Por favor, use este identificador para citar o enlazar este ítem:
http://infotec.repositorioinstitucional.mx/jspui/handle/1027/436| Deep learning point cloud registration based on distance features | |
| JORGE PEREZ GONZALEZ Fernando Luna Madrigal OMAR PIÑA RAMIREZ | |
| Acceso Abierto | |
| Atribución-NoComercial-SinDerivadas | |
| https://latamt.ieeer9.org/index.php/transactions/article/view/2856/361 | |
| Rigid registration Point cloud Deep Learning | |
| In this paper, a new method of rigid point cloudregistration called Points Registration Learning (PREL) is pre-sented. This algorithm is based on Deep Neural Networks trainedby sparse autoencoders and fed with a set of Euclidean andMahalanobis distance maps. Unlike other reported methods, wedo not assume closeness between point clouds or point pairs. Thisallows registering point clouds with a high degree of displacementor occlusion. PREL algorithm does not require an iterativeprocess, it estimates points distribution non-parametrically and itdoes not require a finer adjustment using other methods such asIterative Closest Point (ICP). To evaluate the proposed algorithm,two kinds of point cloud sets were used: one of them correspondsto real scenes acquired with an RGB-D camera and the otherset are surface reconstructions. When comparing PREL, ICPand Deep Closest Point (DCP) with Root Mean Square Error(RMSE), using points sets with a high degree of occlusion anddisplacement, ICP method shows an average RMSE of 98.8,followed by DCP with 32.51 and PREL with 0.75. These resultssuggest that PREL algorithm can be useful to reconstruct scenes,to scan objects and to register point clouds in any application,given the learning ability of the proposed algorithm. JCR del journal reportado al año de publicación del artículo (2019): 2.474 | |
| IEEE LATIN AMERICA TRANSACTIONS | |
| 2019-12 | |
| Artículo | |
| IEEE LATIN AMERICA TRANSACTIONS, volume 17, issue 12 | |
| Inglés | |
| Perez González, J., Luna Madrigal, F., Piña Ramírez, O., 2019. Deep learning point cloud registration based on distance features. IEEE LATIN AMERICA TRANSACTIONS, vol. 17, no. 12 | |
| OTRAS | |
| Versión publicada | |
| publishedVersion - Versión publicada | |
| Aparece en las colecciones: | Artículos |
Cargar archivos:
| Fichero | Tamaño | Formato | |
|---|---|---|---|
| JP_Deep learning_2019.pdf | 1.28 MB | Adobe PDF | Visualizar/Abrir |