Student

Image Super Resolution (SR) has become a very gripping and predominant area of research to overcome the issues of limited resolution. Super resolution is a kind of technique used for obtaining an image of very high resolution by combining different low resolution images of the similar scene. It not only increases the size of an image but also restores the degraded image. It can be used sometimes in biomedical imaging to help doctors for making the correct diagnosis, in forensic investigations to extract even the minute information and in many other applications. This paper introduces a new approach based on wavelets where the interpolated high frequency sub band images obtained by discrete wavelet transform are incremented to the high frequency sub bands obtained by stationary wavelet transform. An intermediate stage has been introduced for estimating the high frequency sub band images by exploiting the difference image obtained by deducting the interpolated low frequency sub band image from the input image. Considerable improvement in the value of PSNR and MSE has been observed for biomedical images