Hi I am Gholamreza. I would like to get details on matlab code for maximum likelihood for satellite image classification ..My friend Justin said matlab code for maximum likelihood for satellite image classification will be available here and now i am living at Canada and i last studied in the college UBC and now am doing Research i need help on providing me with MATLAB code.
Posts: 14,118
Threads: 61
Joined: Oct 2014
The problem of deconvolution of satellite images is badly posed and must be regularized. In this sense, we use a regularization model of edge preservation using a function φ, which implies two hyperparameters. Our goal is to estimate optimal parameters to automatically reconstruct images. We propose to use the maximum likelihood estimator (MLE), applied to the observed image. We need sampling of previous and subsequent distributions. Since the convolution prevents the use of standard samplers, we have developed a modified Geman-Yang algorithm, using an auxiliary variable and a cosine transform. We present a Markov chain Monte Carlo maximum likelihood technique (MCMCML) that is able to simultaneously achieve estimation and reconstruction.
Classification is the task of assigning a set of given data elements to a given set of labels or classes. The main stages of image classification may include determination of an appropriate classification system, selection of training samples, preprocessing of images and extraction of characteristics and selection of suitable classification approaches.