基于多阈值算法和分水岭算法的图像分割  

Image Segmentation by Using Multi Threshold Algorithm and Watershed Algorithm

作　　者：樊秋月[1] FAN Qiu-yue(Guangdong Institute of Science and Technology, ZhuHai 519090, Guangdong, China)

机构地区：[1]广东省科技干部学院机械与电子工程学院,广东珠海519090

出　　处：《内蒙古师范大学学报（自然科学汉文版）》2016年第5期653-656,共4页Journal of Inner Mongolia Normal University(Natural Science Edition)

基　　金：广东省高等职业教育品牌专业建设项目(2050305)

摘　　要：图像分割是图像处理中的关键技术,有利于目标的定位和识别,为了提高图像分割的精度,针对单一分割方法存在的不足,提出一种基于多阈值算法和改进分水岭算法的图像分割方法.首先采用多阈值算法分割图像,得到图像的初步分割轮廓,然后采用分水岭算法实现再一次图像分割,最后选择多种类型的图像对分割方法的精度进行验证.实验结果表明,该方法可以将用户感兴趣的目标准确地分割出来,图像分割精度和速度优于单一的多阈值算法或改进分水岭算法,更有利于图像的后期处理.Image segmentation is the key technology in image processing,and it is conducive to target location and identification.In order to improve the accuracy of image segmentation,this paper proposes a new image segmentation method by using multi threshold algorithm and improved watershed algorithm to solve shortcomings of single methods.Firstly,multi threshold algorithm is used to segment the image,and the initial segment of image is obtained,and secondly,marker watershed algorithm is used to realize the image segmentation again,lastly,the segmentation accuracy is verified by selecting various types images.The results show that the proposed method can accurately segment object of interest,image segmentation accuracy and speed is better than multi threshold algorithm and improved marker watershed algorithm,it is more conducive to the image of the late processing.

关 键 词：图像分割 阈值分割算法 分水岭算法 边缘检测 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]