Author(s)

Wei Chen, Yingze Hong, Hualiang Lin, Jiahao He, Kan Luo and Ying Ma

Corresponding Author

Wei Chen

ABSTRACT

The key to solve the online counting of aluminum-profiles is how to achieve the tracking of multi-target which run at high speed. In this paper, machine vision technology is used to solve the online tracking problem. Firstly, the connected domain labeling algorithm is used to extract the surface adhesion objects one by one. By using this algorithm, the tracking of aluminum-profiles can be reduced to the tracking of center of mass of the suborbicular area. Secondly, the motive distance of the center of mass is estimated by the correlation analysis of binary projection curve. Finally, the coordinates and displacement of the center of mass are used as the characteristic parameters to achieve accurate location and fast tracking of multi-target. Through the on-line counting test on aluminum-profiles with specifications of 20mm and above, the results show that this algorithm can ensure the count accuracy and real-time.

KEYWORDS

Multi-target tracking, Motion estimation, Connection domain, Aluminum-profile