弱纹理环境下基于线条的图像位姿恢复

Line-based pose recovery of image in texture-less environment

近年来,随着图像传感器的快速发展,基于视觉的定位技术越来越受到研究学者的关注。典型的视觉定位技术是即时定位与建图(Simultaneous Localization and Mapping,SLAM)。传统的SLAM一般是通过对图像中的特征点进行提取并计算描述子,但是在室内场景下,环境多为弱纹理情形,如走廊,墙壁等。文中提出了一种轻量级的、快速的基于线条特征进行图像间的位姿恢复的方法。实验结果显示,和基于特征点方法相比较,文中提出的方法在弱纹理环境下能够恢复出欧拉角的误差不大于5度,计算时间大约减少了2~3倍。

In recent years,with the rapid development of image sensors,vision-based positioning technology has attracted more and more attention of researchers.The typical visual positioning technology is Simultaneous Localization and Mapping(SLAM).Traditional methods of SLAM usually extract feature points from images and calculates descriptors,however in indoor scenes,the environment is mostly texture-less,such as corridors,walls,etc.Therefore,this paper proposes a lightweight and fast method for poserecovery between images based on line features.The experimental results show that the proposed method can recover Euler angle in texture-less environment with an error of less than 5 degrees and the calculation time is reduced by 2~3 times,compared with feature points-based methods.