Persistence-Based Skeletonization of Images
Offered By: Applied Algebraic Topology Network via YouTube
Course Description
Overview
Explore persistence-based techniques for image skeletonization in this 30-minute lecture by Vitaliy Kurlin. Delve into the challenge of dividing images into convex polygons with subpixel precision vertices, allowing edges to follow any direction while adhering to object boundaries. Examine recent algorithms grounded in topological persistence, including the Convex Constrained Mesh (CCM) and the Persistent Resolution-Independent Mesh (PRIM). Learn how CCM processes straight line segments to produce a complete mesh of convex polygons, avoiding small angles and offering approximation guarantees. Discover how PRIM automatically detects straight-line edges based on contrast strength differences. Gain insights into advanced image processing techniques that combine algebraic topology and computer vision.
Syllabus
Vitaliy Kurlin (6/2/20): Persistence-based skeletonization of images
Taught by
Applied Algebraic Topology Network
Related Courses
Bradley Nelson: Parameterized Vietoris-Rips Filtrations via CoversApplied Algebraic Topology Network via YouTube Geometry and Topology of Periodic Point Sets, for Example Crystals
Applied Algebraic Topology Network via YouTube Dimensionality Reduction via Distributed Persistence - DIPOLE
Applied Algebraic Topology Network via YouTube Embeddings and Tverberg-Type Problems: New Algorithms and Undecidability Results
Applied Algebraic Topology Network via YouTube Pawel Dlotko - A Few Simple Stories on Topology in Action
Applied Algebraic Topology Network via YouTube