Plug-and-play method for segmenting concrete bridge cracks using the segment anything model with a fractal dimension matrix prompt

Abstract

This paper addresses the diverse scenarios of bridge crack segmentation, proposing a method for detecting cracks on land and underwater using the Segment Anything Model (SAM) prompted by a fractal dimension matrix. The proposed method does not require additional training and obtains fractal feature information of cracks through fractal dimension matrix calculation. These feature information serve as prompt information for SAM to establish a plug-and-play crack segmentation method. The method achieves high detection performance, with a mean accuracy, IoU, and F1-Score of 99.6 %, 0.89, and 0.95 for land cracks, and 97.6 %, 0.89, and 0.95 for underwater cracks, respectively. This represents a significant improvement over methods that do not use the fractal dimension matrix for SAM prompts. Additionally, the method requires no additional training, showcasing excellent generalizability and practical potential for real-world applications in diverse environments.