A technique for automatically extracting regions of interest from ground penetrating radar data of roads based on the row-column variance method
-
摘要: 随着经济和社会的快速发展,道路承受的负载急剧增大,使得道路内部逐渐产生一系列的病害。探地雷达(ground penetrating radar,GPR)是一种无损探测技术,可将道路下方目标的回波信息呈现在雷达剖面图上,其中的空洞、脱空、疏松体等病害信息构成了探地雷达道路数据的感兴趣区域(region of interest,ROI)。传统的人工提取ROI方法对人员的技术要求高,同时针对海量数据的人工识别给一般人员的精力提出了不小的挑战。为此,本文提出一种通过在行列方差基础上加入阈值分割数据的方法,实现ROI的自动提取。实验结果表明,提出的方法有效地提取出多类型、多目标ROI位置信息。该方法在提高探地雷达道路检测效率方面具有较大的应用潜力。Abstract: With the rapid development of the economy and society,traffic loads have increased sharply,gradually causing a series of pavement diseases.Ground penetrating radar (GPR),which is a non-destructive testing technique,can present the echo information of subsurface targets on the GPR profile.The echo information of diseases,such as voids,cavities underneath the pavement,and loosely infilled voids,constitutes a region of interest(ROI) on the GPR profile.The traditional manual ROI extraction method features high technical requirements and high laborious intensity due to massive data.Therefore,this study proposed an automatic ROI extraction method that combines the threshold segmentation data and the row-column variance.The experimental results show that the method proposed in this study can effectively extract the location information of multi-type and multi-target ROIs.This method has great potential for improving road detection efficiency based on GPR.
-
-
计量
- 文章访问数: 39
- PDF下载数: 3
- 施引文献: 0
下载: