東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Classification and detection in mult...

Zavaljevski, Aleksandar.

Linked to FindBook

Google Book

Amazon

博客來

Classification and detection in multispectral images.

Record Type:	Language materials, printed : Monograph/item
Title/Author:	Classification and detection in multispectral images./
Author:	Zavaljevski, Aleksandar.
Description:	242 p.
Notes:	Chair: Atam P. Dhawan.
Contained By:	Dissertation Abstracts International58-01B.
Subject:	Engineering, Biomedical. -
Online resource:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=9718248
ISBN:	0591262967

Classification and detection in multispectral images.
Zavaljevski, Aleksandar.

Classification and detection in multispectral images. - 242 p.

Chair: Atam P. Dhawan.

Thesis (Ph.D.)--University of Cincinnati, 1996.

A novel multi-level adaptive pixel classification and target detection (MLACD) method for multispectral images is presented in this dissertation. The MLACD method takes into account both spectral and spatial characteristics of the data by dealing with it on the global, neighborhood and pixel/subpixel levels. The analysis on each of the levels is complex and includes adaptation loops.

ISBN: 0591262967Subjects--Topical Terms:

1017684
Engineering, Biomedical.

Classification and detection in multispectral images.
LDR:03110nam 2200325 a 45 001 926619
005 20110422
008 110422s1996 eng d
020 $a 0591262967
035 $a (UnM)AAI9718248
035 $a AAI9718248
040 $a UnM $c UnM
100 1 $a Zavaljevski, Aleksandar. $3 1250204
245 1 0 $a Classification and detection in multispectral images.
300 $a 242 p.
500 $a Chair: Atam P. Dhawan.
500 $a Source: Dissertation Abstracts International, Volume: 58-01, Section: B, page: 0359.
502 $a Thesis (Ph.D.)--University of Cincinnati, 1996.
520 $a A novel multi-level adaptive pixel classification and target detection (MLACD) method for multispectral images is presented in this dissertation. The MLACD method takes into account both spectral and spatial characteristics of the data by dealing with it on the global, neighborhood and pixel/subpixel levels. The analysis on each of the levels is complex and includes adaptation loops.
520 $a The general MLACD framework is demonstrated on several specific imaging tasks. The first is target detection on small targets of pixel and subpixel size. Second, a new method is developed for lines of communication extraction. Third, a new model for multispectral magnetic resonance (MR) brain images is proposed, as well as a procedure for model parameters estimation. Fourth, an MR brain images classification scheme is proposed as well as a procedure for elastic transformation of the brain slices. Fifth, a system for manual MR brain images segmentation is developed that uses the Sun SPARC workstation and PIXAR image computer. Finally, a new multi-level method for evaluation of the perfusion images is proposed.
520 $a MLACD methods are applied in two areas: remote sensing and medical imaging. In remote sensing, procedures are applied to the hyperspectral Airborne Visible/Infrared Imaging System (AVIRIS) data for pixel classification and small target detection. Sensitivity analysis and comparative evaluation of the proposed detector are also presented. These methods are also used for extraction of the lines of communication in AVIRIS data. It is shown that the proposed detector has improved receiver operating characteristics compared to conventional methods.
520 $a The system for manual segmentation of MR brain images is applied to several normal and pathological human and pig brains. Perfusion images are evaluated using both simulated and clinical data. Proposed methods are then applied to classification of several human brains as well as pig brains. A synthetic display that combines information of brain classification and perfusion is implemented. Experimental results confirm that it is possible to perform the classification using extended set of tissue classes with acceptable accuracy.
590 $a School code: 0045.
650 4 $a Engineering, Biomedical. $3 1017684
650 4 $a Engineering, Electronics and Electrical. $3 626636
650 4 $a Health Sciences, Radiology. $3 1019076
690 $a 0541
690 $a 0544
690 $a 0574
710 2 0 $a University of Cincinnati. $3 960309
773 0 $t Dissertation Abstracts International $g 58-01B.
790 $a 0045
790 1 0 $a Dhawan, Atam P., $e advisor
791 $a Ph.D.
792 $a 1996
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=9718248