Identifying and mapping individual medicinal plant Lamiophlomis rotata at high elevations by using unmanned aerial vehicles and deep learning
Abstract Background The identification and enumeration of medicinal plants at high elevations is an important part of accurate yield calculations. However, the current assessment of medicinal plant reserves continues to rely on field sampling surveys, which are cumbersome and time-consuming. Recentl...
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2023-04-01
|
| Series: | Plant Methods |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s13007-023-01015-z |
| _version_ | 1797853765395546112 |
|---|---|
| author | Rong Ding Jiawei Luo Chenghui Wang Lianhui Yu Jiangkai Yang Meng Wang Shihong Zhong Rui Gu |
| author_facet | Rong Ding Jiawei Luo Chenghui Wang Lianhui Yu Jiangkai Yang Meng Wang Shihong Zhong Rui Gu |
| author_sort | Rong Ding |
| collection | DOAJ |
| description | Abstract Background The identification and enumeration of medicinal plants at high elevations is an important part of accurate yield calculations. However, the current assessment of medicinal plant reserves continues to rely on field sampling surveys, which are cumbersome and time-consuming. Recently, unmanned aerial vehicle (UAV) remote sensing and deep learning (DL) have provided ultrahigh-resolution imagery and high-accuracy object recognition techniques, respectively, providing an excellent opportunity to improve the current manual surveying of plants. However, accurate segmentation of individual plants from drone images remains a significant challenge due to the large variation in size, geometry, and distribution of medicinal plants. Results In this study, we proposed a new pipeline for wild medicinal plant detection and yield assessment based on UAV and DL that was specifically designed for detecting wild medicinal plants in an orthomosaic. We used a drone to collect panoramic images of Lamioplomis rotata Kudo (LR) in high-altitude areas. Then, we annotated and cropped these images into equally sized sub-images and used a DL model Mask R-CNN for object detection and segmentation of LR. Finally, on the basis of the segmentation results, we accurately counted the number and yield of LRs. The results showed that the Mask R-CNN model based on the ResNet-101 backbone network was superior to ResNet-50 in all evaluation indicators. The average identification precision of LR by Mask R-CNN based on the ResNet-101 backbone network was 89.34%, while that of ResNet-50 was 88.32%. The cross-validation results showed that the average accuracy of ResNet-101 was 78.73%, while that of ResNet-50 was 71.25%. According to the orthomosaic, the average number and yield of LR in the two sample sites were 19,376 plants and 57.93 kg and 19,129 plants and 73.5 kg respectively. Conclusions The combination of DL and UAV remote sensing reveals significant promise in medicinal plant detection, counting, and yield prediction, which will benefit the monitoring of their populations for conservation assessment and management, among other applications. |
| first_indexed | 2024-04-09T19:54:55Z |
| format | Article |
| id | doaj.art-16baad53898747ba8944fd47926825ae |
| institution | Directory Open Access Journal |
| issn | 1746-4811 |
| language | English |
| last_indexed | 2024-04-09T19:54:55Z |
| publishDate | 2023-04-01 |
| publisher | BMC |
| record_format | Article |
| series | Plant Methods |
| spelling | doaj.art-16baad53898747ba8944fd47926825ae2023-04-03T05:28:23ZengBMCPlant Methods1746-48112023-04-0119111610.1186/s13007-023-01015-zIdentifying and mapping individual medicinal plant Lamiophlomis rotata at high elevations by using unmanned aerial vehicles and deep learningRong Ding0Jiawei Luo1Chenghui Wang2Lianhui Yu3Jiangkai Yang4Meng Wang5Shihong Zhong6Rui Gu7School of Pharmacy, Chengdu University of Traditional Chinese MedicineWest China Biomedical Big Data Center, West China Hospital/West China School of Medicine, Sichuan UniversitySchool of Ethnic Medicine, Chengdu University of Traditional Chinese MedicineChengdu Pushi Pharmaceutical Technology Co., LtdSchool of Pharmacy, Chengdu University of Traditional Chinese MedicineInstitute of Geological Survey of Sichuan ProvincialSchool of Pharmacy, Southwest Minzu UniversitySchool of Ethnic Medicine, Chengdu University of Traditional Chinese MedicineAbstract Background The identification and enumeration of medicinal plants at high elevations is an important part of accurate yield calculations. However, the current assessment of medicinal plant reserves continues to rely on field sampling surveys, which are cumbersome and