Developing An Object Detection Using Convolutional Neural Networks Demonstration Kit

Abstract

Artificial intelligence-based image processing technology plays a vital role in industrial manufacturing processes, where it is widely applied in quality inspection on production lines, defect detection in products, and automated process control. The integration of convolutional neural networks (CNNs) has significantly enhanced image analysis accuracy and efficiency. As a result, there is a growing demand in the labor market for personnel with knowledge and expertise in this field. However, the learning process for this technology is complex and often difficult to grasp, while access to learning materials and equipment remains limited and expensive, falling short of learners' needs. Therefore, this study aims to design and develop a demonstration kit for object detection using convolutional neural networks, which serves as an instructional tool for programming with Python and for using a camera to inspect scratches on copper plates through image processing for image classification. The objective is to improve the efficiency of image processing systems. The target group for this study consists of 27 second-year students in the Mechatronics and Robotics program. The research design follows a pre-experimental approach involving a pre-test and post-test comparison. The results show that the developed demonstration kit functions effectively and enhances students' learning outcomes. Moreover, learners expressed positive perceptions toward the learning activities involving the training kit.