测绘通报 ›› 2018, Vol. 0 ›› Issue (4): 112-115,120.doi: 10.13474/j.cnki.11-2246.2018.0122

• 行业观察 • 上一篇    下一篇

航空影像直角房屋轮廓半自动提取

李大军1, 杜神斌1, 郭丙轩2, 薛万唱3, 游琦4   

  1. 1. 东华理工大学测绘工程学院, 江西 南昌 330013;
    2. 武汉大学测绘遥感信息工程国家重点实验室, 湖北 武汉 430079;
    3. 武汉讯图科技有限公司, 湖北 武汉 430079;
    4. 福州市勘测院, 福建 福州 350000
  • 收稿日期:2017-10-25 出版日期:2018-04-25 发布日期:2018-05-03
  • 通讯作者: 郭丙轩。E-mail:mobilemap@163.com E-mail:mobilemap@163.com
  • 作者简介:李大军(1965-),男,博士,教授,研究方向为GIS理论与应用研究。E-mail:djli@ecit.cn
  • 基金资助:

    国家重点研发计划(2016YFB0502200);国家自然科学基金(41127901);东华理工大学研究生创新基金(DHYC-2012007)

Semi-automatic Extraction of Rectangular House Contours from Aerial Images

LI Dajun1, DU Shenbin1, GUO Bingxuan2, XUE Wanchang3, YOU Qi4   

  1. 1. East China University of Technology, Nanchang 330013, China;
    2. State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan 430079, China;
    3. Wuhan Xuntu Technology Co. Ltd., Wuhan 430079, China;
    4. Fuzhou Investigation and Survering Institute, Fuzhou 350000, China
  • Received:2017-10-25 Online:2018-04-25 Published:2018-05-03

摘要:

以精确提取房屋轮廓为目的,提出了一种航空影像中直角房屋轮廓半自动提取算法。该方法首先通过人工使用鼠标依次在房屋边缘线的大致位置画直线,再利用影像分割分别得到所对应的窗口影像;进而利用LSD算法对窗口影像提取直线;然后以其为初值,在保证其整体位移和数值偏差最小的情况下利用最小二乘算法在原始影像图上进行线拟合和直角化;最后连接角点,完成房屋轮廓信息的提取。试验结果表明,该方法不仅有效地减轻了工作人员的劳动强度,而且定位精度高于人工切准。

关键词: 房屋轮廓, 半自动提取, 影像分割, LSD算法, 最小二乘算法, 直角化

Abstract:

In order to extract the contours of the house for the purpose accurately,this paper proposes a semi-automatic extraction algorithm for rectangular house contours in aerial images.The first step of this algorithm is using mouse manually to draw straight lines in the approximate position of house contours in turns,and next step is using image segmentation to achieve the corresponding window images.Then using LSD algorithm to extract straight lines from window images.After that,considering it as the initial value,the least squares algorithm is applied to carry on curve fitting and right angle processing on the original image in case of ensuring its overall displacement and numerical deviation are the minimum.The last step is connecting angular points,and the extraction of the house contours information is done.The experimental results show that this method can reduce labor intensity efficiently and the position accuracy is higher than artificial work.

Key words: house contours, semi-automatic extraction, image segmentation, LSD algorithm, least squares algorithm, right angle processing

中图分类号: