Critical steps in camera pose estimation: an evaluation using LTI-LIB2 library
An evaluation of camera pose estimation methods using a chessboard pattern is presented. Steps evaluated in the estimation process are landmark point detection and camera parameter estimation, due to their critical role in the entire process. The ChESS method and a custom heuristic method are compar...
| Autores Principales: | Cabrera-Quirós, Laura, Campos-Gómez, Rafael, Castro-Godínez, Jorge |
|---|---|
| Formato: | Artículo |
| Idioma: | Español |
| Publicado: |
Editorial Tecnológica de Costa Rica
2014
|
| Materias: | |
| Acceso en línea: |
https://revistas.tec.ac.cr/index.php/tec_marcha/article/view/1656 https://hdl.handle.net/2238/4467 |
| Sumario: |
An evaluation of camera pose estimation methods using a chessboard pattern is presented. Steps evaluated in the estimation process are landmark point detection and camera parameter estimation, due to their critical role in the entire process. The ChESS method and a custom heuristic method are compared for chessboard pattern detection. Both methods are objectively contrasted using True Positive and False Negative criteria. Meanwhile, Zhang’s method for pose estimation based on planar surface point distribution is used as a first approach, and then refined with a nonlinear regression through the Levenberg-Marquardt algorithm. This pose estimation algorithm is evaluated through a comparison with a stable tool, such as the Camera Calibration Toolbox for Matlab®. |
|---|