Virginia Tech
Application of Machine Learning and Hyperspectral Imaging in Plant Phenomics Research
Abstract
dc:description.abstractDigital imaging technology, spatial analyses tools, and computer vision methods can be used to extract traits related to branching pattern, canopy cover, and pod location in edamame in a high throughput fashion. Using genome-wide association study, we identified several single nucleotide polymorphisms (SNPs) that were associated with those traits, which can be used in marker-assisted selection to further develop edamame varieties that are better adapted to mechanical harvesting and can produce higher yield, along with understanding the physiological mechanisms for better shoot architecture traits for better yield. Edamame has a short, optimal harvest window of only about one week. To determine harvesting time, experienced growers depend on touch, taste, and observation, which is low-throughput. We tested spectroscopy-based machine learning method to identify the optimal harvest time of edamame. To further explore the power of spectral imaging, we tested a hyperspectral imaging method to quantify Deoxynivalenol (DON) in small grains. Currently gas chromatography and mass spectroscopy (GC-MS) have been used to quantify the DON levels in small grains, which is lab-based and costly low throughput technique. We found that hyperspectral imaging combined with computer vision and machine learning methods can be used to quantify the levels of DON in wheat kernels in a high through put fashion.
Degree
thesis:*- Name thesis:degree_name
- Doctor of Philosophy
- Level thesis:degree_level
- doctoral
- Discipline thesis:degree_discipline
- Crop and Soil Environmental Sciences
- Department dc:contributor.department
- Crop and Soil Environmental Sciences
- Grantor dc:publisher
- Virginia Tech
- Year dc:date.issued
- 2023
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Dhakal, Kshitiz
- Chair dc:contributor.committeechair
-
- Li, Song
- Committee members dc:contributor.committeemember
-
- Zhao, Bingyu
- Zhang, Bo
- Oakes, Joseph Carroll
- Morota, Gota
Subjects
dc:subject × 7Rights
dc:rights- Statement dc:rights
-
- In Copyright
- Licence dc:rights.uri
- Language dc:language.iso
- en
Identifiers
dc:identifier.*- Dc Identifier Other
- vt_gsexam:36542