A Process Of Gaining High-Resolution Land Satellite Image By Convolutional Neural Network Approach

Abstract

An algorithm of satellite image was one of the important topic in the remote sensing field.  A deep learning methods in this field such as object detection and classification of image has led to the process to the issues of remote sensing.  CNNs are the important deep learning process which used in the classification of image.  The use of Convolutional Neural Network in satellite image division was new one.  Because of the computational difficulties of three dimension convolutional neural network that aim to remove both spectral and spatial data, two dimension CNNs highlighting on the removal of spatial data were preferred.  High resolution images consist of hard spectral data and also spatial data.  In this paper, a two and three dimension CNN design using spatial and spectral data was used to remove perfect data from high resolution images.  The design was analysed on a worldview 2 satellite picture adding agricultural material like uses of land like roads & buildings, tea and nut groves.  An output of the Convolutional Neural Network depend model are correlated against of the RF and SVM algorithms.  An accuracy of classification are existed by 900 points produced through web interface developed for the purpose of crowd sourcing.   It was 94.7% for the two and three dimension CNN design, 87.9% for the support vector machine and 89.7% for Random Forest.