Development of image stitching algorithm /
by Nicole J. Montecalvo and Dennis P. Anacan.
- Indang, Cavite : Cavite State University- Main Campus, 2012.
- xiv, 98 pages : illustrations ; 28 cm.
Thesis (Bachelor of Science in Computer Science) Cavite State University.
Includes bibliographical references.
College of Engineering and Information Technology (CEIT) College of Engineering and Information Technology (CEIT)
ANACAN, DENNIS P. & MONTECALVO, NICOLE J., Development of Image Stitching Algorithm. Undergraduate Thesis. Bachelor of Science in Computer Science. Cavite State University, Indang, Cavite. March 2012. Adviser: Mark Steve Poniente.
The study combines both direct and feature based approaches in producing a panoramic image. The researcher used Scale Invariant Feature Transform to extract keypoints as this can match keypoints under different scale and change in illumination that Traditional Feature Matching. Brute Force algorithm was used to exhaustively match the keypoints. The researcher developed an algorithm in image stitching used in image alignment and verification. The researchers implemented the algorithm in the software that can stitch and produce a panoramic image from a set of input images with overlapping fields. Incremental and Iterative Model was used in developing the software. The advantage of this methodology was that the previously learned idea can be integrated into the software development. The development of the software was under Windows XP Operating System but the executable file can also run under Windows 2007 Operating System. It was written in C++ programming language using GNU Compiler Collection (GCC), utilized Open Computer Vision and wxWidgets library for image processing procedures and creating Graphical User Interface (GUI) respectively. Fishbone diagram was used to illustrate the need for creating the software in which the algorithm was implemented.