Development of Computer Vision Algorithms for Measurement and Inspection of Spur Gears (Dept.M)

Abstract

Gears are one of the most common mechanisms for transmitting power and motion. Error in gears causes two main problems, increased acoustic noise in operation and increased wear, both of which are sufficiently troublesome to cause concern. Therefore, precision measurement of gears plays a vital role in gear measurement and inspection. The current methods of gear measurement are either time consuming or expensive. In addition, no single measurement method is available and capable of accurately measuring all gear parameters while significantly reducing the measurement time. The aim of this paper is to utilize the computer vision technology to develop a non-contact and rapid measurement system capable of measuring and inspecting most of spur gear parameters with an appropriate accuracy. The vision system has been established in the metrology lab and it is used to capture images for gears to be measured or inspected. A software (named GearVision) has been especially developed in-house using Microsoft Visual C++ to analyze the captured images and to perform the measurement and inspection processes. The introduced vision system has been calibrated for both metric and pixel units. After calibration, the system was verified by measuring two sample gears and comparing the calculated parameters with the actual values of gear parameters. The maximum variations between the calculated parameters and the actual values were +0.101 mm for a spur gear with 156 mm outside diameter. This variation can be decreased by measuring small gears.