Method and system for creating a high-fidelity depiction of a scene containing an object

Abstract

A method and system for creating a high-fidelity depiction of a scene including an object within the scene are provided. The method includes uniformly illuminating a target surface of the object with light to obtain reflected, backscattered illumination. The method also includes sensing via a volumetric sensor, brightness of the surface due to a diffuse component of the backscattered illumination to obtain brightness information. Backscattered illumination from the target surface is inspected to obtain geometric measurements which include sensor noise. Rotationally and positionally invariant measured surface albedo including albedo noise of the object is computed based on the brightness and the geometric measurements. A machine-learning model such as a diffusion sensor model is applied to the geometric measurements and the measured surface albedo to remove the sensor noise and the albedo noise, respectively, to obtain a prediction of actual geometry and actual albedo, respectively, of the object.