Real-time Sorting of Stacked Objects Using Depth Information
Changlong Wang, Runing Wang, Jun Zheng · 2019
Aiming at real-time sorting of stacked objects, a series of algorithms are proposed, including plane segmentation and template matching with low computational load and high speed, using only depth information. Contour recognition and planar clustering are used to segment different objects. Then the U-Chord Height (UCH) template matching method is designed to realize accurate classification of objects. Thus robust and real-time sorting of stacked objects can be achieved.