The Method of Forming a Panorama with Increased Resolution in the Spatial Direction

摘要

The creation of one panorama from several images is today a popular user application in modern multimedia devices. Now, existing stitching methods strive to ensure maximum visual quality, while the algorithms allow the appearance of unreliable objects, such as additional lines and distorted proportions, which entails the loss of information. The task of the proposed method is to maximize the number of pixels in the original images. This gives advantages in the subsequent analysis of the merged images in areas such as security, pattern recognition, etc. In this paper, we present a method for forming a panorama with an increased resolution in the spatial direction with the possibility of a wide basis for the location of electrooptical converters for video analysis from moving platforms. Our method connects the pixel coordinates of the two images using graphs to optimally find the seam boundaries. An algorithm has also been developed to calculate a consistent alignment of overlapping frames.