Millimeter-Level Plant Disease Detection From Aerial Photographs via Deep Learning and Crowdsourced Data
Computer vision models that can recognize plant diseases in the field would be valuable tools for disease management and resistance breeding. Generating enough data to train these models is difficult, however, since only trained experts can accurately identify symptoms. In this study, we describe an...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2019-12-01
|
| Series: | Frontiers in Plant Science |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/article/10.3389/fpls.2019.01550/full |
| _version_ | 1818485261665828864 |
|---|---|
| author | Tyr Wiesner-Hanks Harvey Wu Ethan Stewart Chad DeChant Nicholas Kaczmar Hod Lipson Michael A. Gore Rebecca J. Nelson Rebecca J. Nelson |
| author_facet | Tyr Wiesner-Hanks Harvey Wu Ethan Stewart Chad DeChant Nicholas Kaczmar Hod Lipson Michael A. Gore Rebecca J. Nelson Rebecca J. Nelson |
| author_sort | Tyr Wiesner-Hanks |
| collection | DOAJ |
| description | Computer vision models that can recognize plant diseases in the field would be valuable tools for disease management and resistance breeding. Generating enough data to train these models is difficult, however, since only trained experts can accurately identify symptoms. In this study, we describe and implement a two-step method for generating a large amount of high-quality training data with minimal expert input. First, experts located symptoms of northern leaf blight (NLB) in field images taken by unmanned aerial vehicles (UAVs), annotating them quickly at low resolution. Second, non-experts were asked to draw polygons around the identified diseased areas, producing high-resolution ground truths that were automatically screened based on agreement between multiple workers. We then used these crowdsourced data to train a convolutional neural network (CNN), feeding the output into a conditional random field (CRF) to segment images into lesion and non-lesion regions with accuracy of 0.9979 and F1 score of 0.7153. The CNN trained on crowdsourced data showed greatly improved spatial resolution compared to one trained on expert-generated data, despite using only one fifth as many expert annotations. The final model was able to accurately delineate lesions down to the millimeter level from UAV-collected images, the finest scale of aerial plant disease detection achieved to date. The two-step approach to generating training data is a promising method to streamline deep learning approaches for plant disease detection, and for complex plant phenotyping tasks in general. |
| first_indexed | 2024-12-10T16:06:19Z |
| format | Article |
| id | doaj.art-532813fb826a425dbb29e3eb3e252ba7 |
| institution | Directory Open Access Journal |
| issn | 1664-462X |
| language | English |
| last_indexed | 2024-12-10T16:06:19Z |
| publishDate | 2019-12-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Plant Science |
| spelling | doaj.art-532813fb826a425dbb29e3eb3e252ba72022-12-22T01:42:13ZengFrontiers Media S.A.Frontiers in Plant Science1664-462X2019-12-011010.3389/fpls.2019.01550481893Millimeter-Level Plant Disease Detection From Aerial Photographs via Deep Learning and Crowdsourced DataTyr Wiesner-Hanks0Harvey Wu1Ethan Stewart2Chad DeChant3Nicholas Kaczmar4Hod Lipson5Michael A. Gore6Rebecca J. Nelson7Rebecca J. Nelson8Plant Breeding and Genetics Section, School of Integrative Plant Science, Cornell University, Ithaca, NY, United StatesDepartment of Computer Science, Columbia University, New York, NY, United StatesPlant Breeding and Genetics Section, School of Integrative Plant Science, Cornell University, Ithaca, NY, United StatesDepartment of Computer Science, Columbia University, New York, NY, United StatesPlant Breeding and Genetics Section, School of Integrative Plant Science, Cornell University, Ithaca, NY, United StatesDepartment of Mechanical Engineering and Institute of Data Science, Columbia University, New York, NY, United StatesPlant Breeding and Genetics Section, School of Integrative Plant Science, Cornell University, Ithaca, NY, United StatesPlant Breeding and Genetics Section, School of Integrative Plant Science, Cornell