Band selection of hyperspectral chlorophyll-a concentration inversion based on Parallel ant colony algorithm

In chlorophyll-A concentration inversion which based on hyperspectral remote sensing image, must use ground synchronous spectrum measurement data to find the best three band combination by iterative-method, we proposed the problem to analysis the existing band selection methods of chlorophyll-A concentration inversion and put forward a band selection method of the hyperspectral chlorophyll-A concentration inversion three-band model which based on parallel ant colony algorithm. The method is based on analysis of the optical properties and three band model theory, determines the interval of the sensitive band from hundreds of bands, then takes band reflectance of the sampling points and Correlation coefficient of sample points of chlorophyll-A concentration as the initial pheromone list, and updates distance parameter list constantly to let the parallel work ants find the best band combinations. Experimental results show that the method can select the optimal three-band fast and efficient by the remote sensing data and samples of chlorophyll-A concentration data, and the steps of chlorophyll-A concentration inversion can be simplified greatly.