Inspection of Objects using Computer Vision

Abstract

This study is motivated mainly by the need for more efficient and advanced techniques in an inspection of an object because accurate and timely information is needed for any industry to improve their quality and increase the production of goods. The objective of this study is to provide an inexpensive and comprehensive review of defect inspection techniques. Nowadays, new computer vision technologies and image processing technologies have been very important in the improvement and automation of manual processes in technical areas such as industries. In this study, a system that can replace the currently deployed manual inspection procedure by using image processing techniques is studied. Digital image processing is used in many fields, mainly for detection of faulty parts or missing parts. Inspection using digital image processing is the key factor behind all the industrial applications. In this study we have used enhanced ‘Speeded Up Robust Features’ or "SURF" algorithm, our model included counting the features in test image and original image, then matching percentage calculated using a metric of counting the size of inlier matching features towards outlier features. The study mainly deals with analysis to find faulty objects rather than a statistical approach. This study gives us idea that manual inspection technique can be replaced with the proposed idea and hence productivity and accuracy of inspection can be increased.