time-consuming. Recently, unmanned aerial vehicle (UAV) remote sensing and deep learning (DL) have provided ultrahigh-resolution imagery and high-accuracy object recognition techniques, respectively, providing an excellent opportunity to improve the current manual surveying of plants. However, accurate segmentation of individual plants from drone images remains a significant challenge due to the large variation in size, geometry, and distribution of medicinal plants. Results In this study, we proposed a new pipeline for wild medicinal plant detection and yield assessment based on UAV and DL that was specifically designed for detecting wild medicinal plants in an orthomosaic. We used a drone to collect panoramic images of Lamioplomis rotata Kudo (LR) in high-altitude areas. Then, we annotated and cropped these images into equally sized sub-images and used a DL model Mask R-CNN for object detection and segmentation of LR. Finally, on the basis of the segmentation results, we accurately counted the number and yield of LRs. The results showed that the Mask R-CNN model based on the ResNet-101 backbone network was superior to ResNet-50 in all evaluation indicators. The average identification precision of LR by Mask R-CNN based on the ResNet-101 backbone network was 89.34%, while that of ResNet-50 was 88.32%. The cross-validation results showed that the average accuracy of ResNet-101 was 78.73%, while that of ResNet-50 was 71.25%. According to the orthomosaic, the average number and yield of LR in the two sample sites were 19,376 plants and 57.93 kg and 19,129 plants and 73.5 kg respectively. Conclusions The combination of DL and UAV remote sensing reveals significant promise in medicinal plant detection, counting, and yield prediction, which will benefit the monitoring of their populations for conservation assessment and management, among other applications.https://doi.org/10.1186/s13007-023-01015-zLamiophlomis rotataOrthomosaicMedicinal plant detectionMedicinal plant mappingYield predictionDeep learning |
| spellingShingle | Rong Ding Jiawei Luo Chenghui Wang Lianhui Yu Jiangkai Yang Meng Wang Shihong Zhong Rui Gu Identifying and mapping individual medicinal plant Lamiophlomis rotata at high elevations by using unmanned aerial vehicles and deep learning Plant Methods Lamiophlomis rotata Orthomosaic Medicinal plant detection Medicinal plant mapping Yield prediction Deep learning |
| title | Identifying and mapping individual medicinal plant Lamiophlomis rotata at high elevations by using unmanned aerial vehicles and deep learning |
| title_full | Identifying and mapping individual medicinal plant Lamiophlomis rotata at high elevations by using unmanned aerial vehicles and deep learning |
| title_fullStr | Identifying and mapping individual medicinal plant Lamiophlomis rotata at high elevations by using unmanned aerial vehicles and deep learning |
| title_full_unstemmed | Identifying and mapping individual medicinal plant Lamiophlomis rotata at high elevations by using unmanned aerial vehicles and deep learning |
| title_short | Identifying and mapping individual medicinal plant Lamiophlomis rotata at high elevations by using unmanned aerial vehicles and deep learning |
| title_sort | identifying and mapping individual medicinal plant lamiophlomis rotata at high elevations by using unmanned aerial vehicles and deep learning |
| topic | Lamiophlomis rotata Orthomosaic Medicinal plant detection Medicinal plant mapping Yield prediction Deep learning |
| url | https://doi.org/10.1186/s13007-023-01015-z |
| work_keys_str_mv | AT rongding identifyingandmappingindividualmedicinalplantlamiophlomisrotataathighelevationsbyusingunmannedaerialvehiclesanddeeplearning AT jiaweiluo identifyingandmappingindividualmedicinalplantlamiophlomisrotataathighelevationsbyusingunmannedaerialvehiclesanddeeplearning AT chenghuiwang identifyingandmappingindividualmedicinalplantlamiophlomisrotataathighelevationsbyusingunmannedaerialvehiclesanddeeplearning AT lianhuiyu identifyingandmappingindividualmedicinalplantlamiophlomisrotataathighelevationsbyusingunmannedaerialvehiclesanddeeplearning AT jiangkaiyang identifyingandmappingindividualmedicinalplantlamiophlomisrotataathighelevationsbyusingunmannedaerialvehiclesanddeeplearning AT mengwang identifyingandmappingindividualmedicinalplantlamiophlomisrotataathighelevationsbyusingunmannedaerialvehiclesanddeeplearning AT shihongzhong identifyingandmappingindividualmedicinalplantlamiophlomisrotataathighelevationsbyusingunmannedaerialvehiclesanddeeplearning AT ruigu identifyingandmappingindividualmedicinalplantlamiophlomisrotataathighelevationsbyusingunmannedaerialvehiclesanddeeplearning |