Hybrid classification using combination of optimized spectral angle mapping algorithm and interpolation method on multispectral and hyper spectral image

Tembhurne, Omprakash; Malik, Latesh

doi:10.1109/iccca.2012.6179210

Cited by 3 publications

(3 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…So, the generalized angle (spectral similarity measure) as in [12] θ c×n ={θ ki } ( k = 1,…, c , and i = 1,…, n ) between the pixel spectra x i and the reference spectrum r k is defined as (1) where is the inner product of x i and r k . A pixel spectrum with the minimum or zero spectral angles of the reference spectrum is assigned to the class defined by the reference vector as in [13]. The dissimilarity between species increases with increasing the spectral angle θ ki .…”

Section: Spectral Angle Methods (Sam)mentioning

confidence: 99%

Application of Multispectral Images to Search for Construction Objects on the Spectral Signatures Base

Nazyrova¹,

Aitkozha²

2022

SJAITU

View full text Add to dashboard Cite

The work is devoted to the study of Landsat-8 multispectral images of not high resolution using the spectral angle method on the base of spectral signatures libraries to detect objects under construction in an urban area. The physical basis of the research method is that all objects have different reflection coefficients depending on the wavelength. This property makes it possible to identify various substances by their spectral signatures. In the work, an automatic comparison of the curves of the spectral reflectivity of objects on a low resolution space multispectral image was made to identify the identity of the characteristic energy absorption and reflection zones for detecting objects in the construction process. The article also describes the stages of image preprocessing, cross-track illumination correction of the image, atmospheric correction, and mathematical operations of bands transformation, which provide more opportunities for analysis and recognition of objects using a spectral study of a space image. The study accurately determines the presence or absence of the desired materials, since the search is based on the molecular structure of the substance. Also, the use of multispectral images allows you to analyze the entire city at the same time. The initial data was taken from a 2021 Landsat-8 satellite image with 11 bands, with a resolution of 30 meters, which was enhanced to 15 meters during pre-processing. The results of the search and detection of objects under construction in the city are given. The detection results can be used as input data for further in-depth analysis.

show abstract

Section: Spectral Angle Methods (Sam)mentioning

confidence: 99%

Application of Multispectral Images to Search for Construction Objects on the Spectral Signatures Base

Nazyrova¹,

Aitkozha²

2022

SJAITU

View full text Add to dashboard Cite

show abstract

“…The SAM algorithm is a linear model that is proposed based on the theoretical assumption that a single pixel in HSI can only represent one target [5]. The SAM algorithm aims to measure the similarity of two spectrums by calculating the angle between the two spectral vectors according to Eq.…”

Section: Technology Background 21 Sam Algorithmmentioning

confidence: 99%

“…Among these methods, the spectral angle mapper (SAM) algorithm has been widely used. In the SAM algorithm, one sample would be assigned to a class only if the angle between the unlabeled sample spectrum and the specific class reference spectrum is minimum or close to zero [5]. The SAM algorithm follows a linear classification model and can show better performance for HSI classification with little spectrum mixture.…”

Section: Introductionmentioning

confidence: 99%

Spectral–spatial hyperspectral classification based on multi-center SAM and MRF

Tang

Xiao

et al. 2015

Opt Rev

View full text Add to dashboard Cite

In this paper, a novel framework for an accurate spectral-spatial classification of hyperspectral images is proposed to address nonlinear classification problems. The algorithm is based on the spectral angle mapper (SAM), which is achieved by introducing the multi-center model and Markov random fields (MRF) into a probabilistic decision framework to obtain an accurate classification. Experimental comparisons between several traditional classification methods and the proposed MSAM-MRF algorithm have demonstrated that the performance of the proposed MSAM-MRF algorithm outperforms the traditional classification algorithms.

show abstract

A Kernel Spectral Angle Mapper algorithm for remote sensing image classification

Liu

Yang

2013

2013 6th International Congress on Image and Signal Processing (CISP)

View full text Add to dashboard Cite

Hybrid classification using combination of optimized spectral angle mapping algorithm and interpolation method on multispectral and hyper spectral image

Cited by 3 publications

References 10 publications

Application of Multispectral Images to Search for Construction Objects on the Spectral Signatures Base

Application of Multispectral Images to Search for Construction Objects on the Spectral Signatures Base

Spectral–spatial hyperspectral classification based on multi-center SAM and MRF

A Kernel Spectral Angle Mapper algorithm for remote sensing image classification

Contact Info

Product

Resources

About