面向基础设施的三维探地雷达属性成像及信息可视化研究  

Three-Dimensional Ground Penetrating Radar Attribute Imaging and Information Visualization for Urban Infrastructure

作　　者：朱家松[1,2,3] 雷占占 罗享寰 ZHU Jiasong;LEI Zhanzhan;LUO Xianghuan(College of Civil and Transportation Engineering,Shenzhen University,Shenzhen 518060,Guangdong,China;Institute of Urban Smart Transportation&Safety Maintenance,Shenzhen 518060,Guangdong,China;Guangdong Provincial Key Laboratory of Urban Informatics,Shenzhen 518060,Guangdong,China;Shenzhen Metro Group Co.,Shenzhen 518026,Guangdong,China)

机构地区：[1]深圳大学土木与交通工程学院,广东深圳518060 [2]深圳大学城市智慧交通与安全运维研究院,广东深圳518060 [3]广东省城市空间信息工程重点实验室,广东深圳518060 [4]深圳市地铁集团有限公司,广东深圳518026

出　　处：《隧道建设（中英文）》2023年第10期1702-1711,共10页Tunnel Construction

基　　金：国家重点研发计划(2018YFB210100);深圳市地铁集团有限公司科研咨询服务(STJS-DT413-KY002/2021);国家自然科学基金(42204148);深圳市2021年工程建设领域科技计划项目(基于三维探地雷达的地铁隧道基础设施综合检测与健康管理研究)。

摘　　要：为解决探地雷达(GPR)三维图像(C-scan)信息维度低、成像效果主观等问题,以图像像素值的有效测量信息阈值及信息映射函数为研究对象,建立GPR C-scan的成像优化策略。通过以图像的SSIM和PSNR为评价指标,以室内钢筋混凝土墙体和道路路面为试验案例,量化分析变成像参数的可视化效果。试验结果表明,适合城市典型应用场景的GPR信息可视化方法包括映射阈值为70%及投影函数为平方函数。此外,提出融合振幅和频率的属性成像方法,丰富C-scan图像信息,组合信息可视化方法,实现准确表现目标物体几何形状并缓解噪声干扰的GPR C-scan成像,有助于减少人为判读误差,提高C-scan解译的可靠性。Ground penetrating radar(GPR)three-dimensional imaging,known as C-scan imaging,often suffers from limited information dimensionality and subjective image quality.In this study,the authors explore two critical imaging parameters,the valid measurement threshold and information transformation function,to formulate an optimization strategy for GPR C-scan imaging.Furthermore,the authors assess image quality using structural similarity and peak signal-to-noise ratio as evaluation metrics.To quantify the effectiveness of different imaging parameters,a case study is conducted involving indoor reinforced concrete walls and road surfaces.The experimental findings indicate that,for typical urban GPR applications,an optimal measurement threshold of 70%coupled with the mean method as the transformation function provides the best results.Moreover,an attribute imaging method that combines amplitude and frequency is proposed to enhance the information within C-scan images.This integrated visualization approach accurately portrays the geometric characteristics of target objects while reducing noise interference in GPR C-scan imaging.These enhancements are advantageous in minimizing human interpretation errors and enhancing the overall reliability of C-scan interpretation.

关 键 词：探地雷达 C-scan 多属性成像 映射优化 基础设施检测 

分 类 号：U45[建筑科学—桥梁与隧道工程]