A statistical method for crack detection in 3D concrete images

2024-02-25

arXiv.org

AI Summary

To address the high computational cost and trade-off between efficiency and recall in crack detection for large-scale concrete CT images, this paper proposes a training-free, lightweight statistical pre-localization method. The core innovation lies in establishing a crack prior detection framework based on local grayscale distribution modeling, integrating multi-scale sliding windows, local contrast enhancement, and adaptive threshold optimization to rapidly identify high-probability crack regions. By circumventing both the training dependency of deep learning models and the high computational complexity of classical algorithms, the method achieves a 92.7% pre-detection coverage rate on semi-synthetic and real-world 3D CT datasets, while reducing computation time by 68%. This significantly enhances both the efficiency and accuracy of subsequent high-resolution crack segmentation.

Technology Category

• Computer Vision: Segmentation
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization
• Search and Optimization: Local Search

Application Category

• Web Mining and Content Analysis: Large pretrained models with web data
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

In practical applications, effectively segmenting cracks in large-scale computed tomography (CT) images holds significant importance for understanding the structural integrity of materials. However, classical methods and Machine Learning algorithms often incur high computational costs when dealing with the substantial size of input images. Hence, a robust algorithm is needed to pre-detect crack regions, enabling focused analysis and reducing computational overhead. The proposed approach addresses this challenge by offering a streamlined method for identifying crack regions in CT images with high probability. By efficiently identifying areas of interest, our algorithm allows for a more focused examination of potential anomalies within the material structure. Through comprehensive testing on both semi-synthetic and real 3D CT images, we validate the efficiency of our approach in enhancing crack segmentation while reducing computational resource requirements.

Problem

• Detecting cracks in 3D concrete CT images with computational efficiency
• Overcoming computational challenges in large-scale high-resolution CT data
• Pre-localizing crack regions to optimize deep-learning segmentation workflows

Innovation

• Statistical framework for crack pre-localization in 3D CT
• Combines Hessian filter with spatial multiple testing
• Reduces computational complexity for segmentation networks