Analysis of hyperspectral scattering images using locally linear embedding algorithm for apple mealiness classification

► We adopt hyperspectral scattering image to assess apple mealiness. ► LLE algorithm is developed to extract hyperspectral image features. ► SVM coupled with LLE algorithm is effective for detecting apple mealiness. ► Classification results by LLE are better than that by mean-LLE and mean methods. H...

Celý popis

Uloženo v:

Podrobná bibliografie
Vydáno v:	Computers and electronics in agriculture Ročník 89; s. 175 - 181
Hlavní autoři:	Huang, Min, Zhu, Qibing, Wang, Bojin, Lu, Renfu
Médium:	Journal Article
Jazyk:	angličtina
Vydáno:	Amsterdam Elsevier B.V 01.11.2012 Elsevier
Témata:	Agronomy. Soil science and plant productions algorithms Apple apples Biological and medical sciences discriminant analysis Fundamental and applied biological sciences. Psychology hyperspectral imagery Hyperspectral scattering least squares Locally linear embedding Mealiness Partial least squares discriminant Support vector machine Hyperspectral scattering Mealiness Apple Support vector machine Partial least squares discriminant Locally linear embedding Hyperspectral imagery Organoleptic properties Hyperspectral characteristic Classification Algorithm Partial least squares
ISSN:	0168-1699, 1872-7107
On-line přístup:	Získat plný text
Tagy:	Přidat tag Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!

Popis
Shrnutí:	► We adopt hyperspectral scattering image to assess apple mealiness. ► LLE algorithm is developed to extract hyperspectral image features. ► SVM coupled with LLE algorithm is effective for detecting apple mealiness. ► Classification results by LLE are better than that by mean-LLE and mean methods. Hyperspectral scattering images between 600nm and 1000nm were acquired for 580 ‘Delicious’ apples for mealiness classification. A locally linear embedding (LLE) algorithm was developed to extract features directly from the hyperspectral scattering image data. Partial least squares discriminant analysis (PLSDA) and support vector machine (SVM) were applied to develop classification models based on the LLE, mean-LLE and mean spectra algorithms. The model based on the LLE algorithm achieved an overall classification accuracy of 80.4%, compared with 76.2% by the mean-LLE algorithm and 73.0% by the mean spectra method for two-class classification (i.e., mealy and nonmealy) coupled with PLSDA. For the SVM models, the LLE algorithm had an overall classification accuracy of 82.5%, compared with 79.4% by the mean-LLE algorithm and 78.3% by the mean spectra method. Hence, the LLE algorithm provided an effective means to extract hyperspectral scattering features for mealiness classification.
Bibliografie:	http://dx.doi.org/10.1016/j.compag.2012.09.003 ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 23
ISSN:	0168-1699 1872-7107
DOI:	10.1016/j.compag.2012.09.003