AI Article Synopsis

  • The paper discusses using deep learning techniques for detecting cracks in road images, focusing on a method called FetNet that employs a Swin transformer for better image segmentation.
  • This method captures detailed features at different levels and refines segmentation results using two special modules—the refinement attention module (RAM) and the panoramic feature module (PFM).
  • In tests with real-world data, FetNet outperformed other deep-learning models, achieving impressive metrics like 90.4% precision and demonstrating strong generalizability for complex road conditions.

Article Abstract

Structural health monitoring for roads is an important task that supports inspection of transportation infrastructure. This paper explores deep learning techniques for crack detection in road images and proposes an automatic pixel-level semantic road crack image segmentation method based on a Swin transformer. This method employs Swin-T as the backbone network to extract feature information from crack images at various levels and utilizes the texture unit to extract the texture and edge characteristic information of cracks. The refinement attention module (RAM) and panoramic feature module (PFM) then merge these diverse features, ultimately refining the segmentation results. This method is called FetNet. We collect four public real-world datasets and conduct extensive experiments, comparing FetNet with various deep-learning methods. FetNet achieves the highest precision of 90.4%, a recall of 85.3%, an F1 score of 87.9%, and a mean intersection over union of 78.6% on the Crack500 dataset. The experimental results show that the FetNet approach surpasses other advanced models in terms of crack segmentation accuracy and exhibits excellent generalizability for use in complex scenes.

Download full-text PDF

Source
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC11174510PMC
http://dx.doi.org/10.3390/s24113268DOI Listing

Publication Analysis

Top Keywords

detection road
8
road crack
8
crack images
8
segmentation method
8
crack
5
images based
4
based multistage
4
multistage feature
4
feature fusion
4
fusion texture
4

Similar Publications

Background: An association exists between obesity and reduced testosterone levels in males. The propose of this research is to reveal the correlation between 15 indices linked to obesity and lipid levels with the concentration of serum testosterone, and incidence of testosterone deficiency (TD) among adult American men.

Methods: The study utilized information gathered from the National Health and Nutrition Examination Survey (NHANES) carried out from 2011 to 2016.

View Article and Find Full Text PDF

Depression symptom severity and behavioral impairment in school-going adolescents in Uganda.

BMC Psychiatry

January 2025

Division of Epidemiology and Social Sciences, Institute for Health and Equity, Medical College of Wisconsin, 8701 Watertown Plank Road, Milwaukee, WI, 53226, USA.

Background: During adolescence, a critical developmental phase, cognitive, psychological, and social states interact with the environment to influence behaviors like decision-making and social interactions. Depressive symptoms are more prevalent in adolescents than in other age groups which may affect socio-emotional and behavioral development including academic achievement. Here, we determined the association between depression symptom severity and behavioral impairment among adolescents enrolled in secondary schools of Eastern and Central Uganda.

View Article and Find Full Text PDF

CSPG4 overexpression implicates higher risks of recurrence and tumorigenesis after surgical intervention of vocal fold Leukoplakia.

Eur Arch Otorhinolaryngol

January 2025

ENT institute and Department of Otorhinolaryngology, Eye & ENT Hospital, Fudan University, 83 FenYang Road, Shanghai, 200031, China.

Background: Vocal fold leukoplakia (VFL), a precancerous lesion of the larynx, is characterized by white plaques on the vocal fold mucous membrane. Currently, there are no reliable biomarkers to predict the recurrence and malignant transformation of VFL. Considering chondroitin sulfate proteoglycan 4 (CSPG4) as a biomarker for malignant tumors such as laryngeal squamous cell carcinoma (LSCC), we conducted this cohort study to evaluate the prognostic influence of CSPG4 expression on VFL patients.

View Article and Find Full Text PDF

The current work introduces the hybrid ensemble framework for the detection and segmentation of colorectal cancer. This framework will incorporate both supervised classification and unsupervised clustering methods to present more understandable and accurate diagnostic results. The method entails several steps with CNN models: ADa-22 and AD-22, transformer networks, and an SVM classifier, all inbuilt.

View Article and Find Full Text PDF

Fabrication of Ag based Surface Enhanced Raman Scattering substrates with periodic mask arrays by electron beam deposition.

Anal Chim Acta

February 2025

Engineering Research Center of Optical Instrument and System, Ministry of Education and Shanghai Key Lab of Modern Optical System, University of Shanghai for Science and Technology, No.516 Jungong Road, Shanghai, 200093, China.

Background: Surface-enhanced Raman scattering (SERS) has attracted much attention as a powerful detection and analysis tool with high sensitivity and fast detection speed. The intensity of the SERS signal mainly depended on the highly enhanced electromagnetic field of nanostructure near the substrate. However, the fabrication of high-quality SERS nanostructured substrates is usually complicated, makes many methods unsuitable for large-scale production of SERS substrates.

View Article and Find Full Text PDF

Want AI Summaries of new PubMed Abstracts delivered to your In-box?

Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!