Algorithm for monitoring the plankton population dynamics based on satellite sensing data

Abstract

Abstract The scientific work describes the algorithms for processing the multispectral water coastal imagery from satellite sensing data with the aim of identifying the phytoplankton population of a spotted structure: determining the contour, distributing color gradation and as a result - determining the concentration of phytoplankton distribution inside the zones and mass centers. Such characteristics let determine the speed of changing contours spots and their concentration, the mass center shift as a consequence of the water masses movement and the processes of phytoplankton growing and dying. All these may be done on the base of the processed image series of the same water area over different time (different dates). The combination of LBP and neural network methods are observed as algorithms for image processing and the results of computer experiments are presented.