FEATURE EXTRACTION FROM HYPERSPECTRAL IMAGERY FOR OBJECT RECOGNITION
| Main Author: | |
|---|---|
| Language: | English |
| Published: |
The Ohio State University / OhioLINK
2011
|
| Subjects: | |
| Online Access: | http://rave.ohiolink.edu/etdc/view?acc_num=osu1306848130 |
| id |
ndltd-OhioLink-oai-etd.ohiolink.edu-osu1306848130 |
|---|---|
| record_format |
oai_dc |
| spelling |
ndltd-OhioLink-oai-etd.ohiolink.edu-osu13068481302021-08-03T06:02:58Z FEATURE EXTRACTION FROM HYPERSPECTRAL IMAGERY FOR OBJECT RECOGNITION Yeu, Yeon Civil Engineering Hypespectral data Photogrammetry Remote Sensing <p>Over the past decade the most significant research in remote sensing has focused on the development of hyperspectral sensor systems and software to analyze images. The meaning of "hyper" in hyperspectral is "over" or "too many", so that "hyperspectral" refers to the large number of measured wavelength bands. The data provides a huge amount of spectral information for identifying and distinguishing spectrally unique materials to detect and map a wide variety of materials through characteristic reflectance spectra. Hyperspectral images have been proposed to help solving the object recognition problem, mainly due to their wide spectral range, which should provide clues about the material. Instead of approaching object recognition from a multi-sensor point of view, this dissertation solely focuses on hyperspectral images and—in contrast to traditional remote sensing approaches—tries to find dissimilarities in the image, rather than homogeneous regions.</p><p>This paper proposes feature extraction for object recognition by modifying the Fisher Linear Discriminant Analysis (LDA). It is a kind of classifier to minimize the variance matrix within classes and maximize the variance matrix between classes. The reference spectrum is obtained using the Fisher LDA to distinguish a chosen class from others. The dissimilarities among classes are maximized through the LDA process, enabling much clearer boundaries of objects. Features can then be hierarchically be extracted.</p> 2011-07-27 English text The Ohio State University / OhioLINK http://rave.ohiolink.edu/etdc/view?acc_num=osu1306848130 http://rave.ohiolink.edu/etdc/view?acc_num=osu1306848130 unrestricted This thesis or dissertation is protected by copyright: all rights reserved. It may not be copied or redistributed beyond the terms of applicable copyright laws. |
| collection |
NDLTD |
| language |
English |
| sources |
NDLTD |
| topic |
Civil Engineering Hypespectral data Photogrammetry Remote Sensing |
| spellingShingle |
Civil Engineering Hypespectral data Photogrammetry Remote Sensing Yeu, Yeon FEATURE EXTRACTION FROM HYPERSPECTRAL IMAGERY FOR OBJECT RECOGNITION |
| author |
Yeu, Yeon |
| author_facet |
Yeu, Yeon |
| author_sort |
Yeu, Yeon |
| title |
FEATURE EXTRACTION FROM HYPERSPECTRAL IMAGERY FOR OBJECT RECOGNITION |
| title_short |
FEATURE EXTRACTION FROM HYPERSPECTRAL IMAGERY FOR OBJECT RECOGNITION |
| title_full |
FEATURE EXTRACTION FROM HYPERSPECTRAL IMAGERY FOR OBJECT RECOGNITION |
| title_fullStr |
FEATURE EXTRACTION FROM HYPERSPECTRAL IMAGERY FOR OBJECT RECOGNITION |
| title_full_unstemmed |
FEATURE EXTRACTION FROM HYPERSPECTRAL IMAGERY FOR OBJECT RECOGNITION |
| title_sort |
feature extraction from hyperspectral imagery for object recognition |
| publisher |
The Ohio State University / OhioLINK |
| publishDate |
2011 |
| url |
http://rave.ohiolink.edu/etdc/view?acc_num=osu1306848130 |
| work_keys_str_mv |
AT yeuyeon featureextractionfromhyperspectralimageryforobjectrecognition |
| _version_ |
1719430019015507968 |