Design of Rapid Image Mosaic Based on CUDA by 100-Megapixel Optical System

Abstract

To avoid the difficulty of image processing in the gigapixel optical system, this paper proposes a CUDA-based acceleration strategy for the rapid mosaic of a large-scale image and realizes the rapid mosaic of a large-scale image using the collaborative processing of CPU and GPU. Finally, the author designed an electronic hardware and software system and a fast splicing algorithm suitable for the rapid processing of large-scale image-based on a 100-megapixel monocentric multiscale optical system. And the real-time processing of data from nine 12 million pixel sub-cameras at 7.5 frames per second is completed to realize the real-time splicing of the video in the whole field of view.