Image de-blurring is the process of obtaining the original image by using the knowledge of the degrading factors. Degradation comes in many forms such as blur, noise, and camera mis-focus.
A major drawback of existing restoration methods for images is that they suffer from poor convergence properties, the algorithms converge to local minima, that they are impractical for real imaging applications.
Added to its disadvantage, some methods make restrictive assumptions on the PSF or the true image that limits the algorithm's portability to different applications. In conventional approach, deblurring filters are applied on the degraded images without the knowledge of blur and its effectiveness.
In this project, concepts of artificial intelligence are applied for restoration problem in which images are degraded by a blur function and corrupted by random noise. Autoencoders are used for the implementation of denoising and deblurring is done through generative adversarial networks where a discriminator is used to analyse each output image given by the generator.
An important advantage of the proposed model is that it can be implemented as an mobile application(android) and as an web application . Major application area of this proposed system is the processing of satellite images.