Method to measure, model, and predict depth and positioning errors of RGB-D Cameras in function of distance, velocity, and vibration
This thesis proposes a versatile methodology for measuring, modeling, and predicting errors as the Root Mean Square Error (RMSE) in depth and the Relative Positioning Error (RPE) using data captured from an RGB-D camera mounted on the top of a low-cost mobile robot platform. The proposed method has...
में बचाया:
| मुख्य लेखक: | |
|---|---|
| अन्य लेखक: | |
| स्वरूप: | doctoralThesis |
| भाषा: | pt_BR |
| प्रकाशित: |
Universidade Federal do Rio Grande do Norte
|
| विषय: | |
| ऑनलाइन पहुंच: | https://repositorio.ufrn.br/handle/123456789/44632 |
| टैग : |
टैग जोड़ें
कोई टैग नहीं, इस रिकॉर्ड को टैग करने वाले पहले व्यक्ति बनें!
|
| id |
ri-123456789-44632 |
|---|---|
| record_format |
dspace |
| institution |
Repositório Institucional |
| collection |
RI - UFRN |
| language |
pt_BR |
| topic |
RGB-D cameras Smart markers Visual mapping Visual localization Depth error Positioning error |
| spellingShingle |
RGB-D cameras Smart markers Visual mapping Visual localization Depth error Positioning error Fernandez, Luis Enrique Ortiz Method to measure, model, and predict depth and positioning errors of RGB-D Cameras in function of distance, velocity, and vibration |
| description |
This thesis proposes a versatile methodology for measuring, modeling, and predicting
errors as the Root Mean Square Error (RMSE) in depth and the Relative Positioning Error (RPE) using data captured from an RGB-D camera mounted on the top of a low-cost
mobile robot platform. The proposed method has three stages. The first one consists of
creating ground truth data for both 3D points (mapping) and camera poses (localization)
using the novel Smart Markers. The next stage is the acquisition of a data set for RMSE
and RPE errors computation using the mobile platform with the RGB-D camera. Finally,
the third step is to model and predict the errors in the measurements of depth and positioning of the camera as a function of distance, velocity, and vibration. For this modeling
and prediction stage, a simple approach based on Multi-Layer Perception neural networks
is used. The modeling results in two networks, the NrmseZ
for the depth error prediction
and the NRPE for the prediction of camera positioning error. Experiments show that the
NrmseZ
and NRPE have an accuracy of ± 1% and ± 2.5%, respectively. The proposed
methodology can be used straight in techniques that require an estimation of the dynamic
error. For example, in applications of probabilistic robotics for mapping and localization,
with RGB-D cameras mounted on Unmanned Aerial Vehicles, Unmanned Ground Vehicles, and also Unmanned Surface Vehicles (including sailboats). Tasks that use RGB-D
sensors, such as environmental monitoring, maintenance of engineering works, and public security, could rely on this approach to obtain the error information associated with
the camera measurements (depth and positioning). |
| author2 |
Gomes, Rafael Beserra |
| author_facet |
Gomes, Rafael Beserra Fernandez, Luis Enrique Ortiz |
| format |
doctoralThesis |
| author |
Fernandez, Luis Enrique Ortiz |
| author_sort |
Fernandez, Luis Enrique Ortiz |
| title |
Method to measure, model, and predict depth and positioning errors of RGB-D Cameras in function of distance, velocity, and vibration |
| title_short |
Method to measure, model, and predict depth and positioning errors of RGB-D Cameras in function of distance, velocity, and vibration |
| title_full |
Method to measure, model, and predict depth and positioning errors of RGB-D Cameras in function of distance, velocity, and vibration |
| title_fullStr |
Method to measure, model, and predict depth and positioning errors of RGB-D Cameras in function of distance, velocity, and vibration |
| title_full_unstemmed |
Method to measure, model, and predict depth and positioning errors of RGB-D Cameras in function of distance, velocity, and vibration |
| title_sort |
method to measure, model, and predict depth and positioning errors of rgb-d cameras in function of distance, velocity, and vibration |
| publisher |
Universidade Federal do Rio Grande do Norte |
| publishDate |
2021 |
| url |
https://repositorio.ufrn.br/handle/123456789/44632 |
| work_keys_str_mv |
AT fernandezluisenriqueortiz methodtomeasuremodelandpredictdepthandpositioningerrorsofrgbdcamerasinfunctionofdistancevelocityandvibration |
| _version_ |
1773963447129604096 |
| spelling |
