• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>IEEE Joint International Information Technology and Artificial Intelligence Conference >An improved image segmentation algorithm based on the watershed transform
【24h】

An improved image segmentation algorithm based on the watershed transform

机译：一种改进的基于流域变换的图像分割算法

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

The watershed transform has a good response to the weak edge, but it is unable to obtain meaningful segmentation results directly by using the watershed transform to the phenomenon of excessive segmentation in image segmentation. A new segmentation algorithm base on improved watershed algorithm is proposed in this paper. By the different marks on the foreground and background objects, we can obtain better effect of segmentation with the application of watershed algorithm. Through the study of the characteristics of gradient image and morphological open close function reconstruction operations in image processing, a new tag extraction method will be designed by the new algorithm, which is by using the method of opening and closing by reconstruction operators to obtain marks and by controlling the closed reconstruction of structural elements in order to interested in the gradient image area the size of the tag. Then, extracted tag using morphological minimum calibration technology forces as an original local minimum value of the gradient image, and shields the original all local minimum value in the gradient image. In the end, watershed which is modified by the gradient of the image will have a image segmentation, in such aspects as outline eliminate over-segmentation and regional positioning have a very good segmentation effect.
机译：流域的变换对弱边缘具有良好的反应，但是通过使用流域转换到图像分割中过度分割的现象，无法直接获得有意义的分割结果。本文提出了一种改进的流域算法的新分割算法。通过前景和背景对象的不同标记，我们可以通过应用流域算法获得更好的分割效果。通过研究梯度图像和形态开放功能重建操作的图像处理中的特征，通过新的算法设计了一种新的标签提取方法，即通过使用重建操作者打开和关闭方法来获得标记的方法通过控制结构元素的闭合重建，以便对梯度图像区域感兴趣的标签的大小。然后，用形态最小校准技术力作为梯度图像的原始局部最小值提取标签，并屏蔽梯度图像中的原始所有局部最小值。最后，通过图像的梯度修改的流域将具有图像分割，在这样的方面，作为轮廓消除过分分割和区域定位具有非常好的分割效果。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号