An Accurate Approach for Satellite Image Classification Using Neuroevolutionary Method |
Author(s): |
| Nisha , Sapthagiri College of Engineering; Pathanjali C, Sapthagiri College of Engineering |
Keywords: |
| NeuroEvolutionary Methods, Classification Techniques, NEAT, Neural Network, HyperNEAT |
Abstract |
|
Image classification is the growing need for the researchers and analysts who are dealing with huge number of satellite images in their everyday life. Deep Learning has shown quite impressive work in image classification field, but still gaining good accuracy is always an issue with this technique. NeuroEvolutionary Method appears as a solution for this problem. In this paper we present a neuro-evolutionary method called Enhanced HyperNEAT (E-HyperNEAT) which will be used to perform image classification and has shown better accuracy than the other existing methods. E-HyperNEAT has two main features that (1) it will evolve the network as well as the weights, and (2) it will remove the weak links in the network to manage with the processing time. So, E-HyperNEAT is more accurate and efficient method for image classification. |
Other Details |
|
Paper ID: IJSRDV4I30858 Published in: Volume : 4, Issue : 3 Publication Date: 01/06/2016 Page(s): 1017-1020 |
Article Preview |
|
|
|
|
