Classifier fusion methods adapted to template matching on satellite images

Download

index.pdf

Date

2013

Author

Ersöz, Ahmet

Metadata

Show full item record

Item Usage Stats

80
views

106
downloads

Classifier combination and selection methods are becoming popular in vision research. Classifier fusion studies started to take the place of continuous development of new algorithms. In this study, template matching methods are used as classifiers and the results of the template matching methods from satellite images are taken as input to the fusion center. Template matching methods are adapted to different classifier fusion methods. In literature, there is not any performance measurement standard for the binary template matching output images. In order to analyze and compare the template matching methods and the classifier fusion methods, two performance measurement methods are proposed. In one of them, pixel-by-pixel intersection of the output image and the ground truth image are considered. In the other method, the output image is considered as a set of objects and the intersection of the object in the output image and the ground truth image is analyzed. The individual performance of the template matching methods is highly dependent on the threshold values used in the methods. When combining the template matching results, choosing optimal threshold is important to analyze the performance of the classifier fusion method. There have been very few studies for optimizing the fusion system performance and the studies have only been presented on decision level fusion. As a final task, a method for optimizing the performance in score (raw output) level fusion system is proposed. The results are quite promising such that the outputs of the proposed method outperformed to most of the score level fusion methods in the literature.

Subject Keywords

Remote-sensing images., Image processing, Template matching (Digital image processing)., Computer algorithms

URI

http://etd.lib.metu.edu.tr/upload/12616276/index.pdf
https://hdl.handle.net/11511/23030

Collections

Graduate School of Natural and Applied Sciences, Thesis

Suggestions

OpenMETU
Core

A Shadow based trainable method for building detection in satellite images Dikmen, Mehmet; Halıcı, Uğur; Department of Geodetic and Geographical Information Technologies (2014) The purpose of this thesis is to develop a supervised building detection and extraction algorithm with a shadow based learning method for high-resolution satellite images. First, shadow segments are identified on an over-segmented image, and then neighboring shadow segments are merged by assuming that they are cast by a single building. Next, these shadow regions are used to detect the candidate regions where buildings most likely occur. Together with this information, distance to shadows towards illuminati...
Deep learning approach for laboratory mice grimace scaling Eral, Mustafa; Halıcı, Uğur; Department of Electrical and Electronics Engineering (2016) Deep learning is extremely attractive research topic in pattern recognition and machine learning areas. Applications in speech recognition, natural language processing, and machine vision fields gained huge acceleration in performance by employing deep learning. In this thesis, deep learning is used for medical purposes in order to scale pain degree of drug stimulated mice by examining facial grimace. For this purpose each frame in the videos in the training set were scaled manually by experts according to ...
Camera electronics and image enhancement software for infrared detector arrays Küçükkömürler, Alper; Akın, Tayfun; Department of Environmental Engineering (2012) This thesis aims to design and develop camera electronics and image enhancement software for infrared detector arrays. It first discusses the camera electronics suitable for infrared detector arrays, then it concentrates on image enhancement software that are implemented including defective pixel correction, contrast enhancement, noise reduction and pseudo coloring. After that, testing and results of the implemented algorithms were presented. Camera electronics and circuit operation frequency are selected c...
Localization and identification of structural nonlinearities using cascaded optimization and neural networks Koyuncu, A.; Ciğeroğlu, Ender; Özgüven, Hasan Nevzat (Elsevier BV, 2017-10-01) In this study, a new approach is proposed for identification of structural nonlinearities by employing cascaded optimization and neural networks. Linear finite element model of the system and frequency response functions measured at arbitrary locations of the system are used in this approach. Using the finite element model, a training data set is created, which appropriately spans the possible nonlinear configurations space of the system. A classification neural network trained on these data sets then local...
Digital output ROIC with single slope ADC for cooled infrared applications Akyurek, Fatih; Bayram, Barış (2017-04-01) The objective of this research is to develop an ADC stage integrated into ROIC which enables ROIC to have digital output. Digital output method isolates noise caused by outside mediums. At the system level, removal of the ADC proximity card reduces system complexity and volume of the IDDCA system which is important for avionic and missile applications. It also reduces the system cost associated with external ADC components. A digital output ROIC utilizing single slope ADC is fabricated using 0.18 A mu m CMO...

Citation Formats

A. Ersöz, “Classifier fusion methods adapted to template matching on satellite images,” M.S. - Master of Science, Middle East Technical University, 2013.