To accelerate research on vision-based obstacle detection and identification for AGV applications, We are pleased to announce our preliminary plan to collect a large-scale RGB-D video data set with our high resolution depth camera that is based on solid-state flash LIDAR technology. Our depth cameras offer much higher resolution than typical LIDAR, and much longer working distance than typical commercially available depth cameras based on active stereo, structured light or TOF. This data set will be publicly available for free for all academic research purposes.
In order to better serve the needs of researchers in the field, we sincerely ask your inputs on all aspects of the creation of this data set. Any comments, suggestions, and recommendations are greatly appreciated.
Data set collection prelimianary plan