Multi-sensor spatial augmented reality for visualizing the invisible thermal information of 3D objects

Abstract

In recent years, infrared imaging technology has been widely used to capture valuable thermal information of target objects for various applications such as construction engineering, advanced manufacturing, and medical diagnosis. However, it still remains a challenging task to visualize the invisible thermal information of three-dimensional objects for further evaluation and decision making. In this paper, we build a multi-sensor system consisting of a long-wave infrared camera, a RGBD camera, and a digital projector for multimodal data acquisition, real-time 3D thermographic reconstruction, and projector-based spatial augmented reality. We propose a hybrid method that combines the benefits of model-based ray casting and frame-based image warping to generate high-quality projected images with complete view coverage and abundant thermal details for visualizing the invisible thermal information on 3D objects. The effectiveness of the proposed method has been validated for two typical thermal imaging applications including human body temperature monitoring and non-destructive evaluation of composite materials.