Title :
Rebar Counting on Production Line Based on Machine Vision
Author :
Zuoxian Nie;Mao-Hsiung Hung;Jing Huang
Author_Institution :
Coll. of Inf. Sci. &
Abstract :
In this paper, we present a novel algorithm that automatically counts rebar on production line based on video analysis. Our algorithm allows that conveyor belt moves at rather higher speed and rebars roll forward or backward on conveyor belt. There are mainly two parts in our proposed algorithm. Firstly, a method is developed to divide the video frame into three partitions, and then a set of heuristic rules are deduced based on these partitions. Secondly, for cases excluded by above rules, a sequence matching algorithm is developed to identify repeated sequence of rebar between adjacent frames, and then number of newly arrived rebar in every frame can be computed. The experimental results show that our method can support high-speed and precise rebar counting.
Keywords :
"Belts","Image segmentation","Partitioning algorithms","Signal processing algorithms","Machine vision","Algorithm design and analysis"
Conference_Titel :
Robot, Vision and Signal Processing (RVSP), 2015 Third International Conference on
Electronic_ISBN :
2376-9807
DOI :
10.1109/RVSP.2015.18