real-time object recognition, pose acquisition, and tracking

Real-time model-based object recognition, pose estimation, and tracking in dense stereo.
Variable-Dimensional Local Shape Descriptors (VD-LSD, by Taai & Greenspan) are used for object recognition and pose estimation. Bounded Hough Transform (BHT, by Shang & Greenspan) is used for tracking.

VD-LSD project page: http://rcvlab.ece.queensu.ca/~qridb/lsdPage.html
(includes a database of 3D models, LIDAR and stereo 3D point clouds of cluttered scenes and 3D point clouds of cluttered scenes.)

Category:

Science & Technology

Tags:

• computer
• vision
• object
• recognition
• pose
• estimation
• tracking
• real-time
• stereo
• 3D
• range
• data
• image