摘要
针对传统图像分割算法对不同类型噪声敏感性缺陷的问题,基于临近像素空间距离的模糊C均值聚类算法即SF-CM(fuzzy C means clustering algorithm based on the space distance of the nearest pixels)算法,采用核化的空间距离公式,将点到点之间的距离转化为点到空间的距离,很好的平衡了考察像素点临近像素点的灰度信息与位置信息间的关系,进一步克服了临近像素的位置差异对考察像素影响不同的缺点。通过在合成图像和自然图像上的大量实验并与几个传统算法进行对比,不仅表现出了很强的抗干扰能力,提高了聚类精度,并且很好的保留了原图像边缘等细节信息,体现出了较强的鲁棒性。
For the traditional image segmentation algorithms is sensitive to different type noises, the fuzzy C means clustering algorithm based on the space distance of nearest pixels （SFCM） uses a nuclear space distance formula which makes the distance between dot to dot translate into the dot to the space. So that it well balanced gray level and location information of the pixels around the query pixel, overcome the defect brought by the location differences of the pixels near the query pixel. Experiments performed on a large number of synthetic and real-world images show that it not only demonstrate a strong anti-interference ability to the noise, improve the clustering accuracy, hut also retaine the original image edge detail information, demonstrate stronger robustness.
出处
《计算机工程与设计》
CSCD
北大核心
2013年第7期2476-2482,共7页
Computer Engineering and Design
基金
国家自然科学基金项目(61170122)
江苏省自然科学基金项目(BK2011003)
