental results show that the intelligent algorithm not only effectively suppresses the noise, but also keeps the sharpness of the image edge and texture infor- mation. Thus it has wonderful filtering effects on images.
出处
《计算机工程与设计》
北大核心
2015年第8期2180-2184,共5页
Computer Engineering and Design
基金
国家科技重大专项基金项目(2011ZX05021-006)
关键词
图像处理
滤波算法
噪声消除
图像边缘
纹理信息
image processing
filtering algorithm
noise elimination
image edge
texture information
