The Automation of Hyperspectral Training Library Construction: A Case Study for Wheat and Potato Crops
The potential of hyperspectral measurements for early disease detection has been investigated by many experts over the last 5 years. One of the difficulties is obtaining enough data for training and building a hyperspectral training library. When the goal is to detect disease at a previsible stage,...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-11-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/13/23/4735 |
| _version_ | 1797507285146140672 |
|---|---|
| author | Simon Appeltans Orly Enrique Apolo-Apolo Jaime Nolasco Rodríguez-Vázquez Manuel Pérez-Ruiz Jan Pieters Abdul M. Mouazen |
| author_facet | Simon Appeltans Orly Enrique Apolo-Apolo Jaime Nolasco Rodríguez-Vázquez Manuel Pérez-Ruiz Jan Pieters Abdul M. Mouazen |
| author_sort | Simon Appeltans |
| collection | DOAJ |
| description | The potential of hyperspectral measurements for early disease detection has been investigated by many experts over the last 5 years. One of the difficulties is obtaining enough data for training and building a hyperspectral training library. When the goal is to detect disease at a previsible stage, before the pathogen has manifested either its first symptoms or in the area surrounding the existing symptoms, it is impossible to objectively delineate the regions of interest containing the previsible pathogen growth from the areas without the pathogen growth. To overcome this, we propose an image labelling and segmentation algorithm that is able to (a) more objectively label the visible symptoms for the construction of a training library and (b) extend this labelling to the pre-visible symptoms. This algorithm is used to create hyperspectral training libraries for late blight disease (<i>Phytophthora infestans</i>) in potatoes and two types of leaf rust (<i>Puccinia triticina</i> and <i>Puccinia striiformis</i>) in wheat. The model training accuracies were compared between the automatic labelling algorithm and the classic visual delineation of regions of interest using a logistic regression machine learning approach. The modelling accuracies of the automatically labelled datasets were higher than those of the manually labelled ones for both potatoes and wheat, at 98.80% for <i>P. infestans</i> in potato, 97.69% for <i>P. striiformis</i> in soft wheat, and 96.66% for <i>P. triticina</i> in durum wheat. |
| first_indexed | 2024-03-10T04:46:23Z |
| format | Article |
| id | doaj.art-5d5c062016b84dc09f853972d0063176 |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-10T04:46:23Z |
| publishDate | 2021-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-5d5c062016b84dc09f853972d00631762023-11-23T02:55:41ZengMDPI AGRemote Sensing2072-42922021-11-011323473510.3390/rs13234735The Automation of Hyperspectral Training Library Construction: A Case Study for Wheat and Potato CropsSimon Appeltans0Orly Enrique Apolo-Apolo1Jaime Nolasco Rodríguez-Vázquez2Manuel Pérez-Ruiz3Jan Pieters4Abdul M. Mouazen5Department of Environment, Faculty of Bioscience Engineering, Ghent University, 9000 Ghent, BelgiumDepartamento de Ingeniería Aeroespacial y Mecánica de Fluidos Área Agroforestal, University of Sevilla, 41013 Sevilla, SpainDepartamento de Ingeniería Aeroespacial y Mecánica de Fluidos Área Agroforestal, University of Sevilla, 41013 Sevilla, SpainDepartamento de Ingeniería Aeroespacial y Mecánica de Fluidos Área Agroforestal, University of Sevilla, 41013 Sevilla, SpainDepartment of Plants and Crops, Faculty of Bioscience Engineering, Ghent University, 9000 Ghent, BelgiumDepartment of Environment, Faculty of Bioscience Engineering, Ghent University, 9000 Ghent, BelgiumThe potential of hyperspectral measurements for early disease detection has been investigated by many experts over the last 5 years. One of the difficulties is obtaining enough data for training and building a hyperspectral training library. When the goal is to detect disease at a previsible stage, before the pathogen has manifested either its first symptoms or in the area surrounding the existing symptoms, it is impossible to objectively delineate the regions of interest containing the previsible pathogen growth from the areas without the pathogen growth. To overcome this, we propose an image labelling and segmentation algorithm that is able to (a) more objectively label the visible symptoms for the construction of a training library and (b) extend this labelling to the pre-visible symptoms. This algorithm is used to create hyperspectral training libraries for late blight disease (<i>Phytophthora infestans</i>) in potatoes and two types of leaf rust (<i>Puccinia triticina</i> and <i>Puccinia striiformis</i>) in wheat. The model training accuracies were compared between the automatic labelling algorithm and the classic visual delineation of regions of interest using a logistic regression machine learning approach. The modelling accuracies of the automatically labelled datasets were higher than those of the manually labelled ones for both potatoes and wheat, at 98.80% for <i>P. infestans</i> in potato, 97.69% for <i>P. striiformis</i> in soft wheat, and 96.66% for <i>P. triticina</i> in durum wheat.https://www.mdpi.com/2072-4292/13/23/4735hyperspectralwheatpotatomachine learninglabelling |
| spellingShingle | Simon Appeltans Orly Enrique Apolo-Apolo Jaime Nolasco Rodríguez-Vázquez Manuel Pérez-Ruiz Jan Pieters Abdul M. Mouazen The Automation of Hyperspectral Training Library Construction: A Case Study for Wheat and Potato Crops Remote Sensing hyperspectral wheat potato machine learning labelling |
| title | The Automation of Hyperspectral Training Library Construction: A Case Study for Wheat and Potato Crops |
| title_full | The Automation of Hyperspectral Training Library Construction: A Case Study for Wheat and Potato Crops |
| title_fullStr | The Automation of Hyperspectral Training Library Construction: A Case Study for Wheat and Potato Crops |
| title_full_unstemmed | The Automation of Hyperspectral Training Library Construction: A Case Study for Wheat and Potato Crops |
| title_short | The Automation of Hyperspectral Training Library Construction: A Case Study for Wheat and Potato Crops |
| title_sort | automation of hyperspectral training library construction a case study for wheat and potato crops |
| topic | hyperspectral wheat potato machine learning labelling |
| url | https://www.mdpi.com/2072-4292/13/23/4735 |
| work_keys_str_mv | AT simonappeltans theautomationofhyperspectraltraininglibraryconstructionacasestudyforwheatandpotatocrops AT orlyenriqueapoloapolo theautomationofhyperspectraltraininglibraryconstructionacasestudyforwheatandpotatocrops AT jaimenolascorodriguezvazquez theautomationofhyperspectraltraininglibraryconstructionacasestudyforwheatandpotatocrops AT manuelperezruiz theautomationofhyperspectraltraininglibraryconstructionacasestudyforwheatandpotatocrops AT janpieters theautomationofhyperspectraltraininglibraryconstructionacasestudyforwheatandpotatocrops AT abdulmmouazen theautomationofhyperspectraltraininglibraryconstructionacasestudyforwheatandpotatocrops AT simonappeltans automationofhyperspectraltraininglibraryconstructionacasestudyforwheatandpotatocrops AT orlyenriqueapoloapolo automationofhyperspectraltraininglibraryconstructionacasestudyforwheatandpotatocrops AT jaimenolascorodriguezvazquez automationofhyperspectraltraininglibraryconstructionacasestudyforwheatandpotatocrops AT manuelperezruiz automationofhyperspectraltraininglibraryconstructionacasestudyforwheatandpotatocrops AT janpieters automationofhyperspectraltraininglibraryconstructionacasestudyforwheatandpotatocrops AT abdulmmouazen automationofhyperspectraltraininglibraryconstructionacasestudyforwheatandpotatocrops |