Show/Hide Menu
Hide/Show Apps
Logout
Türkçe
Türkçe
Search
Search
Login
Login
OpenMETU
OpenMETU
About
About
Open Science Policy
Open Science Policy
Open Access Guideline
Open Access Guideline
Postgraduate Thesis Guideline
Postgraduate Thesis Guideline
Communities & Collections
Communities & Collections
Help
Help
Frequently Asked Questions
Frequently Asked Questions
Guides
Guides
Thesis submission
Thesis submission
MS without thesis term project submission
MS without thesis term project submission
Publication submission with DOI
Publication submission with DOI
Publication submission
Publication submission
Supporting Information
Supporting Information
General Information
General Information
Copyright, Embargo and License
Copyright, Embargo and License
Contact us
Contact us
Hyperspectral imaging and machine learning of texture foods for classification
Download
index.pdf
Date
2011
Author
Ataş, Musa
Metadata
Show full item record
Item Usage Stats
317
views
103
downloads
Cite This
In this thesis the main objective is to design a machine vision system that classifies aflatoxin contaminated chili peppers from uncontaminated ones in a rapid and non-destructive manner via hyperspectral imaging and machine learning techniques. Hyperspectral image series of chili pepper samples collected from different regions of Turkey have been acquired under halogen and UV illuminations. A novel feature set based on quantized absolute difference of consecutive spectral band features is proposed. Spectral band energies along with absolute difference energies of the consecutive spectral bands are utilized as features and compared with other feature extraction methods such as Teager energy operator and 2D wavelet Linear Discriminant Bases (2D-LDB). For feature selection, Fisher discrimination power, information theoretic Minimum Redundancy Maximum Relevance (mRMR) method and proposed Multi Layer Perceptron (MLP) based feature selection schemes are utilized.Finally, Linear Discriminant Classifier (LDC), Support Vector Machines (SVM) and MLP are used as classifiers. It is observed that MLP outperforms other learning models in terms of predictor performance. We verified the performance and robustness of our proposed methods on different real world datasets. It is suggested that to achieve high classification accuracy and predictor robustness, a machine vision system with halogen excitation and quantized absolute difference of consecutive spectral band features should be utilized.
URI
http://etd.lib.metu.edu.tr/upload/12613883/index.pdf
https://hdl.handle.net/11511/20934
Collections
Graduate School of Informatics, Thesis
Suggestions
OpenMETU
Core
Dependence of n-cSi/MoOx Heterojunction Performance on cSi Doping Concentration
Nasser, Hisham; Kokbudak, Gamze; Mehmood, Haris; Turan, Raşit (Elsevier BV; 2017-04-05)
In this work, we demonstrate a strong correlation between crystalline silicon (cSi) base doping concentration and the performance of cSi/MoOx heterojunction solar cell by investigating the structure numerically based on Silvaco TCAD simulation tool and experimentally. The doping concentration of n-type cSi was scanned in the 1 x 10(15) - 2 x 10(16) cm(-3) range. Simulation results show that utilizing highly doped cSi wafer degrades the conversion efficiency of cSi/MoOx solar cell. Efficiency of 11.16% has b...
Stochastic simulation of large biochemical systems by approximate Gillespie algorithm
Purutçuoğlu Gazi, Vilda (IEEE; 2010-06-01)
In recent years the fast innovations for the development of new methods and algorithms in system biology enable the biologists to analyze and interpret the complex biochemical structures. One of the speedy development has been seen in mathematical methods for generating these complex systems on the computer. These techniques help the researcher to ask biologically interesting questions and test their expectations before starting their biological experiments. There are a number of methods which can approxima...
HYPERSPECTRAL CLASSIFICATION USING STACKED AUTOENCODERS WITH DEEP LEARNING
Özdemir, Ataman; Cetin, C. Yasemin Yardimci (2014-06-27)
In this study, stacked autoencoders which are widely utilized in deep learning research are applied to remote sensing domain for hyperspectral classification. High dimensional hyperspectral data is an excellent candidate for deep learning methods. However, there are no works in literature that focuses on such deep learning approaches for hyperspectral imagery. This study aims to fill this gap by utilizing stacked autoencoders. Experiments are conducted on the Pavia University scene. Using stacked autoencode...
Compressive Sensing Imaging with a Graphene Modulator at THz Frequency in Transmission Mode
Özkan, Vedat Ali; Takan, Taylan; Kakenov, N.; KOCABAŞ, COŞKUN; Altan, Hakan (2016-09-30)
In this study we demonstrate compressive sensing imaging with a unique graphene based optoelectronic device which allows us to modulate the THz field through an array of columns or rows distributed throughout its face.
Scaffolding strategies in electronic performance support systems: types and challenges
Çağıltay, Kürşat (2006-02-01)
In the study described in this paper, the major components of an electronic performance support system are described and the use of scaffolding techniques within such electronic environments is explored. Four different types of scaffolding are discussed: conceptual (supportive) scaffolding, metacognitive (reflective) scaffolding, procedural scaffolding and strategic-intrinsic scaffolding. The paper also discusses some of the challenges that developers face when they plan to include scaffolding techniques as...
Citation Formats
IEEE
ACM
APA
CHICAGO
MLA
BibTeX
M. Ataş, “Hyperspectral imaging and machine learning of texture foods for classification,” Ph.D. - Doctoral Program, Middle East Technical University, 2011.