首页> 外文学位 >Detection du reseau routier a partir des images satellitaires a la suite d'une catastrophe majeure.
【24h】

Detection du reseau routier a partir des images satellitaires a la suite d'une catastrophe majeure.

机译:重大灾难发生后,根据卫星图像检测道路网络。

获取原文
获取原文并翻译 | 示例

摘要

With the resolution increase of remote sensing image, the road extraction is done with most precision followed by a better identification of the various transportation ways. However, this precision has a price. It generates noise due to the sensor and the urban context which make difficult the extraction. In a context of natural disaster and major risks, the time of intervention of the first-aid workers or efficient organization of assistance on the ground and the reduction of false detections by the algorithms of extraction are paramount and most important. Generally, the structures (various highway types) which one wants to analyze in an image are of different sizes. Therefore, the existence of a single resolution adapted to all of these objects is to be put aside. The multiresolution analysis which describes the image under various space scales with a strong capacity of target detection will be employed initially to reduce this sensitivity to the noise. In a second time, the output image of the extraction of the routes obtained with the modified algorithm of extraction of Christophe et Inglada (2007) by Spiric (2011) as viewing space. The initialization and modeling are based on the intrinsic characteristics of the road.
机译:随着遥感影像分辨率的提高,道路提取以最高精度进行,然后更好地识别各种运输方式。但是,这种精度是有代价的。由于传感器和城市环境的原因,它会产生噪声,从而使提取变得困难。在发生自然灾害和重大风险的情况下,急救人员的干预时间或有效的现场援助组织以及提取算法减少错误检测的时间至关重要。通常,要在图像中分析的结构(各种公路类型)具有不同的大小。因此,将撇开适用于所有这些对象的单一分辨率的存在。最初将采用多分辨率分析来描述在各种空间尺度下具有强大目标检测能力的图像,以降低对噪声的敏感性。在第二次中,使用Spiric(2011)的Christophe et Inglada(2007)的改进提取算法获得的路线提取的输出图像作为查看空间。初始化和建模基于道路的固有特性。

著录项

  • 作者

    Coulibaly, Idrissa.;

  • 作者单位

    Ecole de Technologie Superieure (Canada).;

  • 授予单位 Ecole de Technologie Superieure (Canada).;
  • 学科 Artificial Intelligence.;Geodesy.;Remote Sensing.
  • 学位 M.Eng.
  • 年度 2013
  • 页码 254 p.
  • 总页数 254
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类
  • 关键词

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