Researchers at the University of Surrey have developed an innovative AI tool named Pose-Enhanced Geo-Localisation (PEnG) that significantly improves location accuracy without relying on GPS. This technology reduces localisation errors from 734 metres to just 22 metres, making it particularly effective in urban areas where GPS signals fail due to tall buildings or tunnels. PEnG utilizes a combination of satellite and street-level images along with relative pose estimation, providing reliable navigation even with a standard monocular camera. This groundbreaking system aims to enhance the reliability of navigation for various applications, including autonomous vehicles and robotics, especially in challenging conditions. With increasing global dependence on GPS, PEnG offers a robust alternative that mitigates potential disruptions from signal interference. Supported by the University of Surrey’s PhD Foundership Award, the team is now focused on developing prototypes for real-world testing, and their findings are open source to foster collaboration in advancing navigation technology.
Source link
Share
Read more