Enhancing DubaiSat-1 Satellite Imagery Using a Single Image Super-Resolution

摘要

DubaiSat-1 (DS1) captures multispectral images, which consist of three visible channels and one NIR with 5-meter resolution and a panchromatic channel with 2.5-meter resolution. Considering these resolutions, some important details might appear blurry on the image. Therefore, the concept of "Super Resolution" has been introduced to reconstruct the image in a way to overcome the inherent resolution limitations of imaging sensors. The aim of this study is to enhance the quality of the image by artificially increasing the number of pixels within the image to sharpen out-of-focus details or smooth rough edges in DubaiSat-1 images that have been enlarged using a general up-scaling process. Usually, images received from satellites go through many image processing and enhancement steps to increase the quality of these images. Many studies have been conducted in this field, and different techniques were suggested to improve the enhancement procedure. Studies were done to combine low resolution images of the same area to come up with an image with a high resolution and better quality. Image enhancement refers to operations used to get image features such edges, boundaries, or contrast in order to make a digital image more useful for display and analysis. In this paper, the proposed method is based on a well-known approach called "Example Based Super-Resolution". A single low resolution DubaiSat-1 image is used to construct its high resolution version. Simulation results illustrate a good performance of super-resolved images over certain magnification factors (i.e. M = 2,3 and 4).