Phenology based classification index method for land cover mapping from hyperspectral imagery

Abstract

Remote sensing imagery classification contributes assistance to real-time applications for comfort and secures the society. The imagery of satellites entirely depends on the sensor type in satellites. Phenology reflection varies based on the land cover type, which absorbs external energy. Multispectral high-resolution imagery has the maximum details about the earth’s surface. This research work defines phenology based classification approach, which can produce precise high precision land cover classification. The need to develop a phenology based methodology reflects on the vegetation development classification and produces a much more suitable land cover map based on reflection values. The RGB channel values of the image do not influence this technique of reflection phenology classification. Phenology Based Classification Index (PBCI) supervised method is used to classify the high-resolution multispectral imagery with improved phenology classification methods. PBCI works on the passive sensor satellite images, without clouds and shadow in classification. The proposed method has compared with existing phenology classification methods using more than seven quality metrics.