Time-Resolved Far Infrared Light Transport Decomposition for Thermal Photometric Stereo

Kenichiro Tanaka,Nobuhiro Ikeya,Tsuyoshi Takatani,Hiroyuki Kubo,Takuya Funatomi,Vijay Ravi,Achuta Kadambi,Yasuhiro Mukaigawa
DOI: https://doi.org/10.1109/tpami.2019.2959304
IF: 23.6
2021-06-01
IEEE Transactions on Pattern Analysis and Machine Intelligence
Abstract:We present a novel time-resolved light transport decomposition method using thermal imaging. Because the speed of heat propagation is much slower than the speed of light propagation, the transient transport of far infrared light can be observed at a video frame rate. A key observation is that the thermal image looks similar to the visible light image in an appropriately controlled environment. This implies that conventional computer vision techniques can be straightforwardly applied to the thermal image. We show that the diffuse component in the thermal image can be separated, and therefore, the surface normals of objects can be estimated by the Lambertian photometric stereo. The effectiveness of our method is evaluated by conducting real-world experiments, and its applicability to black body, transparent, and translucent objects is shown.
computer science, artificial intelligence,engineering, electrical & electronic
What problem does this paper attempt to address?