A vector SIFT operator for interest point detection in hyperspectral imagery

Hyperspectral imagery has been widely used in real applications such as remote sensing, agriculture, surveillance, and geological analysis. Matching hyperspectral images is a challenge task due to the high dimensional nature of the data. The matching task becomes more difficult when images with different dimensions, such as a hyperspectral image and an RGB image, have to be matched. In this paper, we address this problem by investigating structured support vector machine to learn graph model for each type of image. The graph model incorporates both low-level features and stable correspondences within images. The inherent characteristics are depicted by using graph matching algorithm on weighted graph models. We validate the effectiveness of our method through experiments on matching hyperspectral images to RGB images, and hyperspectral images with different dimensions.

show abstract

“…The SIFT [15] and vector SIFT [8] matching algorithms were applied between every band (or neighboring bands?) of images pairs.…”

Section: Methodsmentioning

confidence: 99%

“…Image matching performance can be further improved by developing better the keypoint detection approaches. For example, vector SIFT [8] operator extends SIFT based upon the multiple Gaussian scale space. Modified SIFT [20] descriptor preserves local edges with local contrast and differential excitation function.…”

Section: Related Workmentioning

confidence: 99%

Learning Graph Model for Different Dimensions Image Matching

Zhou

Yang

et al. 2015

Graph-Based Representations in Pattern Recognition

View full text Add to dashboard Cite

show abstract

“…A number of methods have been developed to perform image matching for registration [15,9,18,10,29] and change detection [14,27]. Closer to the work presented in this paper, researchers have investigated local features for detection and classification.…”

Section: Related Workmentioning

confidence: 99%

Estimating the spatial extents of geospatial objects using hierarchical models

Yang

Newsam

2012

2012 IEEE Workshop on the Applications of Computer Vision (WACV)

View full text Add to dashboard Cite

The goal of this work is to estimate the spatial extents of complex geospatial objects such as high schools and golf courses. Gazetteers are deficient in that they currently specify the spatial extents of these objects using a single latitude/longitude point. We propose a framework that uses readily available high resolution overhead imagery to estimate the boundaries of known object instances in order to update the gazetteers. Key to our approach is a hierarchical object model with three levels. The lowest level characterizes an object using local invariant features; an intermediate, latent level characterizes the land-use/landcover (LULC) classes that constitute an object; and, the top level models an object as a distribution over these classes.We evaluate our approach using a manually labeled ground truth dataset of four object types: high schools, golf courses, mobile home parks, and Costco shopping centers.

show abstract

Bio medical image retrieval using support vector machine

Surya¹

2021

International Conference on Recent Trends in Theoretical and Applied Physics: Icrttap

View full text Add to dashboard Cite

A vector SIFT operator for interest point detection in hyperspectral imagery

Cited by 7 publications

References 12 publications

Learning Graph Model for Different Dimensions Image Matching

Learning Graph Model for Different Dimensions Image Matching

Estimating the spatial extents of geospatial objects using hierarchical models

Bio medical image retrieval using support vector machine

Contact Info

Product

Resources

About