Ground Penetrating Radar-Assisted Navigation and Path Planning Research
Ao Li, Longjuan Wang · Procedia Computer Science · 2025
Space positioning and navigation technology is an important component of modern intelligent systems, widely used in autonomous driving, drones, and robots. Traditional positioning and navigation methods face significant limitations. For example, GPS signals are obstructed indoors, underground, or in densely populated urban areas, resulting in reduced accuracy or complete failure of positioning services. To address these limitations, this article introduces a spatial positioning and navigation method based on ground penetrating radar (GPR). Ground penetrating radar, as a non-invasive underground target detection technology, has been widely used in fields such as urban road defect detection and product quality inspection. Given that most underground features remain relatively stable over time and are less susceptible to surface environmental changes, utilizing ground penetrating radar technology to perceive underground environments can enhance localization in challenging environments. This study proposes a new navigation method and improves existing path planning algorithms to meet the needs of this research. The method uses an image matching depth model to compare the similarity between images to achieve localization of a given area, and repeatedly performs image matching and position confirmation during the navigation process. The system achieves robust path navigation. The experimental results show that the improved path planning algorithm can reduce the complexity of the path, which meets the research needs. Moreover, this work achieves path planning and navigation from the starting point to the endpoint without the need for GPS or other navigation systems. It eliminates the need for global matching by focusing only on the corresponding position matching, thereby accelerating the inference process and improving the real-time performance of GPR based navigation.