Super-Resolution Using Deep Learning

Abstract

Digital images have a broad range of applications such as remote sensing, image and data storage for transmission in business applications, medical imaging, acoustic imaging, forensic sciences and industrial automation. Images acquired by satellites are useful in tracking of earth resources, geographical mapping, and prediction of agricultural crops, urban population, weather forecasting, flood and fire control. Space imaging applications include recognition and analyzation of objects contained in images obtained from deep space-probe missions. There are also medical applications such as processing of X-Rays, Ultrasonic scanning, Electron micrographs, Magnetic Resonance Imaging, Nuclear Magnetic Resonance Imaging, etc. If these images are degraded, blurred or not captured accurately, then we are not able to find these images wide variety of applications. So, we need to enhance quality of these images. The proposed method for enhancing these types of images in this report is convolutional neural networks. CNN is a mathematical model or computational model that tries to simulate the structure and functional aspects of biological neural networks. CNN is an adaptive system that changes its structure based on external or internal information that flows through the network during the learning phase. CNN are adjusted or trained to a specific target output which is based on a comparison of the output and the target, until the network output matches the target.