Full metadata record
DC FieldValueLanguage
dc.contributorFaculty of Construction and Environmenten_US
dc.contributorDepartment of Land Surveying and Geo-Informaticsen_US
dc.contributor.advisorWu, Bo (LSGI)-
dc.creatorZeng, Hai-
dc.identifier.urihttps://theses.lib.polyu.edu.hk/handle/200/10529-
dc.languageEnglishen_US
dc.publisherHong Kong Polytechnic University-
dc.rightsAll rights reserveden_US
dc.titleImproved SIFT algorithm for lunar remote sensing image matching with large illumination differenceen_US
dcterms.abstractThis research presents an improved algorithm for Scale Invariant Feature Transform (SIFT) to perform reliable and robust matching for lunar remote sensing images with large illumination difference. The texture features of lunar remote sensing images are mainly dependent on the illumination. The improved SIFT algorithm, therefore, first introduces an approach to estimate the illumination orientation of images from histogram statistics for keypoint orientations. Then both the illumination orientation and computed orientation histogram are used to create a weight function that is used to add to gradient magnitude of each sample in keypoint neighbor when calculating the improved orientation of keypoints. Experiments demonstrated the improved approach is effective to reduce and eliminate illumination effects for keypoints. The same weight function is employed when computing the 128-Dimensioanl descriptor for keypoints. Based on the improved SIFT descriptors, this research also developed an improved matching method by combining the nearest Euclidean distance matching with the template matching and RANSAC filter. Experimental results showed that the improved SIFT algorithm can efficiently and reliably acquires approximate 40% more correct matches than using the original SIFT algorithm. By statistics, the improved algorithm is well performed to adapt for illumination changes even in 180 degrees, as more than 60% of newly obtained matching pairs are associated to (reverse) illumination orientation. At meantime, there are a lot of new matches coming from those regions lacking of texture where original SIFT algorithm cannot successful match them. The improved SIFT algorithm is, therefore, useful in lunar image matching for those textureless regions as well.en_US
dcterms.extentx, 101 pages : color illustrationsen_US
dcterms.isPartOfPolyU Electronic Thesesen_US
dcterms.issued2017en_US
dcterms.educationalLevelM.Sc.en_US
dcterms.educationalLevelAll Masteren_US
dcterms.LCSHRemote sensing -- Mathematicsen_US
dcterms.LCSHImage processing -- Digital techniquesen_US
dcterms.LCSHHong Kong Polytechnic University -- Dissertationsen_US
dcterms.accessRightsrestricted accessen_US

Files in This Item:
File Description SizeFormat 
991022385850303411.pdfFor All Users (off-campus access for PolyU Staff & Students only)7.88 MBAdobe PDFView/Open


Copyright Undertaking

As a bona fide Library user, I declare that:

  1. I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
  2. I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
  3. I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.

By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.

Show simple item record

Please use this identifier to cite or link to this item: https://theses.lib.polyu.edu.hk/handle/200/10529