9:30am - 10:10am

Dr. Yifei Lou

Mathematics, UT Dallas

Title: Graph Regularized Models for Blind Hyperspectral Unmixing

Abstract: Blind hyperspectral image unmixing is the process of identifying the spectra of pure materials (i.e., endmembers) and their proportions (i.e., abundances) at each pixel. I will talk about a graph total variation (gTV) that regularizes the abundance map. To further alleviate the computational cost, we apply the Nystrom method to approximate a fully-connected graph by a small subset of sampled points, and adopt the Merriman-Bence-Osher (MBO) scheme by decomposing a grayscale image into a bit-wise form. A variety of numerical experiments on synthetic and real hyperspectral images are conducted, showcasing the potential of the proposed method.