HRBB4: Indoor Image Data for Visual Odometry/SLAM
Yan Lu, Dezhen Song
Introduction
This dataset is provided for testing visual odometry/SLAM systems in typical office corridor scenarios. The images are recorded using a Nikon 5100 camera mounted on a PackBot on the 4th floor of H.R. Bright Building, Texas A&M University. The trajectory length is around 70 m. The ground truth of camera positions is also provided.
Download (zip)
The image sequence consists of 12,000 frames of 640 x 360 pixels. Ground truth of camera positions is also included.
Download as a zip (4GB) file.
Related publications
Please cite our work if you use this dataset in your research.
  • Yan Lu, Dezhen Song, and Jingang Yi. "High Level Landmark-Based Visual Navigation Using Unsupervised Geometric Constraints in Local Bundle Adjustment", IEEE International Conference on Robotics and Automation (ICRA), 2014
Credits
Hancheng Ge, Changyoung Kim, and Wen Li.
Links
Multilayer Feature Graph, Netbot Lab