Comparative Analysis of Image Deblurring Deep Networks

It is often observed that most of the commonly captured pictures are acquired using the mobile cameras or the CCTV surveillance cameras. These mobile cameras capture the video footages of the activities of people who are stationary or in motion. These footages are often used as the evidence for different criminal cases which need to be restored for the proper information acquisition. Towards this deep research is happening to overcome the various issues influencing the quality of the video frames. We propose to eliminate blur and restore blurred images that are captured by different camera devices using GAN network. In this work we proposed a novel architecture and the performance is compared with pre-existing GAN models to overcome the issues of blurness in the captured video frames. The proposed solution promises to perform better for the restoration of the various types of video frames namely facial, scenaries, crowded areas, etc. from various noises.