THE ENMAP CONTEST: DEVELOPING AND COMPARING CLASSIFICATION APPROACHES FOR THE ENVIRONMENTAL MAPPING AND ANALYSIS PROGRAMME – DATASET AND FIRST RESULTS
Keywords: Hyperspectral Imaging, EnMAP, Classification, Benchmark Datasets, Evaluation Methodology
Abstract. The Environmental Mapping and Analysis Programme EnMAP is a hyperspectral satellite mission, supposed to be launched into space in the near future. EnMAP is designed to be revolutionary in terms of spectral resolution and signal-to-noise ratio. Nevertheless, it will provide a relatively high spatial resolution also. In order to exploit the capacities of this future mission, its data have been simulated by other authors in previous work. EnMAP will differ from other spaceborne and airborne hyperspectral sensors. Thus, the assumption that the standard classification algorithms from other sensors will perform best for EnMAP as well cannot by upheld since proof. Unfortunately, until today, relatively few studies have been published to investigate classification algorithms for EnMAP. Thus, the authors of this study, who have provided some insights into classifying simulated EnMAP data before, aim to encourage future studies by opening the EnMAP contest. The EnMAP contest consists in a benchmark dataset provided for algorithm development, which is presented herein. For demonstrative purposes, this report also represents two classification results which have already been realized. It furthermore provides a roadmap for other scientists interested in taking part in the EnMAP contest.