| Abstract |
High-precision point positioning measurement has a wide range of applications in smart city construction. Global navigation satellite system real-time kinematic (RTK) technology is an important means for high-precision point positioning measurement. It is widely used in surveying projects such as engineering control measurement, topographic surveying, cadastral mapping, and construction staking. To expand the application scenarios and enhance the surveying functions, researchers have developed products like tilt RTK and visual RTK. However, current visual RTK products are still in an immature stage although products have been launched. Most visual RTK products require users to manually hold the surveying pole and walk a certain distance to form the photogrammetry baseline. During this process, users must keep the camera aimed at the target points from different angles. This manual operation is tedious and often leads to unreliable and inconsistent surveying results. Furthermore, the factors influencing the surveying accuracy have not been investigated in a quantitative way. To address these issues, this paper proposes a visual RTK positioning method without the handheld walking operation, instead relies on swinging the surveying pole with its tip in contact with the ground. It realizes high-precision photogrammetry through fully utilizing fix-point rigid body rotation constraints in such swing mode, and is named as inertial photogrammetry RTK pole. We conducted a detailed analysis and comparison of photogrammetry accuracy between this swinging mode and the traditional handheld walking mode. Field test results show that the proposed swinging mode (with only 1.5–2 m baseline) achieves an accuracy comparable to the handheld walking mode with over 8 m baseline, when surveying target points 15 m away. The surveying root mean squared error can approximately reach 6 cm for target points at 15 m, 4 cm for target points at 10 m, and 2 cm for target points at 6 m. The results prove that the proposed method is an accuracy-competitive, time-saving, operation-convenient, and application-friendly approach that can effectively improve current visual RTK. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2025. |
