Machine Design Automation Model for Metal Production Defect Recognition with Deep Graph Convolutional Neural Network

Error detection has a vital function in the production stages. Computer-aided error applications bring significant technological innovation to process control quality of products. As result, product reached an essential point because computer-aided image processing technologies. Artificial intelligence methods, such as Convolutional Neural Network (CNN), can detect and classify errors. However, detecting acceptable small defects on base parts cannot be done with high rate accuracy. At this point, it is possible minor errors help graph convolutional network, which emerged new method. In study, defect elements surfaces metal nut are determined through ensured. First, surface images captured. For this, python-based Raspberry pi card modified camera system were installed. Adapters three different zoom options used system, depending part The obtained second step sent other computer, for via local server. third stage, transformations by graphically separating white black color tones histogram maps these drawn. Value ranges classified according value from defective parts. nine models analyzed. According analysis results, neural network method gives 2.9554% better results than conventional methods.