Sheet Metal Components with Crack Defects: Labeled Image Dataset for Quality Inspection

This dataset contains labeled images of sheet metal components for crack inspection tasks. The dataset includes both defective parts presenting cracks and non-defective parts. Each component is labeled at the image level according to its condition, and images are organized into separate folders reflecting the corresponding class. For each physical component, four images are provided, captured from different viewing angles to ensure multi-perspective inspection coverage. All images were acquired under controlled lighting conditions with backlight illumination, enhancing contour visibility and facilitating crack detection. The dataset is intended for research and development in automated visual inspection, computer vision, and machine learning applications for industrial quality control.