Deep Learning for Crack Detection: A Review of Learning Paradigms, Generalizability, and Datasets

2025-08-13

AI Summary

Deep learning for crack detection suffers from poor generalization, high annotation costs, and limited data modality. To address these challenges, this work systematically surveys the evolution of learning paradigms—from fully supervised to weakly supervised, unsupervised, and few-shot learning—and introduces 3DCrack, the first cross-domain benchmark dataset integrating RGB imagery and 3D LiDAR scans. Leveraging 3DCrack, we conduct a unified evaluation of mainstream approaches, including CNNs, Vision Transformers, self-supervised learning, domain adaptation, and large-model fine-tuning. Experimental results demonstrate that multimodal fusion and paradigm transfer significantly enhance cross-dataset generalization performance. We publicly release the dataset, standardized evaluation protocols, and reproducible baseline implementations. This work establishes a new benchmark and technical foundation for intelligent infrastructure inspection.

Technology Category

• Computer Vision: Multi-modal Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Machine Learning: Multimodal Learning

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Web applications in cross-disciplinary domains and verticals such as mixed reality, smart cities, and digital health
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

Crack detection plays a crucial role in civil infrastructures, including inspection of pavements, buildings, etc., and deep learning has significantly advanced this field in recent years. While numerous technical and review papers exist in this domain, emerging trends are reshaping the landscape. These shifts include transitions in learning paradigms (from fully supervised learning to semi-supervised, weakly-supervised, unsupervised, few-shot, domain adaptation and fine-tuning foundation models), improvements in generalizability (from single-dataset performance to cross-dataset evaluation), and diversification in dataset reacquisition (from RGB images to specialized sensor-based data). In this review, we systematically analyze these trends and highlight representative works. Additionally, we introduce a new dataset collected with 3D laser scans, 3DCrack, to support future research and conduct extensive benchmarking experiments to establish baselines for commonly used deep learning methodologies, including recent foundation models. Our findings provide insights into the evolving methodologies and future directions in deep learning-based crack detection. Project page: https://github.com/nantonzhang/Awesome-Crack-Detection

Problem

• Analyzing learning paradigm shifts in crack detection
• Improving generalizability across diverse crack datasets
• Expanding crack data beyond RGB to 3D sensors

Innovation

• Transitioning learning paradigms to diverse methods
• Enhancing generalizability via cross-dataset evaluation
• Introducing 3DCrack dataset with 3D laser scans