Aurora image denoising with a modified bidimensional empirical mode decomposition method

The bidimensional empirical mode decomposition (BEMD) technique is a recent development in the area of image processing. Like facilitating many different applications, it enables noise removal from an image as well. However, the original BEMD process is associated with high computation time and other artifacts because of the application of two dimensional (2D) scattered data interpolation methods. In this paper, a modified BEMD (MBEMD) process is devised to make the process faster and reduce the decomposition artifacts. This modified BEMD is then utilized for noise removal from aurora images. In the proposed scheme, denoising filters are applied to the image components obtained from the BEMD process and then filtered components are combined to get the reconstructed image. Simulation results demonstrate the efficacy of the proposed approach for successful denoising of aurora images.