Adopting Image Integration Techniques to Simulate Satellite Images

Abstract

Mathematical integration techniques rely on mathematical relationships such as addition, subtraction, division, and subtraction to merge images with different resolutions to achieve the best effect of the merger. In this study, a simulation is adopted to correct the geometric and radiometric distortion of satellite images based on mathematical integration techniques, including Brovey Transform (BT), Color Normalization Transform (CNT), and Multiplicative Model (MM). Also, interpolation methods, namely the nearest neighborhood, Bi-linear, and Bi-cubic were adapted to the images captured by an optical camera. The evaluation of images resulting from the integration process was performed using several types of measures; the first type depends on the determination of quality in the regions of the edges using a contrast measure as well as the number of edges and threshold. The second type is the global one that is based on the parameters of the image region, including the Mean (µ), Standard Deviation (SD), and Signal to Noise Ratio (SNR). The parameters also included the Amount of Information Added (AIA) to the original image, such as those for the total (AIA_t) , edges (AIA_e), and homogenous (AIA_h) regions. The results showed the efficiency of the integration process in the image fusion with different resolutions in one image integrated resolution. The quality measures used were also capable in evaluating the most efficient techniques and determining the accurate information of the resulting image.