A Fuzzy Comprehensive Evaluation Approach for the 3D Modeling Accuracy of UAV Oblique Photogrammetry

Abstract

3D modeling is the important foundation of virtual reality and scene simulation. The technology of UAV oblique photogrammetry provides a new way for 3D modeling, and the accuracy becomes the key of its application. Since there are a lot of uncertainties in 3D modeling of UAV oblique photogrammetry, this paper proposed a fuzzy comprehensive evaluation approach to deal with this uncertain problem. Firstly, according to the composition of 3D modeling system based on UAV oblique photogrammetry, the main impact factors of 3D modeling accuracy are determined, and the indicator system of accuracy evaluation is therefore constructed. Secondly, by using the method of analytic hierarchy process, the weights and membership degrees of each indicator are calculated. Finally, the proposed approach is applied and validated in the case study. After the fuzzy mapping by the dot multiplication performance between the indicator weights and membership degrees, the comprehensive evaluation sets of different cases are obtained, and their corresponding 3D modeling accuracies can be determined based on the maximum membership degree principle. It is concluded that the fuzzy comprehensive evaluation approach proposed in this paper can quickly and correctly assess the 3D modeling accuracy of UAV oblique photogrammetry.