|
|||||
|
|
| 基于过渡区的遥感图像分割 | |
| 引用本文： | 肖晓玲,张翔,吴谨,田金文.基于过渡区的遥感图像分割[J].军民两用技术与产品,2003(10):44-46. |
| 作者姓名： | 肖晓玲 张翔 吴谨 田金文 |
| 作者单位： | 1. 长江大学江汉石油学院,湖北,荆州,434023 2. 长江大学江汉石油学院,湖北,荆州,434023;华中科技大学模式识别与人工智能研究所,图像信息处理与智能控制教育部重点实验室,武汉,430074 3. 华中科技大学模式识别与人工智能研究所,图像信息处理与智能控制教育部重点实验室,武汉,430074 |
| 摘 要： | 许多遥感图像的目标和背景之间界定模糊,表现为过渡平缓,往往导致边界定位和目标的提取出现误差。首先利用有效平均梯度和特殊剪切来获得目标与背景之间的过渡区,由于目标的边界必在过渡区内,所以提取误差会被限制在一定的范围内。最后利用最小距离法将过渡区内的像素判决后划分到目标或背景中去,进一步保证目标的质量,解决了此类问题中目标提取的困难。实验结果表明了这种方法的有效性。 |
| 关 键 词： | 分割 过渡区 最小距离法 |
| 文章编号： | 1009-8119(2003)10-0044-03 |
Transition-Region-based Remote Sensing Image Segmentation |
|
| Xiao Xiaoling Zhang Xiang, Wu Jin Tian Jinwen.Transition-Region-based Remote Sensing Image Segmentation[J].Universal Technologies & Products,2003(10):44-46. | |
| Authors: | Xiao Xiaoling Zhang Xiang Wu Jin Tian Jinwen |
| Affiliation: | Xiao Xiaoling1 Zhang Xiang1,2 Wu Jin2 Tian Jinwen2 |
| Abstract: | In a lot of remote sensing image, there are fuzzy boundary, smooth transition, causing the difficulty that the border is oriented and generating very great errors for the extract of object This paper solves the difficulty of object extraction to a certain extent Because the object that is extracted must be in the transition region, it can limit errors of extraction within certain range The pixels in transition region are divided into object and background respectively by Minimum Distance Rule in order to guarantee the quality of the object further The experimental result has shown the validity of this kind of method |
| Keywords: | segmentation transition region Minimum Distance Rule |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ | |
