Fig. 3: Transmissive multi-view stereoscopic photo-thermoelectric (PTE) endoscope.
From: Robot-assisted, source-camera-coupled multi-view broadband imagers for ubiquitous sensing platform

a Schematic of the module and its fundamental operation mechanism. Operation of the transmissive multi-view stereoscopic PTE endoscope of the Methods section describes the corresponding detailed conditions. Supplementary Fig. 16 shows an evaluation of the fundamental endoscopy performances in response to the target size, shape and structure. b Non-destructive endoscopy of a plastic water pipe (in the sub-terahertz (sub-THz: λ = 1.15 mm) frequency region) in which the light-absorbing impurity was concealed. The impurity was visualised by detecting the locally attenuated transmission signals due to the photo-absorption, which corresponds to its location. c Non-destructive endoscopy of an actual gas pipe (in the far-infrared (FIR: λ = 10.3 µm) frequency region) with discretely spread breakages. The breakages were visualised by detecting local transmission signals corresponding to their locations. The multi-view stereoscopic PTE images were scanned at a speed of 10 Hz (b, c). d Self-driving transmissive multi-view stereoscopic PTE endoscope. Non-destructive unmanned remote endoscopy of a defective miniature L-shaped tunnel with discretely spread breakages was performed. We utilised multiple broadband FIR frequency radiators to illuminate the entire outer surface of the L-shaped tunnel. The breakages were visualised by detecting local transmission signals corresponding to the locations of the breakages. The self-driving endoscope was operated at a speed of 10 mm/s. L: scanning direction. e Three-dimensional PTE image reconstruction of the target tunnel. The experimentally obtained data were processed via the MATLAB software.