2026
1.
Ranjan Sapkota; Zhichao Meng; Martin Churuvija; Xiaoqiang Du; Zenghong Ma; Manoj Karkee
In: Agriculture Communications, vol. 4, no. 1, pp. 100125, 2026, ISSN: 2949-7981.
Abstract | Links | BibTeX | Tags: agricultural automation, Fruitlet detection, Object detection, YOLO comparison, You Only Look Once (YOLO)
@article{sapkota_comprehensive_2026,
title = {Comprehensive performance evaluation of YOLOv12, YOLO11, YOLOv10, YOLOv9 and YOLOv8 on detecting and counting fruitlet in complex orchard environments},
author = {Ranjan Sapkota and Zhichao Meng and Martin Churuvija and Xiaoqiang Du and Zenghong Ma and Manoj Karkee},
url = {https://www.sciencedirect.com/science/article/pii/S2949798126000050},
doi = {https://doi.org/10.1016/j.agrcom.2026.100125},
issn = {2949-7981},
year = {2026},
date = {2026-01-01},
urldate = {2026-01-01},
journal = {Agriculture Communications},
volume = {4},
number = {1},
pages = {100125},
abstract = {This study systematically conducted an extensive real-world evaluation of all configurations of You Only Look Once (YOLO)-based object detection algorithms, including YOLOv8, YOLOv9, YOLOv10, YOLO11, and YOLOv12. Models were assessed using precision, recall, mean Average Precision at 50 % Intersection over Union (mAP@50), and computational efficiency across pre-processing, inference, and post-processing stages for detecting immature green fruitlets in commercial orchards. Field-level fruitlet counting was also validated using images captured with both Intel RealSense and iPhone 14 Pro Max sensors. YOLOv12l achieved the highest recall (0.900), while YOLOv10x and YOLOv9 GELAN-c reported the top precision scores of 0.908 and 0.903, respectively. YOLOv9 GELAN-base and GELAN-e achieved the highest mAP@50 (0.935), followed by YOLO11s (0.933) and YOLOv12l (0.931). In counting validation, YOLO11n demonstrated superior accuracy, with RMSE values of 4.51\textendash4.96 and MAE values of 3.85\textendash7.73 across four apple varieties. Sensor-specific training on Intel RealSense further improved detection performance. YOLO11n also recorded the fastest inference speed (2.4 ms), outperforming YOLOv8n, YOLOv9 GELAN-s, YOLOv10n, and YOLOv12n, affirming its suitability for real-time orchard applications.},
keywords = {agricultural automation, Fruitlet detection, Object detection, YOLO comparison, You Only Look Once (YOLO)},
pubstate = {published},
tppubtype = {article}
}
This study systematically conducted an extensive real-world evaluation of all configurations of You Only Look Once (YOLO)-based object detection algorithms, including YOLOv8, YOLOv9, YOLOv10, YOLO11, and YOLOv12. Models were assessed using precision, recall, mean Average Precision at 50 % Intersection over Union (mAP@50), and computational efficiency across pre-processing, inference, and post-processing stages for detecting immature green fruitlets in commercial orchards. Field-level fruitlet counting was also validated using images captured with both Intel RealSense and iPhone 14 Pro Max sensors. YOLOv12l achieved the highest recall (0.900), while YOLOv10x and YOLOv9 GELAN-c reported the top precision scores of 0.908 and 0.903, respectively. YOLOv9 GELAN-base and GELAN-e achieved the highest mAP@50 (0.935), followed by YOLO11s (0.933) and YOLOv12l (0.931). In counting validation, YOLO11n demonstrated superior accuracy, with RMSE values of 4.51–4.96 and MAE values of 3.85–7.73 across four apple varieties. Sensor-specific training on Intel RealSense further improved detection performance. YOLO11n also recorded the fastest inference speed (2.4 ms), outperforming YOLOv8n, YOLOv9 GELAN-s, YOLOv10n, and YOLOv12n, affirming its suitability for real-time orchard applications.
2025
2.
Ranjan Sapkota; Manoj Karkee
In: IFAC-PapersOnLine, vol. 59, no. 23, pp. 239–244, 2025, ISSN: 2405-8963.