University, Ithaca, NY, United StatesPlant Pathology and Plant-Microbe Biology Section, School of Integrative Plant Science, Cornell University, Ithaca, NY, United StatesComputer vision models that can recognize plant diseases in the field would be valuable tools for disease management and resistance breeding. Generating enough data to train these models is difficult, however, since only trained experts can accurately identify symptoms. In this study, we describe and implement a two-step method for generating a large amount of high-quality training data with minimal expert input. First, experts located symptoms of northern leaf blight (NLB) in field images taken by unmanned aerial vehicles (UAVs), annotating them quickly at low resolution. Second, non-experts were asked to draw polygons around the identified diseased areas, producing high-resolution ground truths that were automatically screened based on agreement between multiple workers. We then used these crowdsourced data to train a convolutional neural network (CNN), feeding the output into a conditional random field (CRF) to segment images into lesion and non-lesion regions with accuracy of 0.9979 and F1 score of 0.7153. The CNN trained on crowdsourced data showed greatly improved spatial resolution compared to one trained on expert-generated data, despite using only one fifth as many expert annotations. The final model was able to accurately delineate lesions down to the millimeter level from UAV-collected images, the finest scale of aerial plant disease detection achieved to date. The two-step approach to generating training data is a promising method to streamline deep learning approaches for plant disease detection, and for complex plant phenotyping tasks in general.https://www.frontiersin.org/article/10.3389/fpls.2019.01550/fullphenotypingunmanned aerial vehiclesplant diseasedeep learningmachine learningcrowdsourcing |
| spellingShingle | Tyr Wiesner-Hanks Harvey Wu Ethan Stewart Chad DeChant Nicholas Kaczmar Hod Lipson Michael A. Gore Rebecca J. Nelson Rebecca J. Nelson Millimeter-Level Plant Disease Detection From Aerial Photographs via Deep Learning and Crowdsourced Data Frontiers in Plant Science phenotyping unmanned aerial vehicles plant disease deep learning machine learning crowdsourcing |
| title | Millimeter-Level Plant Disease Detection From Aerial Photographs via Deep Learning and Crowdsourced Data |
| title_full | Millimeter-Level Plant Disease Detection From Aerial Photographs via Deep Learning and Crowdsourced Data |
| title_fullStr | Millimeter-Level Plant Disease Detection From Aerial Photographs via Deep Learning and Crowdsourced Data |
| title_full_unstemmed | Millimeter-Level Plant Disease Detection From Aerial Photographs via Deep Learning and Crowdsourced Data |
| title_short | Millimeter-Level Plant Disease Detection From Aerial Photographs via Deep Learning and Crowdsourced Data |
| title_sort | millimeter level plant disease detection from aerial photographs via deep learning and crowdsourced data |
| topic | phenotyping unmanned aerial vehicles plant disease deep learning machine learning crowdsourcing |
| url | https://www.frontiersin.org/article/10.3389/fpls.2019.01550/full |
| work_keys_str_mv | AT tyrwiesnerhanks millimeterlevelplantdiseasedetectionfromaerialphotographsviadeeplearningandcrowdsourceddata AT harveywu millimeterlevelplantdiseasedetectionfromaerialphotographsviadeeplearningandcrowdsourceddata AT ethanstewart millimeterlevelplantdiseasedetectionfromaerialphotographsviadeeplearningandcrowdsourceddata AT chaddechant millimeterlevelplantdiseasedetectionfromaerialphotographsviadeeplearningandcrowdsourceddata AT nicholaskaczmar millimeterlevelplantdiseasedetectionfromaerialphotographsviadeeplearningandcrowdsourceddata AT hodlipson millimeterlevelplantdiseasedetectionfromaerialphotographsviadeeplearningandcrowdsourceddata AT michaelagore millimeterlevelplantdiseasedetectionfromaerialphotographsviadeeplearningandcrowdsourceddata AT rebeccajnelson millimeterlevelplantdiseasedetectionfromaerialphotographsviadeeplearningandcrowdsourceddata AT rebeccajnelson millimeterlevelplantdiseasedetectionfromaerialphotographsviadeeplearningandcrowdsourceddata |