计算机与现代化

• 图像处理 • 上一篇    下一篇

一种基于图像的大型车辆车身长度自动测量方法

  

  1. 安徽大学计算机科学与技术学院,安徽合肥230601
  • 收稿日期:2014-04-01 出版日期:2014-05-28 发布日期:2014-05-30
  • 作者简介:袁涵(1989-),男,安徽六安人,安徽大学计算机科学与技术学院硕士研究生,研究方向:计算机视觉; 汤进(1976-),男,安徽合肥人,教授,博士,研究方向:计算机视觉与模式识别。
  • 基金资助:
    国家自然科学基金资助项目(61202228)

An Image-based Method for Automatic Length Measuring of Large Vehicles

  1. School of Computer Science and Technology, Anhui University, Hefei 230601, China
  • Received:2014-04-01 Online:2014-05-28 Published:2014-05-30

摘要: 针对传统汽车尺寸测量技术在实际应用环境限制下的不足,提出一种基于图像的大型车辆车身长度测量方法。利用图像分割技术和基于图像边缘特征的匹配方法完成车身侧面全景图像拼接,并利用双目测距原理计算出车辆与相机间的距离,在此基础上,根据透视投影原理计算出最终的车身长度。实验结果表明,该方法可以很好地在受限的测量环境中自动、快速、准确地测量出大型车辆车身长度。

关键词: 汽车长度测量, 机器视觉, 图像拼接

Abstract: For the deficiency of the length measurement technology of large vehicles limited by actually application environment, this paper proposes an image-based metrology solution for the length measurement of large vehicles. We use the image segment method to extract different portions of the same vehicle from multiple images, and the matching method based on image edge feature to stitch the segmented parts into a complete profile of the vehicle. We assume that the side of vehicle is an approximate plane. We map the image coordinates onto a reference plane on which the distances between any two points can be measured in physical dimension. With the knowledge of the depths of the reference plane and vehicle side, we can map measurements on the reference plane onto the side of the vehicles using linear transformations. We tested the method with many vehicles of different types, and got pleasing results without environmental restriction.

Key words: vehicle length measurement, machine vision, image stitching

中图分类号: