Image-based modeling for high-quality 3D geometry and reflectance

Abstract

Recent advances in computer vision and graphics have enabled unprecedented levels of realism in photorealistic rendering, opening new dimensions in entertainment, education, and industrial applications. Rendering refers to a process of creating digital images of an object or a scene from 3D data using computers and algorithms. The 3D data includes 3D geometry, lighting conditions, shading information, and camera parameters of the object or the scene. Image-based modeling is the inverse process of rendering, i.e., reconstructing 3D data from 2D images. The 3D data to be recovered can be 3D geometry, reflectance of a surface, camera viewpoints, or lighting conditions. In this dissertation, three image-based modeling problems are discussed. First, image-based modeling using flash photography captures 3D geometry and reflectance of a static object using a single camera and a flashlight attached to the camera. An alternating and iterative optimization framework is proposed to jointly solve for several unknown properties such as 3D geometry and reflectance. Second, image-based modeling at microscale reconstructs 3D normals and reflectance of a surface at microscale. A specially designed acquisition system, as well as an image-based modeling algorithm for microscale material appearance, are proposed. Lastly, image-based modeling for human hair describes a novel 3D reconstruction algorithm for modeling high-quality human hair geometry. We believe that our work on these advanced image-based modeling problems will boost hyper-realism in computer graphics.