Скачать или смотреть Automated Machine Vision System for Recognition and Sorting of Small Gears

MECH 898 – Master’s Project
Student: Gustavo Baréa
Supervisor: Prof. B.W. Surgenor
Kingston, ON - April 2017

An automated machine-vision-based system for the recognition and sorting of small gears was designed, assembled and tested. This report documents the design of the system and presents experimental performance results. Nine experiments were conducted to analyze the effects of changing hardware and software parameters, including: conveyor speed, gear feed rate, classification, and identification score thresholds. The conclusion is that the system can achieve a maximum system accuracy of 97% at a feed rate of 40 parts/min, for a given set of parameter settings. With different parameter settings, an accuracy of 96% can be achieved at a feed rate of 65 parts/min, which is the maximum feed rate possible.