Abstract | Links | BibTeX | Tags: agricultural automation, Apple detection, LLM, Synthetic Image Generation, YOLOv10, YOLOv11, YOLOv12, YOLOv12 object detection, You Only Look Once
@article{sapkota_improved_2025,
title = {Improved YOLOv12 with LLM-Generated Synthetic Data for Enhanced Apple Detection and Benchmarking Against YOLOv11 and YOLOv10⁎⁎Sponsor and financial support acknowledgment goes here. Paper titles should be written in uppercase and lowercase letters, not all uppercase.},
author = {Ranjan Sapkota and Manoj Karkee},
url = {https://www.sciencedirect.com/science/article/pii/S2405896325024929},
doi = {https://doi.org/10.1016/j.ifacol.2025.11.793},
issn = {2405-8963},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {IFAC-PapersOnLine},
volume = {59},
number = {23},
pages = {239\textendash244},
abstract = {Accurate fruit detection in orchards remains a major bottleneck in agricultural automation due to the high cost and labor demands of data collection and annotation. This study evaluated the performance of the YOLOv12 object detection model, and compared against the performances YOLOv11 and YOLOv10 for apple detection in commercial orchards based on the model training completed entirely on synthetic images generated by Large Language Models (LLMs). The YOLOv12n configuration achieved the highest precision at 0.916, the highest recall at 0.969, and the highest mean Average Precision (mAP@50) at 0.978. In comparison, the YOLOv11 series was led by YOLO11x, which achieved the highest precision at 0.857, recall at 0.85, and mAP@50 at 0.91. For the YOLOv10 series, YOLOv10b and YOLOv10l both achieved the highest precision at 0.85, with YOLOv10n achieving the highest recall at 0.8 and mAP@50 at 0.89. These findings demonstrated that YOLOv12, when trained on realistic LLM-generated datasets surpassed its predecessors in key performance metrics. The technique also offered a cost-effective solution by reducing the need for extensive manual data collection in the agricultural field. In addition, this study compared the computational efficiency of all versions of YOLOv12, v11 and v10, where YOLOv11n reported the lowest inference time at 4.7 ms, compared to YOLOv12n’s 5.6 ms and YOLOv10n’s 5.9 ms. Although YOLOv12 is new and more accurate than YOLOv11, and YOLOv10, YOLO11n still stays the fastest YOLO model among YOLOv10, YOLOv11 and YOLOv12 series of models.},
keywords = {agricultural automation, Apple detection, LLM, Synthetic Image Generation, YOLOv10, YOLOv11, YOLOv12, YOLOv12 object detection, You Only Look Once},
pubstate = {published},
tppubtype = {article}
}
Accurate fruit detection in orchards remains a major bottleneck in agricultural automation due to the high cost and labor demands of data collection and annotation. This study evaluated the performance of the YOLOv12 object detection model, and compared against the performances YOLOv11 and YOLOv10 for apple detection in commercial orchards based on the model training completed entirely on synthetic images generated by Large Language Models (LLMs). The YOLOv12n configuration achieved the highest precision at 0.916, the highest recall at 0.969, and the highest mean Average Precision (mAP@50) at 0.978. In comparison, the YOLOv11 series was led by YOLO11x, which achieved the highest precision at 0.857, recall at 0.85, and mAP@50 at 0.91. For the YOLOv10 series, YOLOv10b and YOLOv10l both achieved the highest precision at 0.85, with YOLOv10n achieving the highest recall at 0.8 and mAP@50 at 0.89. These findings demonstrated that YOLOv12, when trained on realistic LLM-generated datasets surpassed its predecessors in key performance metrics. The technique also offered a cost-effective solution by reducing the need for extensive manual data collection in the agricultural field. In addition, this study compared the computational efficiency of all versions of YOLOv12, v11 and v10, where YOLOv11n reported the lowest inference time at 4.7 ms, compared to YOLOv12n’s 5.6 ms and YOLOv10n’s 5.9 ms. Although YOLOv12 is new and more accurate than YOLOv11, and YOLOv10, YOLO11n still stays the fastest YOLO model among YOLOv10, YOLOv11 and YOLOv12 series of models.
3.
Dawood Ahmed; Basit Muhammad Imran; Martin Churuvija; Manoj Karkee
An Integrated Visual Servoing Framework for Precise Robotic Pruning Operations in Modern Commercial Orchard Journal Article
In: IFAC-PapersOnLine, vol. 59, no. 23, pp. 28–33, 2025, ISSN: 2405-8963.
