Classification of Satellite Photographs Utilizing the K-Nearest Neighbor Algorithm
S. R. CENTRAL ASIAN JOURNAL OF MATHEMATICAL THEORY AND COMPUTER SCIENCES, 4 (6):
53-71(Juni 2023)
Zusammenfassung
Categorizing the various components of a satellite image is necessary for producing thematic maps, which requires the image to be analysed and classified first. We have suggested making use of Kohonen maps, which are able to train themselves utilising techniques of unsupervised and competitive learning in order to make this process more effective than the alternatives that came before it. The previous K-medoid clustering method is outperformed by these maps, which allow for more accurate picture categorization. The clustering functionality is handled by the Kohonen network, which does this by automatically analysing the similar characteristics of the pixels and allocating them to the same class as their similar counterparts. In addition to this, it helps reduce the dimensionality of the data. We have combined this with the K-Nearest Neighbor (KNN) classification technique, which is the one that is used the most frequently, in order to finally classify the processed data as being either irrigation land, green land, arid land, or aqua..
%0 Journal Article
%1 noauthororeditor
%A R, S. Silvia Priscila | S. Suman Rajest | R. Regin | Shynu T | Steffi.
%D 2023
%J CENTRAL ASIAN JOURNAL OF MATHEMATICAL THEORY AND COMPUTER SCIENCES
%K Algorithm, Classification Clustering, K-Nearest Kohonen Maps, Neighbor Technique Topological,
%N 6
%P 53-71
%T Classification of Satellite Photographs Utilizing the K-Nearest Neighbor Algorithm
%U https://cajmtcs.centralasianstudies.org/index.php/CAJMTCS/article/view/473/537
%V 4
%X Categorizing the various components of a satellite image is necessary for producing thematic maps, which requires the image to be analysed and classified first. We have suggested making use of Kohonen maps, which are able to train themselves utilising techniques of unsupervised and competitive learning in order to make this process more effective than the alternatives that came before it. The previous K-medoid clustering method is outperformed by these maps, which allow for more accurate picture categorization. The clustering functionality is handled by the Kohonen network, which does this by automatically analysing the similar characteristics of the pixels and allocating them to the same class as their similar counterparts. In addition to this, it helps reduce the dimensionality of the data. We have combined this with the K-Nearest Neighbor (KNN) classification technique, which is the one that is used the most frequently, in order to finally classify the processed data as being either irrigation land, green land, arid land, or aqua..
@article{noauthororeditor,
abstract = {Categorizing the various components of a satellite image is necessary for producing thematic maps, which requires the image to be analysed and classified first. We have suggested making use of Kohonen maps, which are able to train themselves utilising techniques of unsupervised and competitive learning in order to make this process more effective than the alternatives that came before it. The previous K-medoid clustering method is outperformed by these maps, which allow for more accurate picture categorization. The clustering functionality is handled by the Kohonen network, which does this by automatically analysing the similar characteristics of the pixels and allocating them to the same class as their similar counterparts. In addition to this, it helps reduce the dimensionality of the data. We have combined this with the K-Nearest Neighbor (KNN) classification technique, which is the one that is used the most frequently, in order to finally classify the processed data as being either irrigation land, green land, arid land, or aqua..},
added-at = {2023-10-03T14:45:28.000+0200},
author = {R, S. Silvia Priscila | S. Suman Rajest | R. Regin | Shynu T | Steffi.},
biburl = {https://www.bibsonomy.org/bibtex/2610d81b6a180f03ce22e2c0ab3a88eaf/centralasian_20},
interhash = {91b14638cfd1a872ba785602392854e4},
intrahash = {610d81b6a180f03ce22e2c0ab3a88eaf},
issn = {2660-5309},
journal = {CENTRAL ASIAN JOURNAL OF MATHEMATICAL THEORY AND COMPUTER SCIENCES},
keywords = {Algorithm, Classification Clustering, K-Nearest Kohonen Maps, Neighbor Technique Topological,},
language = {english},
month = {June},
number = 6,
pages = {53-71},
timestamp = {2023-10-03T14:45:57.000+0200},
title = {Classification of Satellite Photographs Utilizing the K-Nearest Neighbor Algorithm},
url = {https://cajmtcs.centralasianstudies.org/index.php/CAJMTCS/article/view/473/537},
volume = 4,
year = 2023
}