Comparison and analysis of transformation models on UAV image mosaic

Author

Yaping Wang ; Peihua Shao ; Yijin Chen

Author_Institution

Coll. of Geosci. & Surveying Eng., China Univ. of Min. & Technol. (Beijing), Beijing, China

fYear

2013

fDate

20-22 Dec. 2013

Firstpage

1042

Lastpage

1045

Abstract

In the precondition of getting the same feature points, the selection of different transformation models can greatly affect the speed and effectiveness of image mosaic. In this paper, the focus was on the determination of a transformation model from a given set of feature points. The high-resolution images captured by UAV (unmanned aerial vehicle) in the water-resource-investigation field were selected as source data. Four different transformation models-Affine Transformation, Quadratic Polynomial Transformation, Delaunay Triangulation and Homography Matrix were compared and analyzed for the optimal mosaic model of stereo pair in this situation. The experiment results indicated that Homography Matrix was the best in this case. Finally, it was given some useful advice for the selection of transformation models for the entire air strip mosaic.

Keywords

affine transforms; autonomous aerial vehicles; image resolution; image segmentation; matrix algebra; mesh generation; mobile robots; robot vision; Delaunay triangulation; UAV image mosaic; affine transformation; air strip mosaic; high-resolution images; homography matrix; quadratic polynomial transformation; stereo pair optimal mosaic model; transformation models; unmanned aerial vehicle; water-resource-investigation field; Analytical models; Atmospheric modeling; Educational institutions; Feature extraction; Image registration; Polynomials; Transmission line matrix methods; Affine Transformation; Delaunay Triangulation; Homography Matrix; Quadratic Polynomial Transformation; UAV image mosaic;

fLanguage

English

Publisher

ieee

Conference_Titel

Mechatronic Sciences, Electric Engineering and Computer (MEC), Proceedings 2013 International Conference on

Conference_Location

Shengyang

Print_ISBN

978-1-4799-2564-3

Type

conf

DOI

10.1109/MEC.2013.6885216

Filename

6885216