색상기반 대응점 후보군 검출방법을 사용한 파노라마 영상생성

Abstract

The method of panorama image generation is a technique in which multiple consecutive images are generated into a single image. Since multiple images are generated as one image, a process of extracting and matching interest points in neighboring images is important. The inaccurate matching process not only increases the panorama image generation time but also generates unnatural panorama images. In this paper, we propose a natural and faster panorama image generation method reducing the number of incorrectly matching point candidates by using the Hue value of the adjacent matching point candidates.