测绘通报 ›› 2019, Vol. 0 ›› Issue (12): 56-59,64.doi: 10.13474/j.cnki.11-2246.2019.0386

• 学术研究 • 上一篇    下一篇

无人机影像特征联合提取与双向匹配研究

吉文来, 赵湘玉, 李明峰   

  1. 南京工业大学测绘科学与技术学院, 江苏 南京 211816
  • 收稿日期:2019-06-24 修回日期:2019-08-25 发布日期:2020-01-03
  • 作者简介:吉文来(1967-),男,高级工程师,主要研究方向为精密工程测量、基坑变形监测技术与信息化。E-mail:jiwenlai1967@163.com
  • 基金资助:
    南京市科技计划(201716027);江苏省测绘地理信息科研项目(JSCHKY201809)

Research on united extraction and cross matching of UAV image features

JI Wenlai, ZHAO Xiangyu, LI Mingfeng   

  1. School of Geomatics Science and Technology, Nanjing Tech University, Nanjing 211816, China
  • Received:2019-06-24 Revised:2019-08-25 Published:2020-01-03

摘要: 针对无人机影像数据量大、存在旋转和尺度变化的问题,提出了基于特征联合提取与双向匹配的影像配准方法。分析了SURF算法和BRISK算法的特点,研究了融合SURF特征点检测和BRISK特征点描述的影像特征联合提取和PROSAC双向匹配的算法和步骤,并结合试验分析验证了该方法的有效性。

关键词: 无人机影像, 特征点提取, 特征点匹配, 双向匹配, 影像配准

Abstract: Aiming at the problem of UAV images with large number and changes of rotation and scaling, a method of image registration based on united extraction of features and cross matching is proposed. The characteristics of SURF algorithm and BRISK algorithm are analyzed. The algorithms and steps of united extraction of image features with SURF feature point detection and BRISK feature point description and PROSAC cross matching are studied. The validity of this method is verified by experimental analysis.

Key words: UAV image, feature point extraction, feature point matching, cross matching, image registration

中图分类号: