東華大學圖書館 |

語系: 繁體中文

說明(常見問題)

回圖書館首頁

手機版館藏查詢

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

FindBook

Google Book

Amazon

博客來

Spectral analysis and multispectral/hyperspectral imaging to detect blueberry fruit maturity stages for early blueberry yield estimation.

紀錄類型:	書目-電子資源 : Monograph/item
正題名/作者:	Spectral analysis and multispectral/hyperspectral imaging to detect blueberry fruit maturity stages for early blueberry yield estimation./
作者:	Yang, Ce.
面頁冊數:	1 online resource (135 pages)
附註:	Source: Dissertations Abstracts International, Volume: 76-01, Section: B.
Contained By:	Dissertations Abstracts International76-01B.
標題:	Agronomy. -
電子資源:	http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3583006click for full text (PQDT)
ISBN:	9781321048339

Spectral analysis and multispectral/hyperspectral imaging to detect blueberry fruit maturity stages for early blueberry yield estimation.
Yang, Ce.

Spectral analysis and multispectral/hyperspectral imaging to detect blueberry fruit maturity stages for early blueberry yield estimation. - 1 online resource (135 pages)

Source: Dissertations Abstracts International, Volume: 76-01, Section: B.

Thesis (Ph.D.)--University of Florida, 2013.

Includes bibliographical references

Blueberry industry has been increasingly important to both Florida and United States economically since 1990's (USDA, 2012). Because of the mild sub-tropical climate, blueberry harvesting window in Florida is uniquely early, yielding high profits in the fresh market. However, it is relatively short, usually lasting only five to six weeks. After that, the blueberry price drops rapidly. Therefore, early estimation of fruit yield is crucial for the market and for labor planning. This dissertation explores methods for detection of blueberry with all maturity stages by their spectral properties as well as spatial information. Spectral analysis offers necessary wavelengths for blueberry detection. Spectra of blueberry fruit and leaf samples were obtained and analyzed. The samples were divided into leaf, mature fruit, near-mature fruit, near-young fruit and young fruit. Normalized indices were used as the candidate variables for classification. Classification models were built and their performances were compared. Four to six wavelengths were chosen using different methods and accuracies of more than 94% were obtained for the classification task. However, a spectrophotometer is very expensive and can only be used in a laboratory. In contrast, computer vision enables in-field data acquisition. In 2011, multispectral images with three bands: near infrared (760 - 900 nm), red (630 - 690 nm) and green (520 - 600 nm) were obtained. Different color components were input features for classification. Accuracies of 84% and 73% were obtained for fruit and background classes, respectively. However, the color features did poorly in separating eight classes: mature fruit, intermediate fruit, young fruit, leaf, branch, soil, sky, and reference board. Hyperspectral imaging was proved to be more capable of detecting visually similar object. Therefore, hyperspectral images were acquired in 2012 and 2013. Band selection was necessary to find the most important bands for further application in the field. Three sets of bands were selected using three band selection methods and obtained prediction accuracies of more than 88%. The results showed that the selected band sets were capable of classifying blueberry maturity stages and background. It is beneficial to use spatial information upon the spectral properties of objects in the view. Therefore, spectral-spatial image analysis was considered for the detection of fruits with different maturity stages. Two spectral-spatial image analysis procedures were carried out and evaluated based on the labeled images and obtained more than 78% pixel detection accuracy. The spectral-spatial detection improved the prediction accuracy by up to 30% compared to spectral detection.

Electronic reproduction.
Ann Arbor, Mich. :
ProQuest,
2023

Mode of access: World Wide Web

ISBN: 9781321048339Subjects--Topical Terms:

2122783
Agronomy.
Subjects--Index Terms:

Blueberry industryIndex Terms--Genre/Form:

542853
Electronic books.

Spectral analysis and multispectral/hyperspectral imaging to detect blueberry fruit maturity stages for early blueberry yield estimation.
LDR:04159nmm a2200385K 4500 001 2360580
005 20230928115717.5
006 m o d
007 cr mn ---uuuuu
008 241011s2013 xx obm 000 0 eng d
020 $a 9781321048339
035 $a (MiAaPQ)AAI3583006
035 $a AAI3583006
040 $a MiAaPQ $b eng $c MiAaPQ $d NTU
100 1 $a Yang, Ce. $3 3701201
245 1 0 $a Spectral analysis and multispectral/hyperspectral imaging to detect blueberry fruit maturity stages for early blueberry yield estimation.
264 0 $c 2013
300 $a 1 online resource (135 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 76-01, Section: B.
500 $a Publisher info.: Dissertation/Thesis.
500 $a Advisor: Lee, Won Suk 'Daniel'.
502 $a Thesis (Ph.D.)--University of Florida, 2013.
504 $a Includes bibliographical references
520 $a Blueberry industry has been increasingly important to both Florida and United States economically since 1990's (USDA, 2012). Because of the mild sub-tropical climate, blueberry harvesting window in Florida is uniquely early, yielding high profits in the fresh market. However, it is relatively short, usually lasting only five to six weeks. After that, the blueberry price drops rapidly. Therefore, early estimation of fruit yield is crucial for the market and for labor planning. This dissertation explores methods for detection of blueberry with all maturity stages by their spectral properties as well as spatial information. Spectral analysis offers necessary wavelengths for blueberry detection. Spectra of blueberry fruit and leaf samples were obtained and analyzed. The samples were divided into leaf, mature fruit, near-mature fruit, near-young fruit and young fruit. Normalized indices were used as the candidate variables for classification. Classification models were built and their performances were compared. Four to six wavelengths were chosen using different methods and accuracies of more than 94% were obtained for the classification task. However, a spectrophotometer is very expensive and can only be used in a laboratory. In contrast, computer vision enables in-field data acquisition. In 2011, multispectral images with three bands: near infrared (760 - 900 nm), red (630 - 690 nm) and green (520 - 600 nm) were obtained. Different color components were input features for classification. Accuracies of 84% and 73% were obtained for fruit and background classes, respectively. However, the color features did poorly in separating eight classes: mature fruit, intermediate fruit, young fruit, leaf, branch, soil, sky, and reference board. Hyperspectral imaging was proved to be more capable of detecting visually similar object. Therefore, hyperspectral images were acquired in 2012 and 2013. Band selection was necessary to find the most important bands for further application in the field. Three sets of bands were selected using three band selection methods and obtained prediction accuracies of more than 88%. The results showed that the selected band sets were capable of classifying blueberry maturity stages and background. It is beneficial to use spatial information upon the spectral properties of objects in the view. Therefore, spectral-spatial image analysis was considered for the detection of fruits with different maturity stages. Two spectral-spatial image analysis procedures were carried out and evaluated based on the labeled images and obtained more than 78% pixel detection accuracy. The spectral-spatial detection improved the prediction accuracy by up to 30% compared to spectral detection.
533 $a Electronic reproduction. $b Ann Arbor, Mich. : $c ProQuest, $d 2023
538 $a Mode of access: World Wide Web
650 4 $a Agronomy. $3 2122783
650 4 $a Agriculture. $3 518588
650 4 $a Engineering. $3 586835
653 $a Blueberry industry
653 $a Florida
653 $a Precision agriculture
655 7 $a Electronic books. $2 lcsh $3 542853
690 $a 0285
690 $a 0473
690 $a 0537
710 2 $a ProQuest Information and Learning Co. $3 783688
710 2 $a University of Florida. $b Agricultural and BIological Engineering. $3 3701202
773 0 $t Dissertations Abstracts International $g 76-01B.
856 4 0 $u http://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=3583006 $z click for full text (PQDT)