DESIGN AND FABRICATION OF AN EMBEDDED SYSTEM WITH MACHINE VISION FOR PROCESS AUTOMATION AND QUALITY INSPECTION

Abstract

In this new age of the fourth industrial revolution manufacturing market has become more competitive with state-of-the-art technologies. The new technologies are making the manufacturing process more robust and efficient in nature. Process automation is one of the key areas where industries can give more emphasis to face the growing challenges of the market. Here systems embedded with machine vision can play a pivotal role in quality assurance and efficient process automation. Also reducing the cost of production significantly in all sorts of manufacturing processes. For the computational purpose, an open-source Linux operated portable credit-cardsized computer is used along with its camera module. The camera is used to capture the images of the objects and pass the information to the computer. The computer analyzes the captured images through a machine learning algorithm and detects unwanted objects to removes them from the production line as part of quality inspection and sorting. The object is defined by its color. Objects having color defined as defective are removed from the conveyor. This study will provide guidance for implementing automated color error detection based system to sort defective items in many fields where continuous evaluation and improvement of the quality are demanded.