Impact of gimbal tilt angle on 3D model reconstruction in cross and oblique photography flight lines

Résumé fourni par la source

With the wide application of drone technology for 3D reconstruction in smart cities and engineering monitoring, the requirements for the accuracy, efficiency, and clarity of model reconstruction keep improving continuously. The gimbal tilt angle is crucial for model reconstruction. However, there has been a prolonged lack of systematic research on the gimbal tilt angle, leading to actual drone photography without a quantitative rationale. In this work, we systematically study the impact of different gimbal tilt angles in cross and oblique photography flight lines on the accuracy, efficiency, and clarity of model reconstruction. Moreover, we quantitatively evaluate the reconstruction capability of the two flight plans and three mainstream 3D reconstruction software (ContextCapture, MipMap, and DJI Terra) with the same dataset. The results show that the cross flight line with a gimbal tilt angle of −60 deg yields optimal outcomes, while the oblique photography flight line with a combination of gimbal tilt angles of −90 deg and −45 deg yields optimal outcomes. The cross flight lines are better than the oblique photography flight lines for image acquisition. Furthermore, ContextCapture achieves the highest model accuracy, surpassing MipMap and DJI Terra by 34.9% and 38%, respectively. DJI Terra achieved the highest model reconstruction efficiency, improving by 62.9% and 38% compared with ContextCapture and MipMap, respectively. This work systematically reveals the quantitative impact of gimbal tilt angles on model reconstruction. Besides, the research provides a scientific rationale for the selection of the gimbal tilt angle, flight lines, and 3D reconstruction software in drone photogrammetry.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.