Applicability of the kobotoolbox tool for validation of land use and land cover classification mapping

Abstract

For the validation of land cover and land use classification maps from satellite images by the unsupervised method, on-the-spot visits are necessary for data collection and verification of information from the theoretical plan. The objective of this article is to compare data collection methodologies in the field made with and without the support of the KoBoToolbox tool, aiming at the application of the Kappa Index to assess the thematic accuracy of the maps. In the field, the KoBoCollect application was used and it was found that the user will be able to automate the entire recording of the collected data. It was demonstrated the practicality of collecting data, so that, by formatting the tables with the collected information it is easier to organize the data for the use of the map algebra technique in geoprocessing, and subsequently the application of the validation formula for such maps. Through this research, it was possible to present options for the enrichment of data collection methodologies in the field, requiring further research that presents information or propositions of tools whose output files have extensions compatible with those of GIS, in order to automate the methods of map validation.