An Autonomous Robotized System for a Thermographic Camera

Konferenz: ISR/ROBOTIK 2010 - ISR 2010 (41st International Symposium on Robotics) and ROBOTIK 2010 (6th German Conference on Robotics)
07.06.2010 - 09.06.2010 in Munich, Germany

Tagungsband: ISR/ROBOTIK 2010

Seiten: 8Sprache: EnglischTyp: PDF

Persönliche VDE-Mitglieder erhalten auf diesen Artikel 10% Rabatt

Pretto, Alberto; Menegatti, Emanuele; Pagello, Enrico (University of Padova, Dep. of Information Engineering, Via Gradenigo 6/B, Padova, Italy)
Pretto, Alberto; Menegatti, Emanuele; Pagello, Enrico (IT+Robotics Srl, Contrà Valmerlara 21, Vicenza, Italy)
Bison, Paolo; Grinzato, Ermanno; Cadelano, Gianluca (CNR-ITC, Corso Stati Uniti 4, Padova, Italy)

This paper presents a new approach to automatically monitor an indoor environment on thermodynamic basis. It uses temperature as the driving parameter and is especially suited for comfort analysis or evaluation of moisture. The system measures all fundamental environment parameters (e.g., air temperature, relative humidity and air speed) by imaging with a thermal camera (IR camera) a set of special targets arranged in a grid (the reference grid), which can be placed close to a wall or in any other place of the room. The thermal camera is mounted on a pan-tilt unit to realize the monitoring process in an automatic way. The system processes the thermal images in real-time and autonomously controls the pan-tilt unit. A fast automatic learning procedure enables to recognize the special targets on the grid also in challenging environments and in different environment conditions, while a Particle Filter is used to update the state of the system (i.e., position of the intersection point between the optical axis of the camera and the planar surface of the grid). The system is able to perform a reliable global localization of the position of the thermal camera. During the scanning of the wall surfaces, a set of positions are automatically and sequentially reached by the moving IR camera: for each position a thermal image is recorded. Images are hence rectified in order to obtain a more accurate temperature sampling. We successfully tested our system in several challenging environments.