摘要
图像分割是一种特殊的图像技术,也是一种关键的图像分析技术。文中论述了图像分割在图像处理中的重要作用及好的分割方法的评价标准,探讨了分水岭算法的原理及应用。针对分水岭算法容易产生过分割的问题,首先运用形态学中的基本运算对原始图像进行预处理以增加图像对比度,然后利用分水岭算法分割检测目标,最后利用Matlab工具箱函数仿真。实验结果表明,与改进前的分水岭算法相比,该方法简单有效,可以获得更加符合人的视觉特性的分割效果。
Image segmentation is a special image technology, and also is a kind of key technique of image analysis. It discusses the impor tant role of the image segmentation in image processing and good segmentation evaluation criterion. And then it introduces the principles and application of watershed algorithm. For watershed algorithm is vulnerable to the problems of over split, the application of morphologi- cal opening and closing operation is performed on the original image processing to increase image contrast first. Then use the watershed algorithm to detect the target segmentation. Finally ,Matlab toolbox function simulation is used to test the improved algorithm. The experi ment results show that compared with watershed algorithm before improvement,this method is simple and effective and it can obtain more consistent with human vision character segmentation.
出处
《计算机技术与发展》
2012年第12期90-92,96,共4页
Computer Technology and Development
基金
陕西省教育科研计划项目(11Jk0485)
关键词
图像分割
分水岭算法
过分割
数学形态学
MATLAB
image segmentation
watershed algorithm
over-segmentation
mathematical morphology
Matiab
