10 February 2017 The relative pose estimation of aircraft based on contour model
Author Affiliations +
Proceedings Volume 10250, International Conference on Optical and Photonics Engineering (icOPEN 2016); 102502T (2017) https://doi.org/10.1117/12.2267118
Event: Fourth International Conference on Optical and Photonics Engineering, 2016, Chengdu, China
Abstract
This paper proposes a relative pose estimation approach based on object contour model. The first step is to obtain a two-dimensional (2D) projection of three-dimensional (3D)-model-based target, which will be divided into 40 forms by clustering and LDA analysis. Then we proceed by extracting the target contour in each image and computing their Pseudo-Zernike Moments (PZM), thus a model library is constructed in an offline mode. Next, we spot a projection contour that resembles the target silhouette most in the present image from the model library with reference of PZM; then similarity transformation parameters are generated as the shape context is applied to match the silhouette sampling location, from which the identification parameters of target can be further derived. Identification parameters are converted to relative pose parameters, in the premise that these values are the initial result calculated via iterative refinement algorithm, as the relative pose parameter is in the neighborhood of actual ones. At last, Distance Image Iterative Least Squares (DI-ILS) is employed to acquire the ultimate relative pose parameters.
© (2017) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Tai Fu, Tai Fu, Xiangyi Sun, Xiangyi Sun, } "The relative pose estimation of aircraft based on contour model", Proc. SPIE 10250, International Conference on Optical and Photonics Engineering (icOPEN 2016), 102502T (10 February 2017); doi: 10.1117/12.2267118; https://doi.org/10.1117/12.2267118
PROCEEDINGS
6 PAGES


SHARE
Back to Top