Statistical learning and optimization methods for improving the efficiency in landscape image clustering and classification problems

Download

index.pdf

Date

2005

Author

Gürol, Selime

Metadata

Show full item record

Item Usage Stats

100
views

60
downloads

Remote sensing techniques are vital for early detection of several problems such as natural disasters, ecological problems and collecting information necessary for finding optimum solutions to those problems. Remotely sensed information has also important uses in predicting the future risks, urban planning, communication.Recent developments in remote sensing instrumentation offered a challenge to the mathematical and statistical methods to process the acquired information. Classification of satellite images in the context of land cover classification is the main concern of this study. Land cover classification can be performed by statistical learning methods like additive models, decision trees, neural networks, k-means methods which are already popular in unsupervised classification and clustering of image scene inverse problems. Due to the degradation and corruption of satellite images, the classification performance is limited both by the accuracy of clustering and by the extent of the classification. In this study, we are concerned with understanding the performance of the available unsupervised methods with k-means, supervised methods with Gaussian maximum likelihood which are very popular methods in land cover classification. A broader approach to the classification problem based on finding the optimal discriminants from a larger range of functions is considered also in this work. A novel method based on threshold decomposition and Boolean discriminant functions is developed as an implementable application of this approach. All methods are applied to BILSAT and Landsat satellite images using MATLAB software.

Subject Keywords

Mathematics.

URI

http://etd.lib.metu.edu.tr/upload/12606595/index.pdf
https://hdl.handle.net/11511/15476

Collections

Graduate School of Applied Mathematics, Thesis

Suggestions

OpenMETU
Core

Digital image processing of remotely sensed oceanographic data Türkmen, Müşerref; Öktem, Hakan; Department of Scientific Computing (2008) Developing remote sensing instrumentation allows obtaining information about an area rapidly and with low costs. This fact offers a challenge to remote sensing algorithms aimed at extracting information about an area from the available re¬mote sensing data. A very typical and important problem being interpretation of satellite images. A very efficient approach to remote sensing is employing discrim¬inant functions to distinguish different landscape classes from satellite images. Various methods on this dire...
Nonlocal operators with local boundary conditions in higher dimensions Aksoylu, Burak; Celiker, Fatih; Kilicer, Orsan (Springer Science and Business Media LLC, 2019-02-01) We present novel nonlocal governing operators in 2D/3D for wave propagation and diffusion. The operators are inspired by peridynamics. They agree with the original peridynamics operator in the bulk of the domain and simultaneously enforce local boundary conditions (BC). The main ingredients are periodic, antiperiodic, and mixed extensions of separable kernel functions together with even and odd parts of bivariate functions on rectangular/box domains. The operators are bounded and self-adjoint. We present al...
Application of the boundary element method to parabolic type equations Bozkaya, Nuray; Tezer-Sezgin, Münevver; Department of Mathematics (2010) In this thesis, the two-dimensional initial and boundary value problems governed by unsteady partial differential equations are solved by making use of boundary element techniques. The boundary element method (BEM) with time-dependent fundamental solution is presented as an efficient procedure for the solution of diffusion, wave and convection-diffusion equations. It interpenetrates the equations in such a way that the boundary solution is advanced to all time levels, simultaneously. The solution at a requi...
Construction of substitution boxes depending on linear block codes Yıldız, Senay; Yücel, Melek D; Department of Cryptography (2004) The construction of a substitution box (S-box) with high nonlinearity and high resiliency is an important research area in cryptography. In this thesis, t-resilient nxm S-box construction methods depending on linear block codes presented in "A Construction of Resilient Functions with High Nonlinearity" by T. Johansson and E. Pasalic in 2000, and two years later in "Linear Codes in Generalized Construction of Resilient Functions with Very High Nonlinearity" by E. Pasalic and S. Maitra are compared and the fo...
Predictive models in ecology: Comparison of performances and assessment of applicability Tan, Can Ozan; Ozesmi, Uygar; Beklioğlu, Meryem; Per, Esra; Kurt, Bahtiyar (Elsevier BV, 2006-04-01) Ecological systems are governed by complex interactions which are mainly nonlinear. In order to capture the inherent complexity and nonlinearity of ecological, and in general biological systems, empirical models recently gained popularity. However, although these models, particularly connectionist approaches such as multilayered backpropagation networks, are commonly applied as predictive models in ecology to a wide variety of ecosystems and questions, there are no studies to date aiming to assess the perfo...

Citation Formats

S. Gürol, “Statistical learning and optimization methods for improving the efficiency in landscape image clustering and classification problems,” M.S. - Master of Science, Middle East Technical University, 2005.