Wyniki wyszukiwania - BazTech

1

Rock type discrimination using Landsat-8 OLI satellite data in mafic-ultramafic terrain

Tamilarasan Kuppusamy, Anbazhagan Siddan, Ranjithkumar Sundararajan

Geology, Geophysics and Environment

|

2023

|

Vol. 49, no. 3

281--298

EN

The mafic-ultramafic terrain of the Bhavani complex in southern India is considered for lithological mapping. The Landsat-8 OLI satellite data was used for the interpretation of different rock types in the study area. The satellite data were digitally processed using ENVI 5.6 image processing software. In the OLI data, excluding bands 8 and 9, the remaining seven bands were used for the generation of colour composite images, band ratios, principal component analysis and SVM classification. Reflectance spectral measurements were carried out in laboratory conditions for five rock samples collected from the study area. The XRF analysis was carried out to estimate the composition of major oxides present in the rock samples. The results obtained from XRF analysis were compared with the rock spectra in characterizing the spectral features of the rock types. The colour composite images (B543, B567, B456, and B457), PCA composite image (PC312 and PC456), band ratios (BR5/5 and BR4/3), colour composite images from band ratios, and SVM classified output are useful in delineation various rock types in the terrain.

2

Object Classification Using Support Vector Machines with Kernel-based Data Preprocessing

Adamiak K., Duch P., Ślot K.

Image Processing & Communications

|

2016

|

Vol. 21, no. 3

45--54

EN

The paper explores possibility of improving Support Vector Machine-based classification performance by introducing an input data dimensionality reduction step. Feature extraction by means of two different kernel methods are considered: kernel Principal Component Analysis (kPCA) and Supervised kernel Principal Component Analysis. It is hypothesized that input domain transformation, aimed at emphasizing between-class differences, would facilitate classification problem. Experiments, performed on three different datasets show that one can benefit from the proposed approach, as it provides lower variability in classification performance at similar, high recognition rates.

3

Data mining methods for gene selection on the basis of gene expression arrays

Muszyński M., Osowski S.

International Journal of Applied Mathematics and Computer Science

|

2014

|

Vol. 24, no. 3

657--668

EN

The paper presents data mining methods applied to gene selection for recognition of a particular type of prostate cancer on the basis of gene expression arrays. Several chosen methods of gene selection, including the Fisher method, correlation of gene with a class, application of the support vector machine and statistical hypotheses, are compared on the basis of clustering measures. The results of applying these individual selection methods are combined together to identify the most often selected genes forming the required pattern, best associated with the cancerous cases. This resulting pattern of selected gene lists is treated as the input data to the classifier, performing the task of the final recognition of the patterns. The numerical results of the recognition of prostate cancer from normal (reference) cases using the selected genes and the support vector machine confirm the good performance of the proposed gene selection approach.

4

Classification of fundus eye images using support vector machines for supporting glaucoma diagnosis

Stąpor K., Brueckner A.

Studia Informatica

|

2004

|

Vol. 25, nr 3

49-59

EN

In this paper the new method for automatic classification of fundus eye images into normal and glaucomatous ones is proposed. The cup region is automatically segmented from fundus eye images taken from classical fundus camera. The proposed method makes use of support vector machines classifier with Gaussian kernel. The mean sensitivity is 85 %, while specificity 90%.

PL

W artykule przedstawiono nową metodę automatycznej klasyfikacji cyfrowych obrazów dna oka na normalne i jaskrowe. Obszar wnęki naczyniowej zostaje automatycznie wysegmentowany na obrazie dna oka pozyskanego z klasycznej funduskamery. Zaproponowana metoda klasyfikacji wykorzystuje maszyny wektorów podpierających z jądrem Gaussowskim. Średnia czułość metody wynosi 85%, a specyficzność 90%.