ri-123456789-446322022-05-02T15:31:59Z Method to measure, model, and predict depth and positioning errors of RGB-D Cameras in function of distance, velocity, and vibration Fernandez, Luis Enrique Ortiz Gomes, Rafael Beserra http://lattes.cnpq.br/3725377549115537 http://lattes.cnpq.br/5849107545126304 Gonçalves, Luiz Marcos Garcia 32541457120 http://lattes.cnpq.br/1562357566810393 Silva, Bruno Marques Ferreira da http://lattes.cnpq.br/7878437620254155 Distante, Cosimo Clua, Esteban Walter Gonzalez http://lattes.cnpq.br/4791589931798048 RGB-D cameras Smart markers Visual mapping Visual localization Depth error Positioning error This thesis proposes a versatile methodology for measuring, modeling, and predicting errors as the Root Mean Square Error (RMSE) in depth and the Relative Positioning Error (RPE) using data captured from an RGB-D camera mounted on the top of a low-cost mobile robot platform. The proposed method has three stages. The first one consists of creating ground truth data for both 3D points (mapping) and camera poses (localization) using the novel Smart Markers. The next stage is the acquisition of a data set for RMSE and RPE errors computation using the mobile platform with the RGB-D camera. Finally, the third step is to model and predict the errors in the measurements of depth and positioning of the camera as a function of distance, velocity, and vibration. For this modeling and prediction stage, a simple approach based on Multi-Layer Perception neural networks is used. The modeling results in two networks, the NrmseZ for the depth error prediction and the NRPE for the prediction of camera positioning error. Experiments show that the NrmseZ and NRPE have an accuracy of ± 1% and ± 2.5%, respectively. The proposed methodology can be used straight in techniques that require an estimation of the dynamic error. For example, in applications of probabilistic robotics for mapping and localization, with RGB-D cameras mounted on Unmanned Aerial Vehicles, Unmanned Ground Vehicles, and also Unmanned Surface Vehicles (including sailboats). Tasks that use RGB-D sensors, such as environmental monitoring, maintenance of engineering works, and public security, could rely on this approach to obtain the error information associated with the camera measurements (depth and positioning). Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES Esta tese propõe uma metodologia versátil para medir, modelar e estimar erros como a Raiz do Erro Quadrático Médio (RMSE) na profundidade e o Erro no Posicionamento Relativo (RPE) da câmera utilizando dados capturados de uma câmera RGB-D montada no topo de uma plataforma robótica móvel de baixo custo. O método proposto tem três etapas, sendo que a primeira consiste na criação de dados que expressem um ground truth, tanto para pontos 3D (mapeamento) quanto para poses de câmera (localização) mediante o uso dos novos marcadores inteligentes. A próxima etapa é a aquisição de um conjunto de dados para cálculo dos erros RMSE e RPE utilizando a plataforma móvel com câmera RGB-D. Por fim, a terceira etapa consiste em modelar e estimar os erros nas medidas de profundidade e posicionamento da câmera em função da distância, velocidade e vibração. Para este estágio de modelagem e estimação, uma abordagem simples baseada em redes neurais do tipo Perceptron Multicamadas é usada. Isso resulta em duas redes, NrmseZ para a predição do erro de profundidade e NRPE para a previsão do erro de posicionamento da câmera. Experimentos mostram que as redes NrmseZ e NRPE têm uma precisão de ± 1% e ± 2,5%, respectivamente. A metodologia proposta pode ser usada diretamente nas técnicas que requerem uma estimativa do erro dinâmico. Como por exemplo em aplicações de robótica probabilística para localização e mapeamento, usando câmeras RGB-D montadas em Veículos Aéreos Não Tripulados, Veículo Terrestre Não Tripulados e também Veículos de Superfície Não Tripulados (incluindo veleiros robóticos). Tarefas que usam sensores RGB-D, tais como monitoramento ambiental, manutenção de obras de engenharia e segurança pública, podem contar com esta abordagem para obter informações sobre o erro associado às medições da câmera (profundidade e posicionamento). 2021-10-18T22:52:36Z 2021-10-18T22:52:36Z 2021-08-02 doctoralThesis FERNANDEZ, Luis Enrique Ortiz. Method to measure, model, and predict depth and positioning errors of RGB-D Cameras in function of distance, velocity, and vibration. 2021. 118f. Tese (Doutorado em Engenharia Elétrica e de Computação) - Centro de Tecnologia, Universidade Federal do Rio Grande do Norte, Natal, 2021. https://repositorio.ufrn.br/handle/123456789/44632 pt_BR Acesso Aberto application/pdf Universidade Federal do Rio Grande do Norte Brasil UFRN PROGRAMA DE PÓS-GRADUAÇÃO EM ENGENHARIA ELÉTRICA E DE COMPUTAÇÃO |