The problem
Apparent object size in an endoscope image changes with distance from the camera. Recovering shape and physical dimensions requires information that constrains distance or scale.
The research idea
A medical suture with a known diameter of 0.8 mm provides a reference. With a calibrated two-light endoscope, suture information estimates distance and reflection parameters; geometric and photometric constraints are then combined for shape recovery.
What was evaluated
The paper evaluates recovery using a polyp model and real endoscope imagery. It examines how a known-size reference can provide absolute scale for image-based shape reconstruction.
Considering applications
The design combines shape-from-shading with a known-size reference. Evaluation can compare estimated dimensions with independent measurements under defined acquisition and reference conditions.
Original research
Polyp Shape Recovery from Single Endoscope Image using Medical Suture
Usami, Hiroyasu, Iwahori, Yuji, Wang, Aili, Bhuyan, M. K., Ogasawara, Naotaka, Kasugai, Kunio
The Open Bioinformatics Journal, 2019 · Vol. 12, No. 1, pp. 1–17