Smart board level machine vision system for online inspection: A Review

Abstract: A simple and low cost machine vision system for fault detection and identification is envisaged. The proposed system works on a Beagle Board single board computer. The system can be used to detect cracks or texture differences in classifying objects, especially in food processing industries. It essentially uses a camera for the image capture and process the image in the Beagleboneblack. Currently an online machine vision system on a conveyor line is made up of several cameras that are networked to a single computer. The camera provides the images at high rates and the software on the computer does the processing and makes the decisions to reject the product or not. The goal of this project is to design a smart board level machine vision system using a beagleboneblack with its own image sensors/cameras. The system will have its own algorithms for common machine vision tasks and will be networked to a main system to get output results.

Keywords: Machine vision system, Beagleboneblack, Image processing, Fault detection.