Automatic identification of shrub vegetation of the Teide National Park

Abstract

The identification of vegetation using remote sensing imagery is an important area of work with numerous applications such as the analysis of deforestation and crop monitoring. This work focuses on locating populations of an endemic shrub of the Canary Islands that grows in Teide National Park: the Teide broom. The objective is to determine automatically the location of this species in order to facilitate conservation efforts. For this purpose, a methodology is presented based on image analysis by using superpixels as the minimum image processing unit, combined with one-class classification. The proposed scheme has been analyzed by applying it to an area located at the northwest of the Teide National Park where the existence of populations of this shrub was known a priori.