Identifying the Branch of Kiwifruit Based on Unmanned Aerial Vehicle (UAV) Images Using Deep Learning Method
It is important to obtain accurate information about kiwifruit vines to monitoring their physiological states and undertake precise orchard operations. However, because vines are small and cling to trellises, and have branches laying on the ground, numerous challenges exist in the acquisition of acc...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-06-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/21/13/4442 |
| _version_ | 1797528394684956672 |
|---|---|
| author | Zijie Niu Juntao Deng Xu Zhang Jun Zhang Shijia Pan Haotian Mu |
| author_facet | Zijie Niu Juntao Deng Xu Zhang Jun Zhang Shijia Pan Haotian Mu |
| author_sort | Zijie Niu |
| collection | DOAJ |
| description | It is important to obtain accurate information about kiwifruit vines to monitoring their physiological states and undertake precise orchard operations. However, because vines are small and cling to trellises, and have branches laying on the ground, numerous challenges exist in the acquisition of accurate data for kiwifruit vines. In this paper, a kiwifruit canopy distribution prediction model is proposed on the basis of low-altitude unmanned aerial vehicle (UAV) images and deep learning techniques. First, the location of the kiwifruit plants and vine distribution are extracted from high-precision images collected by UAV. The canopy gradient distribution maps with different noise reduction and distribution effects are generated by modifying the threshold and sampling size using the resampling normalization method. The results showed that the accuracies of the vine segmentation using PSPnet, support vector machine, and random forest classification were 71.2%, 85.8%, and 75.26%, respectively. However, the segmentation image obtained using depth semantic segmentation had a higher signal-to-noise ratio and was closer to the real situation. The average intersection over union of the deep semantic segmentation was more than or equal to 80% in distribution maps, whereas, in traditional machine learning, the average intersection was between 20% and 60%. This indicates the proposed model can quickly extract the vine distribution and plant position, and is thus able to perform dynamic monitoring of orchards to provide real-time operation guidance. |
| first_indexed | 2024-03-10T09:58:39Z |
| format | Article |
| id | doaj.art-5c7631afb90f4b3fbab3de6a99e071ab |
| institution | Directory Open Access Journal |
| issn | 1424-8220 |
| language | English |
| last_indexed | 2024-03-10T09:58:39Z |
| publishDate | 2021-06-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj.art-5c7631afb90f4b3fbab3de6a99e071ab2023-11-22T02:08:22ZengMDPI AGSensors1424-82202021-06-012113444210.3390/s21134442Identifying the Branch of Kiwifruit Based on Unmanned Aerial Vehicle (UAV) Images Using Deep Learning MethodZijie Niu0Juntao Deng1Xu Zhang2Jun Zhang3Shijia Pan4Haotian Mu5College of Mechanical and Electronic Engineering, Northwest Agriculture & Forestry University, Xi’an 712100, ChinaCollege of Mechanical and Electronic Engineering, Northwest Agriculture & Forestry University, Xi’an 712100, ChinaCollege of Mechanical and Electronic Engineering, Northwest Agriculture & Forestry University, Xi’an 712100, ChinaCollege of Mechanical and Electronic Engineering, Northwest Agriculture & Forestry University, Xi’an 712100, ChinaCollege of Mechanical and Electronic Engineering, Northwest Agriculture & Forestry University, Xi’an 712100, ChinaCollege of Mechanical and Electronic Engineering, Northwest Agriculture & Forestry University, Xi’an 712100, ChinaIt is important to obtain accurate information about kiwifruit vines to monitoring their physiological states and undertake precise orchard operations. However, because vines are small and cling to trellises, and have branches laying on the ground, numerous challenges exist in the acquisition of accurate data for kiwifruit vines. In this paper, a kiwifruit canopy distribution prediction model is proposed on the basis of low-altitude unmanned aerial vehicle (UAV) images and deep learning techniques. First, the location of the kiwifruit plants and vine distribution are extracted from high-precision images collected by UAV. The canopy gradient distribution maps with different noise reduction and distribution effects are generated by modifying the threshold and sampling size using the resampling normalization method. The results showed that the accuracies of the vine segmentation using PSPnet, support vector machine, and random forest classification were 71.2%, 85.8%, and 75.26%, respectively. However, the segmentation image obtained using depth semantic segmentation had a higher signal-to-noise ratio and was closer to the real situation. The average intersection over union of the deep semantic segmentation was more than or equal to 80% in distribution maps, whereas, in traditional machine learning, the average intersection was between 20% and 60%. This indicates the proposed model can quickly extract the vine distribution and plant position, and is thus able to perform dynamic monitoring of orchards to provide real-time operation guidance.https://www.mdpi.com/1424-8220/21/13/4442deep learningunmanned aerial vehiclekiwifruitimage segmentation |
| spellingShingle | Zijie Niu Juntao Deng Xu Zhang Jun Zhang Shijia Pan Haotian Mu Identifying the Branch of Kiwifruit Based on Unmanned Aerial Vehicle (UAV) Images Using Deep Learning Method Sensors deep learning unmanned aerial vehicle kiwifruit image segmentation |
| title | Identifying the Branch of Kiwifruit Based on Unmanned Aerial Vehicle (UAV) Images Using Deep Learning Method |
| title_full | Identifying the Branch of Kiwifruit Based on Unmanned Aerial Vehicle (UAV) Images Using Deep Learning Method |
| title_fullStr | Identifying the Branch of Kiwifruit Based on Unmanned Aerial Vehicle (UAV) Images Using Deep Learning Method |
| title_full_unstemmed | Identifying the Branch of Kiwifruit Based on Unmanned Aerial Vehicle (UAV) Images Using Deep Learning Method |
| title_short | Identifying the Branch of Kiwifruit Based on Unmanned Aerial Vehicle (UAV) Images Using Deep Learning Method |
| title_sort | identifying the branch of kiwifruit based on unmanned aerial vehicle uav images using deep learning method |
| topic | deep learning unmanned aerial vehicle kiwifruit image segmentation |
| url | https://www.mdpi.com/1424-8220/21/13/4442 |
| work_keys_str_mv | AT zijieniu identifyingthebranchofkiwifruitbasedonunmannedaerialvehicleuavimagesusingdeeplearningmethod AT juntaodeng identifyingthebranchofkiwifruitbasedonunmannedaerialvehicleuavimagesusingdeeplearningmethod AT xuzhang identifyingthebranchofkiwifruitbasedonunmannedaerialvehicleuavimagesusingdeeplearningmethod AT junzhang identifyingthebranchofkiwifruitbasedonunmannedaerialvehicleuavimagesusingdeeplearningmethod AT shijiapan identifyingthebranchofkiwifruitbasedonunmannedaerialvehicleuavimagesusingdeeplearningmethod AT haotianmu identifyingthebranchofkiwifruitbasedonunmannedaerialvehicleuavimagesusingdeeplearningmethod |