Please use this identifier to cite or link to this item: http://hdl.handle.net/2080/2419
Title: LANDSAT Enhanced Thematic Mapper Plus Image Registration using SIFT
Authors: Paul, S
Durgam, U K
Pati, U C
Keywords: Automatic image registration
Scale invariant feature transform
Affine transformation
Random sample consensus
Issue Date: Dec-2015
Publisher: IEEE
Citation: 2015 Third International Conference on Image Information Processing (ICIIP -2015), Waknaghat, Himachal Pradesh, India, 21 - 24 Dec 2015
Abstract: Enhanced Thematic Mapper is a eight-band multispectral sensor in LANDSAT satellite capable of providing high resolution images. These remote sensing images have a wide range of applications in agriculture, deforestation, land cover, volcanic flow activity monitoring etc. All these applications require very accurate image registration. But, automatic registration of these images is still a vital challenge as the images have significant translation, rotation, illumination and scaling differences. Scale invariant feature transform (SIFT) is capable of extracting invariant features from image pairs and it has its own descriptor to reliably match the corresponding features. In this paper, an automatic feature based image registration is proposed using scale invariant feature transform with a reliable matching technique to increase the number of correct matched features between image pairs. At first, SIFT based feature matching is implemented with cross matching process to remove the most of the outliers. Then, fine matched features are obtained by using RANSAC algorithm. Finally, simulation results represent the effectiveness of the proposed method.
Description: Copyright belongs to proceeding publisher
URI: http://hdl.handle.net/2080/2419
Appears in Collections:Conference Papers

Files in This Item:
File Description SizeFormat 
Lansat_paul_2015.pdf3.75 MBAdobe PDFView/Open


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.