计算机工程与应用 ›› 2012, Vol. 48 ›› Issue (6): 193-195.

• 图形、图像、模式识别 • 上一篇    下一篇

对象化过程中的快速标号算法研究

李龙翔1,胡晓东2,沈占锋2,明冬萍1,2,宋卓沁1,2   

  1. 1.中国地质大学(北京) 信息工程学院,北京 100083
    2.遥感科学国家重点实验室(中国科学院 遥感应用研究所,北京师范大学),北京 100101
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2012-02-21 发布日期:2012-02-21

Fast labeling method used in objectification

LI Longxiang1, HU Xiaodong2, SHEN Zhanfeng2, MING Dongping1,2, SONG Zhuoqin1,2   

  1. 1.Institute of Information and Engineering, China University of Geosciences(Beijing), Beijing 100083, China
    2.State Key Laboratory of Remote Sensing Science(Institute of Remote Sensing Applications of Chinese Academy of Sciences, Beijing Normal University), Beijing 100101, China
  • Received:1900-01-01 Revised:1900-01-01 Online:2012-02-21 Published:2012-02-21

摘要: 面向对象的遥感信息提取技术由于其更符合人脑思维方式、更容易与地学知识结合,目前已经成为遥感信息自动提取的主流方向。对象化处理是面向对象遥感信息处理的重要步骤,实现了从分割影像到对象矢量影像的转变。提出了一种快速标号方法,可以快速为分割图像中的对象赋给唯一值标签并同时统计对象的其他特征,可以显著地提高去除小班块、矢量化预处理等操作的运行速度,在对象化过程中有广阔的应用前景。

Abstract: Object-oriented information extraction method has become the mainstream of automatic extraction of information in remote sensing, since it is in accord with artificial intelligence and has more approaches to geographic theories. Objectification is a key step in object-oriented automatic extraction of information, in which a segmentation image is transmitted into a vector image. The proposed method is able to assign a unique value to an object represented by a grey value in the segmentation image, at the same time, calculates some other features of this object. Using this method, many processes such as removing small region, pretreatment of vectorization can be achieved much faster.