Scalable Recurrent Neural Network for Hyperspectral Image Classification

Abstract

Image classification is one of the primary tasks in geocomputation, that is being used to categorize for further analysis such as land management, potential mapping, forecast analysis and soil assessment etc.  Image classification is the method by which labels or class identifiers are attached to individual pixels on basis of their characteristics. RNN classifies classes of images without any feature extraction step while other existing classification methods utilize rather complex feature extraction processes. Experiments on a series of satellite image data reveal that the suggested classification method can be a viable alternative to the existing feature extraction-based methods in classification performance and speed.