The Active Wave-front Sampling based 3D endoscope

Abstract

This thesis investigates the potential of Active Wave-front Sampling (AWS) for real time quantified 3D endoscopy. AWS is a technique by which phase information from an aperture area of a lens is obtained by sampling sub-regions of the aperture area in time. The resulting phase information is then processed to obtain an accurate 3D measure of an imaged target object. While AWS has multitudinous applications and can take on many forms, this thesis examines the possibility of using an off-the-shelf monoscope and a single camera in conjunction with a rotating off-axis aperture to generate quantified real time three dimensional surface models of a surgical scenario. This AWS based stereo-endoscope (referred to as the AWS-scope) is shown to possess optical attributes that are superior to those of the current industry standard. Quantified 3D data generated using this AWS-scope is used to create enhanced synthetic stereo image pairs that exhibit superior depth perception characteristics comparable to the current industry standard.