Long Wavelength Infrared (LWIR) cameras can provide a representation of a part of the light spectrum that is sensitive to temperature. These cameras also named Thermal Infrared (TIR) cameras are powerful tools to detect features that cannot be seen by other imaging technologies. For instance they enable defect detection in material, fever and anxiety in mammals and many other features for numerous applications. However, the accuracy of thermal cameras can be affected by many parameters; the most critical involves the relative position of the camera with respect to the object of interest.
Several models have been proposed in order to minimize the influence of some of the parameters but they are mostly related to specific applications. Because such models are based on some prior informations related to context, their applicability to other contexts cannot be easily assessed. The few models remaining are mostly associated with a specific device.
In this paper the authors studied the influence of the camera position on the measurement accuracy. Modeling of the position of the camera from the object of interest depends on many parameters. In order to propose a study which is as accurate as possible, the position of the camera will be represented as a five dimensions model. The aim of this study is to investigate and attempt to introduce a model which is as independent from the device as possible.
The Infrared Images and Other Data Acquisition Station enables a user, who is located inside a laboratory, to acquire visible and infrared images and distances in an outdoor environment with the help of an Internet connection. This station can acquire data using an infrared camera, a visible camera, and a rangefinder. The system can be used through a web page or through Python functions.
The question of how to map the 3D indoor temperature by infrared thermography is solved by a hybrid method
which is a combination of infrared thermography and the well known heat diffusion equation. The idea is to use
infrared thermography to get the surface temperature of each frontier of the 3D domain of interest. A suitable
procedure is devoted to this, allowing an automatic scanning of the whole frontier, the registration of data and
computation. These surface temperatures constitute the boundary conditions of the heat equation solved in the
domain of interest. The solution of the heat equation allows analyzing and controlling the temperature of every point
belonging to the considered domain. This temperature distribution is controlled over the time with a period of the
same order than the necessary time to obtain the frontier temperatures and at the end to contribute to the analysis of
the thermal comfort.
The study is done for the steady-state conditions under various weather situations. In this case the temperature
depends only on space coordinates. With such procedure, we can have an idea about the time necessary to reach
thermal equilibrium; time which has a great impact on the thermal comfort sensation. The results yielded by this
method are compared with those given by others techniques used for temperature measurement. Finally, the method
is used to access 3D temperature distribution for various geometric shapes.