Machine Vision Based Automatic Quality Inspection System for Connecting-rod

Abstract

This research deals with attempting to study, analyse, design, implement and testing of automatic machine vision-based system for quality control of engine Connecting-rod. The connecting-rod or conrod links the piston, which is a part of reciprocating engine, to the crankshaft. In combination with crankshaft, connecting-rods convert up and down motion of piston into rotating motion. In industries, large number of connecting rods with different diameters are being manufactured for various engines. Usually parameters of conrod such as large bore diameter are measured manually by using different gages. In this method, human mistakes contributes to for inaccurate diameter measurement. As automatic quality inspection system for connecting-rod is proposed to estimate its large bore diameter and small bore diameter of different sizes using machine vision. In order to measure the parameters of connecting-rod, both methods Circular Hough transform and Histogram Based Edge Detection have been implemented and compared. However satisfactory results have been achieved with Histogram Based Edge Detection. The proposed algorithm provides accuracy, efficiency and it is also cost-effective as compared to existing manual inspection system.