An Implementation of Futuristic Deep Learning Neural Network in Satellite Images for Hybrid Image Fusion
Aswin Kumer S V1, S K Srivatsa2
1Aswin Kumer S V, Department of Engineering and Communication Engineering, Koneru Lakshmaiah Education Foundation, Vaddeswaram, A.P., India.
2S K Srivatsa, Department of Engineering and Communication Engineering, MIT, Chromepet, India.

Manuscript received on 11 April 2019 | Revised Manuscript received on 17 May 2019 | Manuscript published on 30 May 2019 | PP: 484-487 | Volume-8 Issue-1, May 2019 | Retrieval Number: A4388058119/19©BEIESP
Open Access | Ethics and Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: Image fusion is the process of registering two different images of same scene to combine the individual image quality details to single image. This work is going to reduce the complexities while implementing image fusion by applying the Deep Neural Network (DNN) to analyse the input images to get the resultant output image. At first, the DNN checks whether the input images are in same size by checking the dimensions of the input images. Then, the DNN checks the resolution of the input images and it resize the low resolution image corresponds to the high resolution image for better image quality output, if the input images are in different sizes. Then it performs image rectification using the left side and right side images. After rectification, the DNN executes the image registration by matching the coordinates of the input images. Finally, the image fusion is performed with the input images and the resultant image is enhanced to improve the quality of an image and it can be evaluated by Mean Square Error (MSE), Peak Signal to Noise Ratio (PSNR), Correlation Coefficient (CC), Maximum Squared Error (MAXERR), Structured Similarity Index Measurement (SSIM) and Ratio of Squared Norms (L2RAT).
Index Terms: Deep Neural Network (DNN), Hybrid Image Fusion, Image Enhancement, Image Rectification.

Scope of the Article: Deep Learning