Panoramic image stitching

• Main Author: Wong, Joniece Jing Yi
• Other Authors: Chua Chin Seng
• Format: Final Year Project
• Language: English
• Published: 2014
• Subjects: DRNTU::Engineering::Electrical and electronic engineering::Control and instrumentation
• Online Access: http://hdl.handle.net/10356/61093
• Institution: Nanyang Technological University

Panoramic images were first created in the middle of the 19th century and it became popular in modern days. It has progressed greatly through the use of computer image processing methods to create panoramas digitally. The word ‘panorama’ comes from the Greek words pân and hòrama, which meant ‘to see everything’. [2] An image stitching program is developed by the student to stitch images with an overlapped region. In this final year project, the Scale Invariant Feature Transform (SIFT) algorithm is used to detect interest points and compute their descriptors. These descriptors are used for matching interest points between images, and the images are combined using the information provided by the SIFT descriptors. This report also includes a brief introduction on two other popular interest point detectors, such as the Harris Corner Detector, the Local Binary Detector (LBP), and will focus mainly on the Scale Invariant Feature Transform (SIFT).