| North Carolina State University Undergraduate Symposium |
2010 Summer NC State Undergraduate Research Symposium |
| Close Details |
| Session Time : 8/5/10 1:00 PM - 8/5/10 4:00 PM |
| Content Area : NSF Electrical and Computer Engineering REU |
|
Student Presenters : David John Fenwick Biomedical Engineering |
|
Mentors and/or Co-Authors : Joel Trussell Electrical & Computer Engineering |
|
Abstract Title : Image Stitching - Algorithms and Applications |
|
Abstract : Image stitching, the process of combining multiple images into a single composite image, is a widely-known and frequently used technique in the field of image processing. While many people are familiar with creating panoramic images using their camera, there are many additional problems that can be solved using existing image stitching techniques. Applications include object recognition, target tracking, computer assisted animation and the creation of ultra high-resolution images. Using MATLAB, we developed a platform to investigate all stages of the image stitching process. The code can stitch together images that are related by basic image transforms (combinations of translations and rotations). Feature point detection uses auto-correlation combined with a harmonic mean scalar interest measure. The control points that result from the feature detection are matched in multiple images using normalized cross-correlation. The matching control points are used to produce a collection of 2-D transformations to map the images to a single coordinate system corresponding to the composite image. |