Computational & Technology Resources
an online resource for computational,
engineering & technology publications
Civil-Comp Proceedings
ISSN 1759-3433
CCP: 103
Edited by: Y. Tsompanakis
Paper 20

A Computer Vision based Experimental Device for Plastic Bottle Identification and Sorting

E. Scavino1, M.A.M. Arebey1, H. Basri2, A. Hussain1, M.A. Hannan1 and R. Mohd Saleh1

1Department of Electrical, Electronic and System Engineering
2Department of Civil and Structural Engineering
Faculty of Engineering and Built Environment
National University of Malaysia, Bangi, Malaysia

Full Bibliographic Reference for this paper
E. Scavino, M.A.M. Arebey, H. Basri, A. Hussain, M.A. Hannan, R. Mohd Saleh, "A Computer Vision based Experimental Device for Plastic Bottle Identification and Sorting", in Y. Tsompanakis, (Editor), "Proceedings of the Third International Conference on Soft Computing Technology in Civil, Structural and Environmental Engineering", Civil-Comp Press, Stirlingshire, UK, Paper 20, 2013. doi:10.4203/ccp.103.20
Keywords: computer vision, pattern recognition, sorting, plastic bottles.

An experimental machine vision apparatus was used to identify and sort recyclable plastic bottles sliding on a tilted plane. A prototype singulating device was installed before the image acquisition area, in order to deliver one bottle at a time to the identification and sorting system. Colour images were taken with a commercially available webcam and the recognition was performed using the software developed, based on the shape and dimensions of object images. The identification was fulfilled by comparison of the geometrical data of the bottle image with the data stored in an existing database. New occurrences, corresponding to non-identified bottles, were stored in the database then manually verified in order to avoid duplicates in the database. Thus conceived, the database is intended to automatically increase in size and the system to become more complete and independent. The identified bottles were introduced into a tilted slide then deflected by opening a suitable lateral gate for sorting in accordance to the plastic type. The identification and sorting system was tested on a set of 150 different bottles of 5 different kinds of plastic of various sizes. Particular attention was focused on the efficiency of the image recognition software under various lighting conditions, as well as on the long term reliability of the mechanical and pneumatic components of the sorting system. Up to date, an efficiency of 97% was observed for the image and pattern recognition system, with shortcomings only due to very poor lighting conditions, while the hardware system showed no particular breakdowns after thousands of cycles.

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description