摘要
图像分割是图像处理中的一项关键技术,在图像工程中占据着重要的位置。随着信息技术的快速发展,它在地质、环保、气象、医疗等领域有着广泛的应用。常见的分割算法包括阈值分割方法、边缘检测方法、区域提取方法和结合特定理论工具分割方法。利用了分水岭算法的原理及特点,针对算法所产生的图像过分割问题,采用形态学处理函数,以典型的pears图像为例,用MATLAB 7.0图像处理软件,进行了仿真研究。结果表明,利用分水岭算法和形态学处理函数,使得图像中相连的多个对象被分割成多个单对象,实现了图像的有效分割,并尽可能地减少或消除了过分割现象。
Image segmentation plays a very important role in image processing, and occupies an important position in image project. With the rapid development of information technology., it is widely used in geological, environmental, meteorological, medical and other fields. Common segmentation includes threshold segmentation, edge detection methods, regional extraction methods and theories of specfic tools segmentation. The paper analyses the principle and characteristic of watershed algorithm. For images gererated by over - segmentation, taking pears image as example, and using image processing software with MATLAB7. 0 to carry out simulation study. The results show that the conjoint objects in this image are divided into many single objects by using watershed algorithm and the morophology processing functhion, the image is effectively segmented, and the over - segmentation is reduced or eliminated as much as possible.
出处
《计算机仿真》
CSCD
北大核心
2009年第5期255-258,共4页
Computer Simulation
基金
黑龙江省教育厅科学技术研究项目资助(1151355)
关键词
图像分割
数学形态学
分水岭
计算机仿真
Image segmentation
Mathematical morphology
Watershed
Computer simulation
