2018/03/02 by Fahime Amiri, Amiri, Fahime, Mohammad Hossein Kahaei +1
Chemistry · Earth and Planetary Sciences · Engineering · Environmental Science · #FOS: Electrical engineering #Remote Sensing and Land Use #Remote Sensing in Agriculture #Remote-Sensing Image Classification #Signal Processing (eess.SP) #Spectroscopy and Chemometric Analyses #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.1803.00873
openalex publication_date 2018/03/02 · openalex created_date 2022/09/30 · openalex updated_date 2026/07/28
This paper presents a semi-supervised hyperspectral unmixing solution that\nintegrate the spatial information in the abundance estimation procedure. The\nproposed method is applied on a nonlinear model based on polynomial\npostnonlinear mixing model where characterizes each pixel reflections composed\nof nonlinear function of pure spectral signatures added by noise. We\npartitioned the image to classes where contains similar materials so share the\nsame abundance vector. The spatial correlation between pixels belonging to each\nclass is modelled by Markov Random Field. A Bayesian framework is proposed to\nestimate the classes and corresponding abundance vectors alternatively. We\nproposed sparse Dirichlet prior for abundance vector that made it possible to\nuse this algorithm in semi-supervised scenario where the exact involved\nmaterials are unknown. In this approach, we just need to have a large library\nof pure spectral signatures including the desired materials. An MCMC algorithm\nis used to estimate the abundance vector based on generated samples. The result\nof implementation on simulated data shows the prominence of proposed approach.\n