AI-Based Mobile Phone Screen Defect Detection Using OpenCV, YOLO, and Convolutional Neural Networks

Mobile phone screens are essential components of modern smartphones and may develop defects such as cracks,scratches, dark spots, and surface damage. Manual inspection can be time-consuming and may produce inconsistent results under changing lighting conditions. This paper presents a proposed artificial intelligence-based framework for automated mobile phone screen defect detection using OpenCV, You Only Look Once (YOLO), and convolutional neural networks (CNNs). OpenCV supports image acquisition and preprocessing, while YOLO is used to detect potential defects and estimate their locations. A CNN-based classifier may be incorporated when separate classification of detected regions is required. The system is designed to report defect categories, bounding boxes, and confidence scores, as well as indicate when no defect is detected above a selected threshold. The proposed framework describes the processing pipeline, system architecture, and evaluation strategy. Its effectiveness must be established through implementation and testing on a representative, accurately annotated dataset. Index Terms—Mobile phone screen, defect detection, OpenCV, YOLO, convolutional neural network, deep learning, computer vision, automated visual inspection.