Acquisition of 3D surface temperature distribution of a car body

Thermal analysis using finite element modeling (FEM) is widely used in automotive industry nowadays. FEM can give simulation results at design stage. However, as FEM is a kind of simulations, it normally suffers from deviations with experimental results. Therefore, experimental measurement is necessary. To measure the 3D temperature distribution on a car body, we normally use a number of temperature sensors and then record the temperature values according to the positions of the sensors. It can provide the temperature values at different positions. Using this information, we can obtain the 3D temperature distribution on a car body. However, this method is time consuming and inconvenient, as we have to place a number of sensors on a car body, measure its corresponding positions as well as process the measured temperature values. In contrast, we firstly take a sequence of 2D thermal infrared images of the car body to have 2D temperature distribution and then reconstruct the 3D temperature distribution from the 2D images using computer vision method. It provides a more convenient method to acquire 3D temperature distribution compared to using a number of temperature sensors. The experimental results show the feasibility of our method.