Probabilistic Neural Networks for Segmentation of Features in Corn Kernel Images

dc.contributor.author Steenhoek, Loren
dc.contributor.author Misra, Manjit
dc.contributor.author Misra, Manjit
dc.contributor.author Batchelor, William
dc.contributor.author Davidson, Jennifer
dc.contributor.department Agricultural and Biosystems Engineering
dc.date 2018-02-14T13:40:40.000
dc.date.accessioned 2020-06-29T22:40:42Z
dc.date.available 2020-06-29T22:40:42Z
dc.date.copyright Mon Jan 01 00:00:00 UTC 2001
dc.date.embargo 2014-09-08
dc.date.issued 2001-01-01
dc.description.abstract <p>A method is presented for clustering of pixel color information to segment features within corn kernel images. Features for blue–eye mold, germ damage, sound germ, shadow in sound germ, hard starch, and soft starch were identified by red, green, and blue (RGB) pixel value inputs to a probabilistic neural network. A data grouping method to obtain an exemplar set for adjustment of the Probabilistic Neural Network (PNN) weights and optimization of a universal smoothing factor is described. Of the 14,427 available exemplars (RGB pixel values sampled from previously collected images), 778 were used for adjustment of the network weights, 737 were used for optimization of the PNN smoothing parameter, and 12,912 were reserved for network validation. Based on a universal PNN smoothing factor of 0.05, the network was able to provide an overall pixel classification accuracy of 86% on calibration data and 75% on unseen data. Much of the misclassification was due to overlap of pixel values among classes. When an additional network layer was added to combine similar classes (blue–eye mold and germ damage, sound germ and shadow in sound germ, and hard and soft starch), network results were significantly enhanced so that accuracy on validation data was 94.7%. Image quality was shown to be important to the success of this algorithm as lighting and camera depth of field effects caused artifacts in the segmented images.</p>
dc.description.comments <p>This article is from <em>Applied Engineering in Agriculture</em> 17 (2001): 225–234, doi:<a href="http://dx.doi.org/10.13031/2013.5447" target="_blank">10.13031/2013.5447</a>. Posted with permission.</p>
dc.format.mimetype application/pdf
dc.identifier archive/lib.dr.iastate.edu/abe_eng_pubs/543/
dc.identifier.articleid 1827
dc.identifier.contextkey 6092570
dc.identifier.s3bucket isulib-bepress-aws-west
dc.identifier.submissionpath abe_eng_pubs/543
dc.identifier.uri https://dr.lib.iastate.edu/handle/20.500.12876/1321
dc.language.iso en
dc.source.bitstream archive/lib.dr.iastate.edu/abe_eng_pubs/543/2001_Steenhoek_ProbabilisticNeural.pdf|||Sat Jan 15 00:53:51 UTC 2022
dc.source.uri 10.13031/2013.5447
dc.subject.disciplines Agriculture
dc.subject.disciplines Bioresource and Agricultural Engineering
dc.subject.disciplines Electrical and Computer Engineering
dc.subject.keywords Electrical and Computer Engineering
dc.subject.keywords Grain damage
dc.subject.keywords Evaluation
dc.subject.keywords Machine vision
dc.subject.keywords Probabilistic
dc.subject.keywords Neural network
dc.subject.keywords Corn
dc.subject.keywords Color
dc.subject.keywords Pattern recognition
dc.title Probabilistic Neural Networks for Segmentation of Features in Corn Kernel Images
dc.type article
dc.type.genre article
dspace.entity.type Publication
relation.isAuthorOfPublication a1d94dd4-52b8-4e60-b8a8-e742eefa9a59
relation.isOrgUnitOfPublication 8eb24241-0d92-4baf-ae75-08f716d30801
File
Original bundle
Now showing 1 - 1 of 1
Name:
2001_Steenhoek_ProbabilisticNeural.pdf
Size:
272.41 KB
Format:
Adobe Portable Document Format
Description:
Collections