Detection of Pavement Cracks by Deep Learning Models of Transformer and UNet

TL;DR

This paper compares transformer-based and UNet deep learning models for pavement crack detection, finding that SwinUNet achieves the highest accuracy but with higher memory use, providing insights for future surface crack detection methods.

Contribution

It evaluates nine deep learning models, including transformer architectures, for pavement crack detection, highlighting the strengths of SwinUNet in accuracy and convergence.

Findings

Transformer models converge faster and are more accurate.

SwinUNet outperforms other models in accuracy.

Transformer models consume more memory and have lower efficiency.

Abstract

Fracture is one of the main failure modes of engineering structures such as buildings and roads. Effective detection of surface cracks is significant for damage evaluation and structure maintenance. In recent years, the emergence and development of deep learning techniques have shown great potential to facilitate surface crack detection. Currently, most reported tasks were performed by a convolutional neural network (CNN), while the limitation of CNN may be improved by the transformer architecture introduced recently. In this study, we investigated nine promising models to evaluate their performance in pavement surface crack detection by model accuracy, computational complexity, and model stability. We created 711 images of 224 by 224 pixels with crack labels, selected an optimal loss function, compared the evaluation metrics of the validation dataset and test dataset, analyzed the data details, and checked the segmentation outcomes of each model. We find that transformer-based models generally are easier to converge during the training process and have higher accuracy, but usually exhibit more memory consumption and low processing efficiency. Among nine models, SwinUNet outperforms the other two transformers and shows the highest accuracy among nine models. The results should shed light on surface crack detection by various deep-learning models and provide a guideline for future applications in this field.