ALGORITMOS PARA CONSTRUÇÃO DE PANORAMA DE IMAGENS 360 E VISUALIZAÇÃO

• Main Authors: Alan Kazuo Hiraga, Francisco Assis da Silva, Almir Olivette Artero
• Format: Article
• Language: Portuguese
• Published: Universidade do Oeste Paulista 2013-07-01
• Series: Colloquium Exactarum
• Subjects: 360 panoroma; images retification; images adjustment; SIFT; RANSAC
• Online Access: http://revistas.unoeste.br/revistas/ojs/index.php/ce/article/view/863/953

This paper presents a methodology for the construction of mosaic images to form 360 panorama and an application for viewing. We used the SIFT and RANSAC algorithms commonly found in the literature for matching images. The algorithms for projection, adjustment of images and Blend for smoothing the joints were implemented in this paper using the OpenCV Computer Vision library. The methodology of this research techniques were applied to reduce the distortions caused in the joints of successive images, as well as increase the quality of the final panorama and get better performance. The viewer application developed in C# shows the 360 image resulting from images stitching into a cylinder with the point of view inside. The results of the experiments performed using the proposed technique has proved satisfactory, the mosaic formed by the joining of multiple pictures suffers little distortion, however, these distortions do not interfere with the final formation of the panorama 360, which provides good visual quality.