| 基于改进分水岭算法的球团矿图像分割 |
| |
| 作者姓名： | 秦长泽 刘 琼 |
| |
| 作者单位： | 武汉科技大学信息科学与工程学院 |
| |
| 基金项目： | 武汉科技大学大学生科技创新基金研究项目(13ZRC123) |
| |
| 摘 要： | 针对传统图像分割算法不能对球团矿图像中粘连球团进行准确有效分割的问题,提出一种改进的分水岭分割算法。该算法将标记和分水岭算法结合起来,利用标记带来的先验知识限定图像分割区域的数目,并对分水岭算法的梯度图像进行修改,从而能够对球团边缘准确分割,同时避免产生过分割现象。实验结果表明,改进分水岭算法对图像中粘连球团的边缘有良好的分割效果。
|
| 关 键 词： | 图像分割 分水岭算法 球团矿 标记 梯度图像 |
Image segmentation of pellets based on improved watershed algorithm |
| |
| Authors: | Qin Changze and Liu Qiong |
| |
| Institution: | Qin Changze;Liu Qiong;College of Information Science and Engineering,Wuhan University of Science and Technology; |
| |
| Abstract: | An improved watershed segmentation algorithm is proposed to solve the problem that cannot be solved by traditional image segmentation algorithms to segment the adhesive pellets in pellet images effectively and accurately. It combines markers with watershed algorithm, takes advantage of the prior knowledge from markers to limit the number of image segmentation regions, and modifies the gradient image in order to correctly segment the pellet edges and avoid the over-segmenting phenomenon in the meantime.Experimental results show that the improved watershed algorithm has better segmentation effect. |
| |
| Keywords: | image segmentation watershed algorithm pellet marker gradient image |
| 本文献已被 CNKI 等数据库收录！ |
| 点击此处可从《》浏览原始摘要信息 |
| 点击此处可从《》下载免费的PDF全文 |
