A Model for SIFT Optimization Using FPAA-Driven Block Convolution

dc.contributor.advisorKoziol, Scott M.
dc.contributor.authorEustice, Dylan Scott, 1992-
dc.contributor.departmentElectrical and Computer Engineering.en_US
dc.contributor.otherBaylor University.en_US
dc.contributor.schoolsHonors College.en_US
dc.date.accessioned2014-06-02T18:45:34Z
dc.date.available2014-06-02T18:45:34Z
dc.date.copyright2014
dc.date.issued2014-06-02
dc.description.abstractThe Scale Invariant Feature Transform (SIFT) is a useful algorithm for describing local features in an image. This research successfully demonstrates a model for optimizing SIFT using a Field Programmable Analog Array (FPAA). A method is presented which theoretically can reduce SIFT run time by nearly 50% by greatly limiting the image area required to search for SIFT features. Block convolution is a simplified, non-overlapping method of convolution which is computed with the FPAA. The block convolution between the input image and a specially designed kernel is used to determine potential regions of interest. The performance of several different types of kernels will be compared, most of which were produced using an evolutionary search algorithm. A threshold is then applied to the filtered image produced by the FPAA and regions which are unlikely to produce matches are masked. By removing the necessity to search in areas of the image where a match is unlikely to be found, we see a more efficient implementation of SIFT that also demonstrates the usefulness of new FPAA technology.en_US
dc.identifier.urihttp://hdl.handle.net/2104/9015
dc.language.isoen_USen_US
dc.rightsBaylor University projects are protected by copyright. They may be viewed from this source for any purpose, but reproduction or distribution in any format is prohibited without written permission. Contact libraryquestions@baylor.edu for inquiries about permission.en_US
dc.rights.accessrightsWorldwide access.en_US
dc.rights.accessrightsAccess changed 3/2/2017.
dc.subjectEvolutionary computingen_US
dc.subjectReconfigureable computingen_US
dc.subjectComputer visionen_US
dc.titleA Model for SIFT Optimization Using FPAA-Driven Block Convolutionen_US
dc.typeThesisen_US

Files

Original bundle

Now showing 1 - 2 of 2
Loading...
Thumbnail Image
Name:
Eustice, Dylan Thesis Final.pdf
Size:
9.89 MB
Format:
Adobe Portable Document Format
Description:
Thesis
No Thumbnail Available
Name:
Copyright and Availability Form.pdf
Size:
708.05 KB
Format:
Adobe Portable Document Format
Description:
Permissions Form