一种联合空谱特征的沥青路面裂缝检测方法  

A Method for Asphalt Road Surface Crack Detection Using Joint Spatial Spectral Features

作　　者：杜保坤 李广慧 马鹏阁[1] 谢祥兵[3] 钱金旺 DU Baokun;LI Guanghui;MA Pengge;XIE Xiangbing;QIAN Jinwang(School of Electronic Information,Zhengzhou University of Aeronautics,Zhengzhou 450046,China;Henan Vocational College of Water Conservancy And Environment,Zhengzhou 450008,China;School of Civil Engineering and Architecture,Zhengzhou University of Aeronautics,Zhengzhou 450046,China)

机构地区：[1]郑州航空工业管理学院电子信息学院,河南郑州450046 [2]河南水利与环境职业学院,河南郑州450008 [3]郑州航空工业管理学院土木建筑学院,河南郑州450046

出　　处：《郑州航空工业管理学院学报》2024年第4期79-85,共7页Journal of Zhengzhou University of Aeronautics

基　　金：国家自然科学基金(51378474;62301036);中原科技创新领军人才中原千人计划(194200510015);河南省科技攻关项目(192102310012;202102210136;231111212000);郑州航院研究生教育创新计划基金项目(2022CX60)

摘　　要：文章提出了一种基于高光谱图像的沥青路面裂缝检测算法,旨在解决传统方法下裂缝检测精度低、误检较多的问题。本方法利用高光谱成像技术获取道路的空间-光谱信息,并通过裂缝与路面的光谱信息差异提取裂缝,同时利用裂缝的空间特征信息对目标进行阈值分割、形态学运算等以此来提取裂缝主干,进而联合空间-光谱特征检测结果进行数学运算以实现对沥青路面裂缝的精确检测。实验结果表明,与ACI、VIS2算法相比,本文算法的F1-score分别提升了14.11%和29.03%,证明了该方法的有效性,同时针对不同类型和尺寸的裂缝目标表现出了高检测精度、强抗干扰性和显著的鲁棒性。The article introduces a novel hyperspectral imaging-based algorithm for detecting cracks in asphalt pave-ments,aiming to overcome the limitations of traditional methods,such as low accuracy and high false detection rates,thereby improving the efficiency of road maintenance.The method harnesses hyperspectral imaging technology to gather spatial-spectral information from the road,extracting cracks by exploiting the spectral discrepancies between the cracks and the road surface.It also utilizes spatial characteristics of cracks to conduct threshold segmentation and morphological operations,extracting the main structure of the cracks.Subsequently,it combines spatial and spectral feature detection results through mathematical computation to achieve precise identification of asphalt pavement cracks.The experimental outcomes indicate that the proposed algorithm surpasses the ACI and VIS2 methods,en-hancing the F1-score by 14.11%and 29.03%,respectively,thus validating its efficacy.The algorithm exhibits high detection accuracy,strong anti-interference and significant robustness for different types and sizes of crack targets.

关 键 词：高光谱图像 裂缝检测 空谱联合 光谱指数 形态学运算 

分 类 号：TN29[电子电信—物理电子学] TP79[自动化与计算机技术—检测技术与自动化装置]