We provide a framework for constructing a guided robot for usage in hospitals in this thesis. The omnidirectional camera on the robot allows it to recognize and track the person who is following it. Furthermore, when directing the individual to their preferred position in the hospital, the robot must be aware of its surroundings and avoid accidents with other people or items. To train and evaluate our robot's performance, we developed an auto-labeling framework for creating a dataset of panoramic videos captured by the robot's omnidirectional camera. We labeled each person in the video and their real position in the robot's frame, enabling us to evaluate the accuracy of our tracking system and guide the development of the robot's navigation algorithms. Our research expands on earlier work that has established a framework for tracking individuals using omnidirectional cameras. We want to contribute to the continuing work to enhance the precision and dependability of these tracking systems, which is essential for the creation of efficient guiding robots in healthcare facilities, by developing a benchmark dataset. Our research has the potential to improve the patient experience and increase the efficiency of healthcare institutions by reducing staff time spent guiding patients through the facility.

We provide a framework for constructing a guided robot for usage in hospitals in this thesis. The omnidirectional camera on the robot allows it to recognize and track the person who is following it. Furthermore, when directing the individual to their preferred position in the hospital, the robot must be aware of its surroundings and avoid accidents with other people or items. To train and evaluate our robot's performance, we developed an auto-labeling framework for creating a dataset of panoramic videos captured by the robot's omnidirectional camera. We labeled each person in the video and their real position in the robot's frame, enabling us to evaluate the accuracy of our tracking system and guide the development of the robot's navigation algorithms. Our research expands on earlier work that has established a framework for tracking individuals using omnidirectional cameras. We want to contribute to the continuing work to enhance the precision and dependability of these tracking systems, which is essential for the creation of efficient guiding robots in healthcare facilities, by developing a benchmark dataset. Our research has the potential to improve the patient experience and increase the efficiency of healthcare institutions by reducing staff time spent guiding patients through the facility.

Dataset of Panoramic Images for People Tracking in Service Robotics

SHAMSIZADEH, SEPIDEH
2022/2023

Abstract

We provide a framework for constructing a guided robot for usage in hospitals in this thesis. The omnidirectional camera on the robot allows it to recognize and track the person who is following it. Furthermore, when directing the individual to their preferred position in the hospital, the robot must be aware of its surroundings and avoid accidents with other people or items. To train and evaluate our robot's performance, we developed an auto-labeling framework for creating a dataset of panoramic videos captured by the robot's omnidirectional camera. We labeled each person in the video and their real position in the robot's frame, enabling us to evaluate the accuracy of our tracking system and guide the development of the robot's navigation algorithms. Our research expands on earlier work that has established a framework for tracking individuals using omnidirectional cameras. We want to contribute to the continuing work to enhance the precision and dependability of these tracking systems, which is essential for the creation of efficient guiding robots in healthcare facilities, by developing a benchmark dataset. Our research has the potential to improve the patient experience and increase the efficiency of healthcare institutions by reducing staff time spent guiding patients through the facility.
2022
Dataset of Panoramic Images for People Tracking in Service Robotics
We provide a framework for constructing a guided robot for usage in hospitals in this thesis. The omnidirectional camera on the robot allows it to recognize and track the person who is following it. Furthermore, when directing the individual to their preferred position in the hospital, the robot must be aware of its surroundings and avoid accidents with other people or items. To train and evaluate our robot's performance, we developed an auto-labeling framework for creating a dataset of panoramic videos captured by the robot's omnidirectional camera. We labeled each person in the video and their real position in the robot's frame, enabling us to evaluate the accuracy of our tracking system and guide the development of the robot's navigation algorithms. Our research expands on earlier work that has established a framework for tracking individuals using omnidirectional cameras. We want to contribute to the continuing work to enhance the precision and dependability of these tracking systems, which is essential for the creation of efficient guiding robots in healthcare facilities, by developing a benchmark dataset. Our research has the potential to improve the patient experience and increase the efficiency of healthcare institutions by reducing staff time spent guiding patients through the facility.
People Tracking
Auto Collecting Data
Service Robots
File in questo prodotto:
File Dimensione Formato  
Shamsizadeh_Sepideh.pdf

accesso aperto

Dimensione 3.62 MB
Formato Adobe PDF
3.62 MB Adobe PDF Visualizza/Apri

The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/20.500.12608/48008