accuracy-assessment

AcATaMa is an open-source QGIS plugin designed to provide comprehensive support for accuracy assessment and sample-based area estimation of raster thematic maps

🤖 Evaluate real robot policies using Gaussian splatting for soft-body simulations, enhancing the understanding of robot interactions in complex environments.

GRASS GIS shell scripts for per-year maximum-likelihood land-cover classification of a Landsat 8-9 OLI/TIRS time series (2015-2023) of the Sudd Wetlands, South Sudan, using k-means clustering and r.kappa accuracy assessment. Supports the article 'Image Segmentation of the Sudd Wetlands...' (Lemenkova, Analytics 2023, 2(3):745-780).

LaTeX source for the article 'Artificial Neural Networks for Mapping Coastal Lagoon of Chilika Lake, India, Using Earth Observation Data' (Lemenkova, J. Mar. Sci. Eng. 2024, 12(5):709). Comparison of RF, SVM and MLP/ANN classifiers in GRASS GIS.

Python scripts supplementing the MGISA landform mapping workflow (preprocessing, analysis, assessment, ensemble) described in Robillard (2025).

GRASS GIS shell scripts for maximum-likelihood classification of a Landsat 8 OLI/TIRS series (2013, 2017, 2022) analysing environmental patterns in Lake Chad, Central Africa (i.cluster, i.maxlik, r.kappa). Figures for Lemenkova, Die Bodenkultur 2023, 74(1):49-64.

GRASS GIS shell scripts for land-cover classification and mangrove-dynamics analysis of coastal Guinea-Bissau from a Landsat 8-9 OLI/TIRS time series (2017, 2020, 2023) using k-means clustering, i.maxlik classification, r.kappa accuracy and colour composites. Figures for Lemenkova, Transylvanian Rev. Syst. Ecol. Res. 2024, 26(2):17-30.

GRASS GIS scripts, clustering results and accuracy matrices for the article 'Image Segmentation of the Sudd Wetlands in South Sudan for Environmental Analytics by GRASS GIS Scripts' (Lemenkova, Analytics 2023, 2(3):745-780). Landsat 8-9 OLI/TIRS image segmentation, classification and NDVI.

This repository contains Python code for preliminary assessing the accuracy of geodetic (surveying) networks.

Land cover classification from Sentinel-2 with a random forest and k-means, using a spatial-block train/test split so the reported accuracy means something. CPU only, 56 tests.

GRASS GIS shell scripts implementing an automated maximum-likelihood algorithm to retrieve land-cover types in Guinea, West Africa, from a Landsat-8 OLI/TIRS time series (2014, 2018, 2023), with k-means clustering and 10-class classification. Figures for Lemenkova, Ovidius Univ. Annals of Constanta, Civil Eng. 2025, 25(1):19-36.

GRASS GIS shell scripts for per-year maximum-likelihood classification of a Landsat 8-9 OLI/TIRS series (2014-2022) monitoring the ephemeral salt lakes of Chotts Melrhir and Merouane, Algeria (i.cluster, i.maxlik, r.kappa). Figures for Lemenkova, Applied System Innovation 2023, 6(4):61.

GRASS GIS shell scripts monitoring seasonal land-cover fluctuations of saline lakes in northern Tunisia (Gulf of Hammamet and Gulf of Gabes) from a seasonal Landsat 8-9 OLI/TIRS series (2017, 2023) using k-means clustering, i.maxlik classification and r.kappa accuracy. Figures for Lemenkova, Land 2023, 12(11):1995.

GRASS GIS scripts (k-means clustering, i.maxlik and SVM via r.learn) plus a GMT topographic map and an R workflow diagram for land-cover classification of a Landsat 8-9 OLI/TIRS time series (2015-2023) of the Saloum River Delta, Senegal. Figures for Lemenkova, Earth 2024, 5(3):420-462.

GRASS GIS shell scripts for maximum-likelihood classification of Landsat 8 OLI/TIRS imagery (2015, 2023) mapping wetlands of Kenya (i.cluster, i.maxlik, r.kappa). Figures for Lemenkova, Transylv. Rev. Syst. Ecol. Res. 2023, 25(2):1-18.

LaTeX source for the article 'Image Segmentation of the Sudd Wetlands in South Sudan for Environmental Analytics by GRASS GIS Scripts' (Lemenkova, Analytics 2023, 2(3):745-780).

LaTeX source for the article 'Support Vector Machine Algorithm for Mapping Land Cover Dynamics in Senegal, West Africa, Using Earth Observation Data' (Lemenkova, Earth 2024, 5(3):420-462). GRASS GIS SVM vs k-means classification of Landsat OLI/TIRS.

LaTeX source for the article 'Improving Bimonthly Landscape Monitoring in Morocco, North Africa, by Integrating Machine Learning with GRASS GIS' (Lemenkova, Geomatics 2025, 5(1):5). Decision Tree and Extra Trees classification of a bimonthly Landsat time series in GRASS GIS.

GRASS GIS scripts and ANN/ML classification results (RF, SVM, MLP) for the article 'Artificial Neural Networks for Mapping Coastal Lagoon of Chilika Lake, India, Using Earth Observation Data' (Lemenkova, J. Mar. Sci. Eng. 2024, 12(5):709). Landsat 8-9 OLI/TIRS image classification and accuracy assessment.

LaTeX source for the article 'Monitoring Seasonal Fluctuations in Saline Lakes of Tunisia Using Earth Observation Data Processed by GRASS GIS' (Lemenkova, Land 2023, 12(11):1995).