测绘通报 ›› 2022, Vol. 0 ›› Issue (4): 37-43.doi: 10.13474/j.cnki.11-2246.2022.0107

• 学术研究 • 上一篇    下一篇

复杂背景下航拍图像的电力线自动提取算法

陈竹安1,2, 邹梓龙1, 徐志芳1, 彭嘉琪1, 施陈敬3, 洪志强4   

  1. 1. 东华理工大学测绘工程学院, 江西 南昌 330013;
    2. 自然资源部环鄱阳湖区域矿山环境监测与治理重点试验室, 江西 南昌 330013;
    3. 广东国地资源与环境研究院, 广东 广州 510000;
    4. 南昌工学院, 江西 南昌 330108
  • 收稿日期:2021-03-04 修回日期:2021-12-28 出版日期:2022-04-25 发布日期:2022-04-26
  • 通讯作者: 邹梓龙。E-mail:zzl1102@qq.com
  • 作者简介:陈竹安(1978-),男,副教授,主要研究方向为摄影测量、图像处理及遥感数据处理。E-mail:cza53@qq.com
  • 基金资助:
    国家自然科学基金(51708098);江西省教育厅课题(GJJ180396);2021年国家(江西省)大学生创新创业训练计划(S202110405037)

Automatic power line extraction algorithm for aerial image under complex background

CHEN Zhu'an1,2, ZOU Zilong1, XU Zhifang1, PENG Jiaqi1, SHI Chenjing3, HONG Zhiqiang4   

  1. 1. Faculty of Geomatics, East China University of Technology, Nanchang 330013, China;
    2. Key Laboratory of Mine Environmental Monitoring and Improving around Poyang Lake, Ministry of Natural Resources, Nanchang 330013, China;
    3. Guangdong Guodi Institute of Resource and Evironment, Guangzhou 510000, China;
    4. Nanchang Institute of Science and Technology, Nanchang 330108, China
  • Received:2021-03-04 Revised:2021-12-28 Online:2022-04-25 Published:2022-04-26

摘要: 无人机对电力线巡检的关键问题是如何从复杂背景的航拍图像中准确地提取电力线。本文提出了一种基于二维变分模态分解 (2D-VMD) 提取电力线的新算法。首先对原始航拍图像进行预处理,加快数据处理速度;然后采用2D-VMD算法对预处理后的图像进行分解,通过改进后的点锐度算法,选取带有电力线特征的IMF分量图,并利用Roberts算子进行边缘检测;最后利用形态学改进的Hough变换,完成对电力线的提取。试验结果表明,本文方法比传统的Canny算子结合Hough变换方法、LSD方法、Roberts 算法结合形态学改进的Hough变换方法更具精确性、抗噪性、自动化。

关键词: 复杂背景, 二维变分模态分解, Roberts算法, 形态学, Hough变换

Abstract: The key problem of UAV power line inspection is how to accurately extract power lines from aerial images with complex background. This paper proposes a new algorithm for power line extraction based on two-dimensional variational mode decomposition (2D-VMD). Firstly, the original aerial image is preprocessed to speed up data processing. Secondly, 2D-VMD algorithm is utilized to decompose the preprocessed image. The IMF component graph with power line features is selected by the improved point sharpness algorithm and edge detection is performed by Roberts operator. Finally, the power lines are extracted by morphological modification of Hough transform. Experimental results show that the proposed method is more accurate, noise resistant and automatic than the traditional Canny edge detection combined with Hough transform method, line segment detector (LSD) method, Roberts edge detection and morphological improved Hough transform method.

Key words: complex background, two-dimensional variational mode decomposition (2D-VMD), Roberts algorithm, morphology, Hough transform

中图分类号: