Instance Segmentation and 3D Pose Estimation of Tea Bud Leaves for Autonomous Harvesting Robots
In unstructured tea garden environments, accurate recognition and pose estimation of tea bud leaves are critical for autonomous harvesting robots. Due to variations in imaging distance, tea bud leaves exhibit diverse scale and pose characteristics in camera views, which significantly complicates the...
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-01-01
|
| Series: | Agriculture |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2077-0472/15/2/198 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832589473031913472 |
|---|---|
| author | Haoxin Li Tianci Chen Yingmei Chen Chongyang Han Jinhong Lv Zhiheng Zhou Weibin Wu |
| author_facet | Haoxin Li Tianci Chen Yingmei Chen Chongyang Han Jinhong Lv Zhiheng Zhou Weibin Wu |
| author_sort | Haoxin Li |
| collection | DOAJ |
| description | In unstructured tea garden environments, accurate recognition and pose estimation of tea bud leaves are critical for autonomous harvesting robots. Due to variations in imaging distance, tea bud leaves exhibit diverse scale and pose characteristics in camera views, which significantly complicates the recognition and pose estimation process. This study proposes a method using an RGB-D camera for precise recognition and pose estimation of tea bud leaves. The approach first constructs an for tea bud leaves, followed by a dynamic weight estimation strategy to achieve adaptive pose estimation. Quantitative experiments demonstrate that the instance segmentation model achieves an mAP@50 of 92.0% for box detection and 91.9% for mask detection, improving by 3.2% and 3.4%, respectively, compared to the YOLOv8s-seg instance segmentation model. The pose estimation results indicate a maximum angular error of 7.76°, a mean angular error of 3.41°, a median angular error of 3.69°, and a median absolute deviation of 1.42°. The corresponding distance errors are 8.60 mm, 2.83 mm, 2.57 mm, and 0.81 mm, further confirming the accuracy and robustness of the proposed method. These results indicate that the proposed method can be applied in unstructured tea garden environments for non-destructive and precise harvesting with autonomous tea bud-leave harvesting robots. |
| format | Article |
| id | doaj-art-67a36e7a2cf341ac88609f6886b3330a |
| institution | Kabale University |
| issn | 2077-0472 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Agriculture |
| spelling | doaj-art-67a36e7a2cf341ac88609f6886b3330a2025-01-24T13:16:06ZengMDPI AGAgriculture2077-04722025-01-0115219810.3390/agriculture15020198Instance Segmentation and 3D Pose Estimation of Tea Bud Leaves for Autonomous Harvesting RobotsHaoxin Li0Tianci Chen1Yingmei Chen2Chongyang Han3Jinhong Lv4Zhiheng Zhou5Weibin Wu6National Key Laboratory of Agricultural Equipment Technology, College of Engineering, South China Agricultural University, Guangzhou 510642, ChinaNational Key Laboratory of Agricultural Equipment Technology, College of Engineering, South China Agricultural University, Guangzhou 510642, ChinaNational Key Laboratory of Agricultural Equipment Technology, College of Engineering, South China Agricultural University, Guangzhou 510642, ChinaNational Key Laboratory of Agricultural Equipment Technology, College of Engineering, South China Agricultural University, Guangzhou 510642, ChinaNational Key Laboratory of Agricultural Equipment Technology, College of Engineering, South China Agricultural University, Guangzhou 510642, ChinaSchool of Electronic and Information Engineering, South China University of Technology, Guangzhou 510640, ChinaNational Key Laboratory of Agricultural Equipment Technology, College of Engineering, South China Agricultural University, Guangzhou 510642, ChinaIn unstructured tea garden environments, accurate recognition and pose estimation of tea bud leaves are critical for autonomous harvesting robots. Due to variations in imaging distance, tea bud leaves exhibit diverse scale and pose characteristics in camera views, which significantly complicates the recognition and pose estimation process. This study proposes a method using an RGB-D camera for precise recognition and pose estimation of tea bud leaves. The approach first constructs an for tea bud leaves, followed by a dynamic weight estimation strategy to achieve adaptive pose estimation. Quantitative experiments demonstrate that the instance segmentation model achieves an mAP@50 of 92.0% for box detection and 91.9% for mask detection, improving by 3.2% and 3.4%, respectively, compared to the YOLOv8s-seg instance segmentation model. The pose estimation results indicate a maximum angular error of 7.76°, a mean angular error of 3.41°, a median angular error of 3.69°, and a median absolute deviation of 1.42°. The corresponding distance errors are 8.60 mm, 2.83 mm, 2.57 mm, and 0.81 mm, further confirming the accuracy and robustness of the proposed method. These results indicate that the proposed method can be applied in unstructured tea garden environments for non-destructive and precise harvesting with autonomous tea bud-leave harvesting robots.https://www.mdpi.com/2077-0472/15/2/198YOLOv8s-seg modeladaptive pose estimationRGB-D cameraprecise harvesting |
| spellingShingle | Haoxin Li Tianci Chen Yingmei Chen Chongyang Han Jinhong Lv Zhiheng Zhou Weibin Wu Instance Segmentation and 3D Pose Estimation of Tea Bud Leaves for Autonomous Harvesting Robots Agriculture YOLOv8s-seg model adaptive pose estimation RGB-D camera precise harvesting |
| title | Instance Segmentation and 3D Pose Estimation of Tea Bud Leaves for Autonomous Harvesting Robots |
| title_full | Instance Segmentation and 3D Pose Estimation of Tea Bud Leaves for Autonomous Harvesting Robots |
| title_fullStr | Instance Segmentation and 3D Pose Estimation of Tea Bud Leaves for Autonomous Harvesting Robots |
| title_full_unstemmed | Instance Segmentation and 3D Pose Estimation of Tea Bud Leaves for Autonomous Harvesting Robots |
| title_short | Instance Segmentation and 3D Pose Estimation of Tea Bud Leaves for Autonomous Harvesting Robots |
| title_sort | instance segmentation and 3d pose estimation of tea bud leaves for autonomous harvesting robots |
| topic | YOLOv8s-seg model adaptive pose estimation RGB-D camera precise harvesting |
| url | https://www.mdpi.com/2077-0472/15/2/198 |
| work_keys_str_mv | AT haoxinli instancesegmentationand3dposeestimationofteabudleavesforautonomousharvestingrobots AT tiancichen instancesegmentationand3dposeestimationofteabudleavesforautonomousharvestingrobots AT yingmeichen instancesegmentationand3dposeestimationofteabudleavesforautonomousharvestingrobots AT chongyanghan instancesegmentationand3dposeestimationofteabudleavesforautonomousharvestingrobots AT jinhonglv instancesegmentationand3dposeestimationofteabudleavesforautonomousharvestingrobots AT zhihengzhou instancesegmentationand3dposeestimationofteabudleavesforautonomousharvestingrobots AT weibinwu instancesegmentationand3dposeestimationofteabudleavesforautonomousharvestingrobots |