Codes for cloud segmentation in ground-based infrared sky images adquiared using an sky imager mounted on a solar tracker. The codes were run in a High Performances Computer and the library for the paralellization of the code is MPI.
https://arxiv.org/pdf/2012.06930.pdf
CEEGE https://arxiv.org/pdf/2102.06646.pdf
IEEE PES https://arxiv.org/pdf/2102.10151.pdf
The generative models include in this repository are:
Gaussian Mixture Model (clustering) is GMM_segm.py.
K-means Clustering is KMS_segm.py.
Gaussian Discriminat Analysis (Linear) is GDA_segm.py.
Naive Bayes Classifier is NBC_segm.py.
The discriminative model were implemented in their primal formulation. The features were transformed to a feature space using split basis functions. See GDA_segm.
Ridge Regression for Classification is RRC_segm.py.
Suport Vector Machine is SVC_segm.py.
Gaussian Process for Classification cross-validated in parallel using MPI is GPC-MPI_segm.py.
The MRF implemented in this repository are:
Supervised Gaussian MRF is MRF_segm.py.
Unsupervised Gaussian MRF optimized via Independet Conditional Models cross-validated in parallel using MPI is ICM-MRF-MPI_segm.py.
Supervised Gaussian MRF with Simulate Anneling on the implementation is SA-MRF_segm.py.
Unsupervised Gaussian MRF optimized via Independet Conditional Models with Simulate Anneling on the implementation cross-validated in parallel using MPI is SA-ICM-MRF-MPI_segm.py.
See XX for further information.
The utils for loading, organized the vectors and matrices, processing the data, and common dependecies are in the files utils.py and feature_extraction_utils.py.
A sample dataset is publicaly available in DRYAD repository: https://datadryad.org/stash/dataset/doi%253A10.5061%252Fdryad.zcrjdfn9m