Improved Convolutional Neural Network (Icnn) For Classification of Satellite Image Processing

Abstract

 Satellite imagery is significant for some, applications including disaster response, law enforcement, and environmental monitoring. These applications require the manual identification of objects and facilities in the imagery. Since the geographic scopes to be covered are incredible and the investigators accessible to direct the pursuits are not many, automation is required. However traditional object detection and classification algorithms are excessively inaccurate and temperamental to take care of the issue. Deep learning is a group of machine learning algorithms that have shown guarantee for the automation of such assignments. It has made progress in image understanding through convolutional neural networks. In this paper apply them to the issue of object and facility recognition in high-resolution, multi-spectral satellite imagery. In this paper proposed Improved Convolutional Neural Network (ICNN) for classification of satellite images and exploratory results show that the proposed strategy gives a best accuracy result.