Surface Reconstruction and Sampling: Theory and Applications - From Point Clouds to Smooth Surfaces

Explore surface reconstruction mathematics in this 42-minute lecture from the ESI's "Infinite-dimensional Geometry" programme, examining the fundamental relationship between sampling and reconstruction of real-world surfaces. Dive into the challenges of determining optimal point quantities for faithful feature reconstruction and investigate the inverse problem of minimal point sampling on known surfaces. Learn about reconstruction algorithms and their theoretical bounds, while studying sampling techniques for various surface representations including meshes, smooth higher-order boundaries, subdivision limit surfaces, and signed distance functions. Understand practical applications such as scan data size reduction, error measurement, artifact handling, and simulation optimization, all guided by the principle that smooth surfaces contain richer information than their discrete geometric representations.