Product defect categorization using machine vision through machine learning. Issue 1 (December 2020)

• Record Type: Journal Article
• Authors: Karthikeyan, G; Naveen, Monagari; Jegadeeshwaran, R; Sakthivel, G; Saravanakumar, D; Lakshmipathi, J
• Is Part Of: Journal of physics. Volume 1716:Issue 1(2021)
• Publication Date: 2020-12
• Subjects: Machine vision -- machine learning -- kNN -- SVM -- GUI.; Physics -- Congresses; 530.5
• DOI: 10.1088/1742-6596/1716/1/012029
• Languages: English
• ISSNs: 1742-6588

Abstract

Abstract: O-rings are among the seals most often used in the industry. O-rings accuracy measurement and inspection play a significant role in seal quality control. Human tests can be unpredictable and can take time. The goal of this paper is to use detection algorithms based on machine vision technology to monitor the O-rings norm, which also has the correct measurement rule and the classification rule. During this, we find an entirely different variety of good defects, Material shortage, Bounce, Spiral, and Breakage. Extract values for the elements by using MATLAB. Feature selection optimization attribute choice with MATLAB and classification exploitation of KNN, SVM, call Trees and alternative classifier variety with MATLAB is performed to check for the utmost prediction precision. To evaluate the recorded images of O-rings and conduct the measurement and inspection processes a computer vision program is applied. Then the GUI system is built to interface the user with the credibility of accessing the trained model. The proposed GUI is tested via a sequence of O-rings being checked.