• P. NARASIMHA RAO
• P. AKHILA
• P. SAI PRANEETH
• P. SREEJA
• D. RAVALI

Due to increase in the crime rates in the major cities, proof like surveillance cameras feed and audio recordings become very crucial in the investigation. But majority of the Indian cities or households do not contain a CCTV camera or a surveillance camera, even if it does, the captured images or video is not up to the mark as it would be highly pixelated and the culprits would get away due to this loophole. Even in the field of medical research and diagnosis, the MRI or brain tumor scans would contain a lot of noise and the scanned images are not clear which hinders an effective diagnosis by the medical practitioner. The common enemy in the above-mentioned scenarios is the availability of low-resolution image which is not very helpful in our work. So, to tackle the above mentioned and many other problems, we have developed a robust deep learning based Single Image Super Resolution model which try to tackle the problem of low-resolution images by converting the obtained low-resolution images into high resolution versions. In this work we try to develop deep learning-based models that try to take in the low-resolution images and provide a high-resolution version of that particular image. In this work we try to retain very crucial features of the low-resolution image like edges, colour textures, shadows etc. We start of by discussing the first ever models proposed in super resolution domain that is interpolation techniques, then we look over how the convolutional neural network-based model overshadowed the interpolation techniques, and to further enhance the image and give better results we study how Generative Adversarial Networks helps us in generating a high-resolution image with rich features in it.

Generative Adversarial Networks (GAN), Convolution Neural Network, Generator, Discriminator, Parametric Rectified Linear unit (PReLU), Leaky ReLU, Residual Blocks, Batch-Normalization, Dense.

"IMAGE RESTORATION USING RESIDUAL GAN", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.9, Issue 6, page no.c724-c730, June-2022, Available :http://www.jetir.org/papers/JETIR2206289.pdf

"IMAGE RESTORATION USING RESIDUAL GAN", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.9, Issue 6, page no. ppc724-c730, June-2022, Available at : http://www.jetir.org/papers/JETIR2206289.pdf