Deletion method of image low-frequency components using fast median filter algorithm

Abstract

This article proposes a computational method for removing low-frequency component of the raster image. Development of a calculation method is based on the fast median filtering. Presented method implementation removes low-frequency component for a larger image by dividing the image into sections for eliminating of the edge effect. The method can be usedfor enhancing the signal / noise ratio, reducing the dynamic range of images and processing time of large images, as well as allow for section-by-section filtering without the influence of the edge effect. Article also represents quality indicators of luminance image alignment. It was shown that a computer implementation of the proposed modification of the method of removing the low-frequency component of a larger image using a fast median filtering were made on FrameSmooth software.

Keywords

• processing of large images
• background image alignment