A Capstone Design Project Machine Vision System In Inspection Process

Abstract

The purpose of this project is to design an automated inspection system that will identify, accept, or reject parts on a production line. The project simulates a major goal of industry-to monitor deviations in parts geometry. The project begins by integrating five components that work together as a system: a conveyor, a machine vision sensing device or camera, a robot, a programmable logic controller (PLC), and a pneumatic actuator. The conveyor belt moves parts to the inspection area, where the machine vision camera detects them. The camera makes an image of each part and displays that image within the software windows environment. The edge count and measurement distance sensors in the software identify each part as acceptable (PASS) or not acceptable (FAIL). The system, as it is designed, uses a PASS signal to recognize acceptable parts. That signal then is sent to and used by the robot to gather and assemble the acceptable parts in a central location. If the part is unacceptable, it is moved within reach of the actuator, which pushes it off the conveyor belt into a scrap bin. Five components, each performing a different function, are tied together into a system dedicated to ensuring that customers receive a quality product.