|
|
System for People Detection and Localization Using Thermal Imaging Cameras
Charvát, Michal ; Kempter, Guido (oponent) ; Drahanský, Martin (vedoucí práce)
In today's world, there is an increasing need for automatic reliable mechanisms for detecting and localizing people -- from performing people flow analysis in museums, controlling smart homes to guarding hazardous areas like railway platforms. We propose a method for detecting and locating people using low-cost FLIR Lepton 3.5 thermal cameras and a~Raspberry Pi 3B+ computers. This thesis describes the continuation of the "Detection of People in Room Using Low-Cost Thermal Imaging Camera" project, which now supports modelling of complex scenes with polygonal boundaries and multiple thermal cameras observing them. In this paper, we introduce an improved control and capture library for the Lepton 3.5, a new person detection technique that uses the state-of-the-art YOLO (You Only Look Once) real-time object detector based on deep neural networks, furthermore, a new thermal unit with automated configuration using Ansible encapsulated in a custom 3D printed enclosure for safe manipulation, and last but not least, a step by step instruction manual on how to deploy the detection system in a new environment including other supporting tools and improvements. The results of the new system are demonstrated on a~simple people flow analysis performed in the Czech National Museum in Prague.
|
|
|
System for People Detection and Localization Using Thermal Imaging Cameras
Charvát, Michal ; Kempter, Guido (oponent) ; Drahanský, Martin (vedoucí práce)
In today's world, there is an increasing need for automatic reliable mechanisms for detecting and localizing people -- from performing people flow analysis in museums, controlling smart homes to guarding hazardous areas like railway platforms. We propose a method for detecting and locating people using low-cost FLIR Lepton 3.5 thermal cameras and a~Raspberry Pi 3B+ computers. This thesis describes the continuation of the "Detection of People in Room Using Low-Cost Thermal Imaging Camera" project, which now supports modelling of complex scenes with polygonal boundaries and multiple thermal cameras observing them. In this paper, we introduce an improved control and capture library for the Lepton 3.5, a new person detection technique that uses the state-of-the-art YOLO (You Only Look Once) real-time object detector based on deep neural networks, furthermore, a new thermal unit with automated configuration using Ansible encapsulated in a custom 3D printed enclosure for safe manipulation, and last but not least, a step by step instruction manual on how to deploy the detection system in a new environment including other supporting tools and improvements. The results of the new system are demonstrated on a~simple people flow analysis performed in the Czech National Museum in Prague.
|
host ::
RSS.