Image processing is the recent trend in the computing, images play very crucial role in the field of engineering and technology. Many real-world applications used images as computation purpose. To find the high-resolution images still very challenging, due the capacity of the various acquisition devices, it is difficult to obtain the high-resolution images, still noise also present in the images. Various techniques for the image enhancement proposes by the various researchers to remove the noise present in the image. In this work generative adversarial network-based model is use here to improve image resolution. Adversarial network is the deep learning-based model which comprises various convolution layers, ReLU functions and normalization layers. GAN consists of discriminator and generator functions which is dedicated to perform different tasks. Single image super resolution-based GAN pretrained model used here to upscaling the images. This work upscale the different medical images with good PSNR values. All simulation performed in python environment with tensor flow and keras. The model training performs of colab with virtual GPU support.