Motus: A Framework for Human Motion Classification in a Not-controlled Moving Environment
This work introduces a framework proposal based on various algorithms, processes, and methods to classify Motion Capture (MoCap) data. To provide a generalized model for MoCap data classification, the approach is defined step by step: data collecting, data cleansing, segmentation, data pre-processin...
Autores Principales: | Rodríguez-González, Joselyn, Hernández-López, María, Siles-Canales, Francisco |
---|---|
Formato: | Artículo |
Idioma: | Inglés |
Publicado: |
Editorial Tecnológica de Costa Rica (entidad editora)
2020
|
Materias: | |
Acceso en línea: |
https://revistas.tec.ac.cr/index.php/tec_marcha/article/view/5076 https://hdl.handle.net/2238/12070 |
Sumario: |
This work introduces a framework proposal based on various algorithms, processes, and methods to classify Motion Capture (MoCap) data. To provide a generalized model for MoCap data classification, the approach is defined step by step: data collecting, data cleansing, segmentation, data pre-processing, feature selection, model selection, and validation. For each step, we selected and evaluated algorithms, process and methods have shown good performance in previous studies, all of them were proved and validated in BVH databases, but in not freely moving environment. |
---|