SteelBlastQC: Shot-blasted Steel Surface Dataset with Interpretable Detection of Surface Defects

TL;DR

This work tackles automated quality control for shot-blasted steel surfaces by introducing SteelBlastQC, a dataset of 1654 labeled RGB images (512-by-512) distinguishing 'ready for paint' from 'needs shot-blasting'. It benchmarks three approaches—Compact Convolutional Transformer (CCT), Support Vector Machine (SVM) with ResNet-50 features, and a Convolutional Autoencoder (CAE)—with CCT and SVM achieving up to 95% accuracy and CAE providing an unsupervised baseline. A key contribution is the interpretable outputs via heatmaps that highlight defect regions, enabling transparent decision-making in industrial settings. By releasing the dataset and baseline code, the work aims to accelerate research in industrial defect detection and support deployment of automated inspection systems in manufacturing.

Abstract

Automating the quality control of shot-blasted steel surfaces is crucial for improving manufacturing efficiency and consistency. This study presents a dataset of 1654 labeled RGB images (512x512) of steel surfaces, classified as either "ready for paint" or "needs shot-blasting." The dataset captures real-world surface defects, including discoloration, welding lines, scratches and corrosion, making it well-suited for training computer vision models. Additionally, three classification approaches were evaluated: Compact Convolutional Transformers (CCT), Support Vector Machines (SVM) with ResNet-50 feature extraction, and a Convolutional Autoencoder (CAE). The supervised methods (CCT and SVM) achieve 95% classification accuracy on the test set, with CCT leveraging transformer-based attention mechanisms and SVM offering a computationally efficient alternative. The CAE approach, while less effective, establishes a baseline for unsupervised quality control. We present interpretable decision-making by all three neural networks, allowing industry users to visually pinpoint problematic regions and understand the model's rationale. By releasing the dataset and baseline codes, this work aims to support further research in defect detection, advance the development of interpretable computer vision models for quality control, and encourage the adoption of automated inspection systems in industrial applications.

Figures (10)

• Figure 1: Example of data collected: passing quality control (left) and failing quality control (right). The defects depicted on the right are (from top to bottom) a welding line, two types of scratches, corrosion, and two examples of discoloration.
• Figure 2: Our proposed interpretable surface defect detection pipeline using supervised and unsupervised deep learning approaches
• Figure 3: Interpretation heatmap examples of truly and falsely classified images using SVM
• Figure 4: Comparing interpretation heatmap of SVM using RGB (left) and grayscale (right) input image with defective surface
• Figure 5: Interpretation heatmap examples of truly and falsely classified images using CCT