详细信息
RHCrackNet: Refined Hierarchical Feature Fusion and Enhancement Network for Pixel-Level Pavement Anomaly Detection ( EI收录) 被引量:1
文献类型:期刊文献
英文题名:RHCrackNet: Refined Hierarchical Feature Fusion and Enhancement Network for Pixel-Level Pavement Anomaly Detection
作者:Liu, Wenjing[1];Li, Zhenhua[1];Wang, Ji[1];Lu, Qingjie[1]
机构:[1]Guangdong Ocean Univ, Coll Elect & Informat Engn, Dept Elect Engn, Zhanjiang 524088, Peoples R China
年份:2025
卷号:8
期号:4
起止页码:880
外文期刊名:BIG DATA MINING AND ANALYTICS
收录:EI(收录号:20252118462700)、ESCI(收录号:WOS:001489646000012)、Scopus(收录号:2-s2.0-105005485674)、WOS
基金:This work was supported by the Key R&D Plan of Shanxi Province (No. 2023-ZDLGY-15), the GuangDong Basic and Applied Basic Research Foundation (No. 2023A1515110770), the Ocean Youth Talent Innovation Project of Zhanjiang (No. 2023E0010), the Non-funded Science and Technology Research Project of Zhanjiang (No. 2024B01051), and the Program for Scientific Research Start-up Funds of Guangdong Ocean University (Nos. 060302112319 and 060302112309).
语种:英文
外文关键词:Training; Measurement; Accuracy; Semantics; Noise; Network architecture; Feature extraction; Robustness; Convolutional neural networks; Anomaly detection; Convolutional Neural Network (CNN); anomaly detection; feature fusion; feature enhancement; non-local attention
外文摘要:Accurate and automatic detection of pavement anomaly is critical for damage assessment and pavements maintainence. While existing Convolutional Neural Network (CNN) approaches have achieved high performance, their robustness to texture noise is limited, and the completeness of detected pixel-level cracks remains uncertain due to insufficient extraction of contextual information. To address these limitations, we propose a novel pavement anomaly detection network called RHCrackNet. This model incorporates feature fusion modules and feature enhancement modules to dynamically aggregate high-level semantic features with low-level detail features and enhance them through attention mechanisms. In addition, a non-local attention module is introduced to learn long-range dependencies and improve the connectivity of detected subtle cracks. To further enhance performance, auxiliary structure loss and direction loss are developed for supervised training. Experimental results show that RHCrackNet is highly competitive with state-of-the-art methods on six real-world datasets and has good generalization capabilities.
参考文献:
正在载入数据...