A Comprehensive Survey for Real-World Industrial Defect Detection: Challenges, Approaches, and Prospects

2025-07-15

AI Summary

Industrial defect detection faces core challenges including poor generalization to novel defects under closed-set paradigms, high annotation costs, and limited adaptability in multimodal settings. To address these, this paper systematically unifies and models both closed-set and open-set detection frameworks across 2D and 3D modalities. It presents the first comprehensive empirical comparison of these paradigms in realistic industrial scenarios, demonstrating that open-set approaches significantly reduce annotation dependency, enable reliable identification of unknown defects, and enhance deployment robustness. Integrating techniques from computer vision, deep learning, anomaly detection, and multimodal perception, the work encompasses supervised, unsupervised, and weakly supervised methods. We construct the most comprehensive taxonomy of industrial defect detection to date, explicitly identifying key challenges—such as few-shot learning, cross-domain generalization, and 3D geometry-texture coupling—as well as principled development pathways. This provides both theoretical foundations and practical guidance for algorithmic innovation and industrial deployment.

Technology Category

• Computer Vision: Multi-modal Vision
• Data Mining & Knowledge Management: Anomaly/Outlier Detection
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments

Abstract

Industrial defect detection is vital for upholding product quality across contemporary manufacturing systems. As the expectations for precision, automation, and scalability intensify, conventional inspection approaches are increasingly found wanting in addressing real-world demands. Notable progress in computer vision and deep learning has substantially bolstered defect detection capabilities across both 2D and 3D modalities. A significant development has been the pivot from closed-set to open-set defect detection frameworks, which diminishes the necessity for extensive defect annotations and facilitates the recognition of novel anomalies. Despite such strides, a cohesive and contemporary understanding of industrial defect detection remains elusive. Consequently, this survey delivers an in-depth analysis of both closed-set and open-set defect detection strategies within 2D and 3D modalities, charting their evolution in recent years and underscoring the rising prominence of open-set techniques. We distill critical challenges inherent in practical detection environments and illuminate emerging trends, thereby providing a current and comprehensive vista of this swiftly progressing field.

Problem

• Surveying industrial defect detection challenges and solutions
• Comparing closed-set and open-set defect detection methods
• Analyzing trends in 2D and 3D defect detection

Innovation

• Uses computer vision for defect detection
• Implements deep learning in 2D and 3D
• Adopts open-set frameworks for anomalies