The detection of surface defects on chrome faucets is a critical task in ensuring product quality in manufacturing. Due to the reflective nature of chrome, traditional vision systems face significant challenges, necessitating specialized techniques for accurate defect identification.

This study developed an AI-based vision inspection system tailored to these challenges, utilizing a controlled environment with a Basler camera and customized lighting setup to capture high-quality images of the faucets.

The captured images underwent preprocessing, including grayscale conversion, normalization, and segmentation, to enhance feature extraction. Various machine learning models, including Support Vector Machine (SVM), Random Forest, Gradient Boosting, and a Convolutional Neural Network (CNN), were trained on a dataset of 40 images—19 non-defective and 21 defective. The models were optimized and evaluated for their ability to accurately detect defects.

The study concludes that the AI-based system is effective in identifying surface defects on chrome faucets, offering a significant improvement over traditional methods and contributing to higher product quality and manufacturing efficiency.