Satellite Images Analysis for Sargassum Recognition

The sargassum has become a problem in the south of Mexico; it arrives massively, and it is necessary to study available data concerning these algae to plan and take actions that minimize its effects when arrives to the coast. In this work we propose a method based on basic image processing techniques to identify sargassum presence from satellite images. We analyze the bands 4, 5 and 6 of Landsat images to visually detect sargassum. We employ floating algae and normalized difference vegetation indexes and the estimation of the temperature in surface earth to establish some kind of relationship between their presence and the season of the year. In this work, the spectral analysis of images shows that frequency amplitudes are constant except in the ones corresponding to sargassum where some anomalies are detected. This observation was verified in the selected bands. We conclude the proposed approximation can be useful to remotely study some characteristics and detect the sargassum (Figure 1) when it is getting close to coasts of the south of Mexico.  Figure 1. Original image and sargassum detection image. Acknowledgements: This work was supported by UNAM-PAPIIT IT101521