测绘通报 ›› 2021, Vol. 0 ›› Issue (8): 28-32,47.doi: 10.13474/j.cnki.11-2246.2021.0235

• 学术研究 • 上一篇    下一篇

基于GA-SIFT算法的无人机航拍图像实时拼接

王艳, 祁萌   

  1. 成都工业学院, 四川 成都 610031
  • 收稿日期:2020-10-22 出版日期:2021-08-25 发布日期:2021-08-30
  • 通讯作者: 祁萌。E-mail:983629548@qq.com
  • 作者简介:王艳(1986-),女,硕士,讲师,主要研究方向为数字图像处理、计算机视觉。E-mail:1423950761@qq.com
  • 基金资助:
    国家自然科学基金(61705022);四川科技支撑项目(2018JY0507)

Real-time stitching of unmanned vehicle images based on GA-SIFT algorithm

WANG Yan, QI Meng   

  1. Chengdu Technological University, Chengdu 610031, China
  • Received:2020-10-22 Online:2021-08-25 Published:2021-08-30

摘要: 为实现无人机航拍图像的实时拼接,本文深入研究了无人机航拍图像拼接中的关键技术,提出了一种基于遗传算法优化的图像拼接算法。首先利用SIFT算法提取图像的特征点,在特征点粗匹配过程中,采用欧氏距离作为相似度测量,利用遗传算法的并行性优化特征点匹配性能;然后使用RANSAC算法去除误匹配点对并获得转换矩阵,从而完成图像拼接。试验结果表明,采用遗传算法进行特征匹配,可大大降低匹配时间,匹配时间与特征点数量成正比;同时提高了匹配精度,进而提高了图像拼接的实时性和稳健性。

关键词: 无人机航拍图像, 实时拼接, SIFT, 欧氏距离, 遗传算法

Abstract: In order to realize the real-time splicing of unmanned aerial vehicle images, the key technologies in aerial image stitching of an unmanned aerial vehicle are deeply researched in this paper. To solve the problem of time-consuming in the process of feature point matching, an image mosaic algorithm based on genetic algorithm optimization is proposed. Firstly, the SIFT algorithm is used to extract the feature points of the image. In the process of feature point matching, the Euclidean distance is used as the similarity measurement, and the genetic algorithm is used to optimize algorithm. Then the RANSAC algorithm is used to remove the mismatched points to complete image stitching. Experiments show that GA-SITF algorithm greatly reduces the matching time, which is proportional to the number of feature points, and at the same time improves the matching accuracy, thereby improving the real-time and robustness of image stitching.

Key words: unmanned vehicle images, real-time stitching, SIFT, Euclidean distance, GA

中图分类号: