AI Based Product Defect Detection Using Deep Learning a CNN, Yolo V8
Ch. Kodanda Ramu1, K. Mounika2, D. Praveen3, Ch. Vinod4, K.Bhavani Sankar5
1Associate Professor, Department of Computer Science & Engineering, Miracle Educational Society Group of Institutions, Bhogapuram, Vizianagaram, Andhra Pradesh, India - 535216
2,3,4,5B.Tech Student , Department of Computer Science & Engineering, Miracle Educational Society Group of Institutions, Bhogapuram, Vizianagaram, Andhra Pradesh, India - 535216
Email: kvr.chintu1978@gmail.com
Abstract - In current manufacturing sectors customer satisfaction is dependent on high product standards and minimizing financial impact from the release of defective products. The traditional methods of manually inspecting products are cumbersome, require a lot of resource (time and money) and prone to human error particularly when dealing with large scale production. This project proposes a system which will automatically detect product defects utilising artificial intelligence, computer vision, and machine learning techniques. The proposed solution will leverage Deep Learning methodologies, specifically Convolutional Neural Networks (CNN's), to analyse the images of finished goods taken throughout the manufacturing process. The model will be trained on a data set containing images of both defect and defect free products. Once trained the model will be able to identify surface defects including but not limited to cracks, scratched, dented or incorrectly assembled products in real time. The system will also apply several image pre-processing techniques such as resizing, normalising, and reducing noise in order to increase the accuracy of defect detection. The system will be developed using Python and associated libraries (e.g., Flask). By Utilizing AI & CV for QA Processes. Quality Control personnel will have the ability to upload product images and receive immediate feedback on defect detection results through the proposed solution incorporating artificial intelligence as part of quality inspection to provide a scalable, accurate, and cost-effective alternative to current defect detection practices. The results of this project demonstrate the real-world application of AI & Computer Vision in Smart Manufacturing and further illustrate the capabilities of intelligent systems to increase the quality control process of manufacturing.
Key Words: Product Defect Detection, Deep Learning, CNN, YOLOv8, Computer Vision, Smart Manufacturing, Industrial Automation
