Method to classify vegetation cover using satellite images and artificial intelligence

Herrera L.; Auccahuasi W.; Rojas K.; Urbano K.; Cuzcano A.; Carpio J.D.; Flores E.; Flores P.; Benites N.; Zamalloa L.; Sernaque F. · AIP Conference Proceedings (2023) · DOI: 10.1063/5.0125500

Resumen

Space technology is being used with greater emphasis in monitoring land cover, where the use of satellite images is used to analyze large areas of land, we can find optical satellite images that cover large areas of land, we present a methodology to be able to classify areas of vegetation cover present in the cadastre by means of satellite images, the classification is carried out by analyzing the chromatic characteristics that are extracted from the images. For which, two groups of images are created, corresponding to areas with the presence of vegetation and no vegetation. For the classification, the Matlab tool was used, from where a neural network was implemented to perform the classification, as well as a user interface for the use, manipulation and classification of the image, the results allow evaluating through the user interface of such that the neural network will be able to classify it.

Instituciones: Universidad Nacional Mayor de San Marcos; Universidad Continental; Universidad Nacional Federico Villarreal; Universidad Privada Norbert Wiener; Universidad César Vallejo; Universidad Privada del Norte