ENHANCE STEREO VISUAL ODOMETRY PERFORMANCE BY REMOVING UNSTABLE FEATURES

  • Anh Duc Vu Institute of Information and Communication Technology, Le Quy Don Technical University
  • Xuan Phuc Nguyen Institute of System Integration, Le Quy Don Technical University
  • Van Xiem Hoang Faculty of Electronics and Telecommunications, Vietnam National University - University of Engineering and Technology
  • Quang Lam Bui Phu Xuan University
  • Quang Hieu Dang Institute of System Integration, Le Quy Don Technical University
  • Huu Hung Nguyen Institute of System Integration, Le Quy Don Technical University
Keywords: Stereo visual odometry, essential matrix, object detection, unstable feature selection

Abstract

Visual odometry includes two important stages: 1) feature extraction and 2) pose estimation. The performance of visual odometry is dependent on the quality of features including the number of features, the percentage of the correct matching, and the location of detected features. Usually, RANSAC method has been used in pose estimation to remove outlier and select a good set of features that provide higher accuracy. However, in the case the higher wrong matches, the RANSAC seems to be failing. This article proposes the removing unstable feature method by deep learning-based object detection. The proposed method evaluated on the KITTI dataset shows a higher accuracy 6 - 8% compared to the conventional method.

điểm /   đánh giá
Published
2024-07-24
Issue
Vol. 13 No. 01 (2024)
Section
Bài viết