Assessment of Different Object Detectors for the Maturity Level Classification of Broccoli Crops Using UAV Imagery
Broccoli is an example of a high-value crop that requires delicate handling throughout the growing season and during its post-harvesting treatment. As broccoli heads can be easily damaged, they are still harvested by hand. Moreover, human scouting is required to initially identify the field segments...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-02-01
|
| Series: | Remote Sensing |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2072-4292/14/3/731 |
| _version_ | 1797484982258630656 |
|---|---|
| author | Vasilis Psiroukis Borja Espejo-Garcia Andreas Chitos Athanasios Dedousis Konstantinos Karantzalos Spyros Fountas |
| author_facet | Vasilis Psiroukis Borja Espejo-Garcia Andreas Chitos Athanasios Dedousis Konstantinos Karantzalos Spyros Fountas |
| author_sort | Vasilis Psiroukis |
| collection | DOAJ |
| description | Broccoli is an example of a high-value crop that requires delicate handling throughout the growing season and during its post-harvesting treatment. As broccoli heads can be easily damaged, they are still harvested by hand. Moreover, human scouting is required to initially identify the field segments where several broccoli plants have reached the desired maturity level, such that they can be harvested while they are in the optimal condition. The aim of this study was to automate this process using state-of-the-art Object Detection architectures trained on georeferenced orthomosaic-derived RGB images captured from low-altitude UAV flights, and to assess their capacity to effectively detect and classify broccoli heads based on their maturity level. The results revealed that the object detection approach for automated maturity classification achieved comparable results to physical scouting overall, especially for the two best-performing architectures, namely Faster R-CNN and CenterNet. Their respective performances were consistently over 80% mAP@50 and 70% mAP@75 when using three levels of maturity, and even higher when simplifying the use case into a two-class problem, exceeding 91% and 83%, respectively. At the same time, geometrical transformations for data augmentations reported improvements, while colour distortions were counterproductive. The best-performing architecture and the trained model could be tested as a prototype in real-time UAV detections in order to assist in on-field broccoli maturity detection. |
| first_indexed | 2024-03-09T23:13:15Z |
| format | Article |
| id | doaj.art-f52f0853819d49cb972c06c2d0a5f36e |
| institution | Directory Open Access Journal |
| issn | 2072-4292 |
| language | English |
| last_indexed | 2024-03-09T23:13:15Z |
| publishDate | 2022-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Remote Sensing |
| spelling | doaj.art-f52f0853819d49cb972c06c2d0a5f36e2023-11-23T17:42:34ZengMDPI AGRemote Sensing2072-42922022-02-0114373110.3390/rs14030731Assessment of Different Object Detectors for the Maturity Level Classification of Broccoli Crops Using UAV ImageryVasilis Psiroukis0Borja Espejo-Garcia1Andreas Chitos2Athanasios Dedousis3Konstantinos Karantzalos4Spyros Fountas5Laboratory of Agricultural Engineering, Department of Natural Resources Management & Agricultural Engineering, School of Environment and Agricultural Engineering, Agricultural University of Athens, 11855 Athens, GreeceLaboratory of Agricultural Engineering, Department of Natural Resources Management & Agricultural Engineering, School of Environment and Agricultural Engineering, Agricultural University of Athens, 11855 Athens, GreeceLaboratory of Agricultural Engineering, Department of Natural Resources Management & Agricultural Engineering, School of Environment and Agricultural Engineering, Agricultural University of Athens, 11855 Athens, GreeceLaboratory of Agricultural Engineering, Department of Natural Resources Management & Agricultural Engineering, School of Environment and Agricultural Engineering, Agricultural University of Athens, 11855 Athens, GreeceLaboratory of Remote Sensing, Department of Topography, School of Rural, Surveying and Geoinformatics Engineering, National Technical University of Athens, 10682 Athens, GreeceLaboratory of Agricultural Engineering, Department of Natural Resources Management & Agricultural Engineering, School of Environment and Agricultural Engineering, Agricultural University of Athens, 11855 Athens, GreeceBroccoli is an example of a high-value crop that requires delicate handling throughout the growing season and during its post-harvesting treatment. As broccoli heads can be easily damaged, they are still harvested by hand. Moreover, human scouting is required to initially identify the field segments where several broccoli plants have reached the desired maturity level, such that they can be harvested while they are in the optimal condition. The aim of this study was to automate this process using state-of-the-art Object Detection architectures trained on georeferenced orthomosaic-derived RGB images captured from low-altitude UAV flights, and to assess their capacity to effectively detect and classify broccoli heads based on their maturity level. The results revealed that the object detection approach for automated maturity classification achieved comparable results to physical scouting overall, especially for the two best-performing architectures, namely Faster R-CNN and CenterNet. Their respective performances were consistently over 80% mAP@50 and 70% mAP@75 when using three levels of maturity, and even higher when simplifying the use case into a two-class problem, exceeding 91% and 83%, respectively. At the same time, geometrical transformations for data augmentations reported improvements, while colour distortions were counterproductive. The best-performing architecture and the trained model could be tested as a prototype in real-time UAV detections in order to assist in on-field broccoli maturity detection.https://www.mdpi.com/2072-4292/14/3/731object detectionUAV imagesmaturity detectionefficientdetretinanetcenternet |
| spellingShingle | Vasilis Psiroukis Borja Espejo-Garcia Andreas Chitos Athanasios Dedousis Konstantinos Karantzalos Spyros Fountas Assessment of Different Object Detectors for the Maturity Level Classification of Broccoli Crops Using UAV Imagery Remote Sensing object detection UAV images maturity detection efficientdet retinanet centernet |
| title | Assessment of Different Object Detectors for the Maturity Level Classification of Broccoli Crops Using UAV Imagery |
| title_full | Assessment of Different Object Detectors for the Maturity Level Classification of Broccoli Crops Using UAV Imagery |
| title_fullStr | Assessment of Different Object Detectors for the Maturity Level Classification of Broccoli Crops Using UAV Imagery |
| title_full_unstemmed | Assessment of Different Object Detectors for the Maturity Level Classification of Broccoli Crops Using UAV Imagery |
| title_short | Assessment of Different Object Detectors for the Maturity Level Classification of Broccoli Crops Using UAV Imagery |
| title_sort | assessment of different object detectors for the maturity level classification of broccoli crops using uav imagery |
| topic | object detection UAV images maturity detection efficientdet retinanet centernet |
| url | https://www.mdpi.com/2072-4292/14/3/731 |
| work_keys_str_mv | AT vasilispsiroukis assessmentofdifferentobjectdetectorsforthematuritylevelclassificationofbroccolicropsusinguavimagery AT borjaespejogarcia assessmentofdifferentobjectdetectorsforthematuritylevelclassificationofbroccolicropsusinguavimagery AT andreaschitos assessmentofdifferentobjectdetectorsforthematuritylevelclassificationofbroccolicropsusinguavimagery AT athanasiosdedousis assessmentofdifferentobjectdetectorsforthematuritylevelclassificationofbroccolicropsusinguavimagery AT konstantinoskarantzalos assessmentofdifferentobjectdetectorsforthematuritylevelclassificationofbroccolicropsusinguavimagery AT spyrosfountas assessmentofdifferentobjectdetectorsforthematuritylevelclassificationofbroccolicropsusinguavimagery |