Abstract | Links | BibTeX | Tags: agricultural automation, agricultural robotics, perception, precision pruning, robotic manipulator, sensing, visual servoing
@article{ahmed_integrated_2025,
title = {An Integrated Visual Servoing Framework for Precise Robotic Pruning Operations in Modern Commercial Orchard},
author = {Dawood Ahmed and Basit Muhammad Imran and Martin Churuvija and Manoj Karkee},
url = {https://www.sciencedirect.com/science/article/pii/S2405896325024565},
doi = {https://doi.org/10.1016/j.ifacol.2025.11.758},
issn = {2405-8963},
year = {2025},
date = {2025-01-01},
urldate = {2025-01-01},
journal = {IFAC-PapersOnLine},
volume = {59},
number = {23},
pages = {28\textendash33},
abstract = {This study presents a vision-guided robotic control system for automated fruit tree pruning applications. Traditional pruning practices are labor-intensive and limit agricultural efficiency and scalability, highlighting the need for advanced automation. A key challenge is the precise, robust positioning of the cutting tool in complex orchard environments, where dense branches and occlusions make target access difficult. To address this, an Intel RealSense D435 camera is mounted on the flange of a UR5e robotic arm and CoTracker3, a transformer-based point tracker, is utilized for visual servoing control that centers tracked points in the camera view. The system integrates proportional control with iterative inverse kinematics to achieve precise end-effector positioning. The system was validated in Gazebo simulation, achieving a 77.77% success rate within 5mm positional tolerance and 100% success rate within 10mm tolerance, with a mean end-effector error of 4.28 ± 1.36 mm. The vision controller demonstrated robust performance across diverse target positions within the pixel workspace. The results validate the effectiveness of integrating vision-based tracking with kinematic control for precision agricultural tasks. Future work will focus on real-world implementation and the integration of force sensing for actual cutting operations.},
keywords = {agricultural automation, agricultural robotics, perception, precision pruning, robotic manipulator, sensing, visual servoing},
pubstate = {published},
tppubtype = {article}
}
This study presents a vision-guided robotic control system for automated fruit tree pruning applications. Traditional pruning practices are labor-intensive and limit agricultural efficiency and scalability, highlighting the need for advanced automation. A key challenge is the precise, robust positioning of the cutting tool in complex orchard environments, where dense branches and occlusions make target access difficult. To address this, an Intel RealSense D435 camera is mounted on the flange of a UR5e robotic arm and CoTracker3, a transformer-based point tracker, is utilized for visual servoing control that centers tracked points in the camera view. The system integrates proportional control with iterative inverse kinematics to achieve precise end-effector positioning. The system was validated in Gazebo simulation, achieving a 77.77% success rate within 5mm positional tolerance and 100% success rate within 10mm tolerance, with a mean end-effector error of 4.28 ± 1.36 mm. The vision controller demonstrated robust performance across diverse target positions within the pixel workspace. The results validate the effectiveness of integrating vision-based tracking with kinematic control for precision agricultural tasks. Future work will focus on real-world implementation and the integration of force sensing for actual cutting operations.
2024
4.
Uddhav Bhattarai; Qin Zhang; Manoj Karkee
Design, integration, and field evaluation of a robotic blossom thinning system for tree fruit crops Journal Article
In: Journal of Field Robotics, vol. 41, no. 5, pp. 1366–1385, 2024, ISSN: 1556-4967, (_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/rob.22330).
Abstract | Links | BibTeX | Tags: agricultural automation, agricultural robotics, artificial intelligence in agriculture, blossom thinning, robotic thinning
@article{bhattarai_design_2024,
title = {Design, integration, and field evaluation of a robotic blossom thinning system for tree fruit crops},
author = {Uddhav Bhattarai and Qin Zhang and Manoj Karkee},
url = {https://onlinelibrary.wiley.com/doi/abs/10.1002/rob.22330},
doi = {10.1002/rob.22330},
issn = {1556-4967},
year = {2024},
date = {2024-01-01},
urldate = {2024-01-01},
journal = {Journal of Field Robotics},
volume = {41},
number = {5},
pages = {1366\textendash1385},
abstract = {The United States (US) apple industry relies heavily on semi-skilled manual labor force for essential field operations such as training, pruning, blossom and green fruitlet thinning, and harvesting. Blossom thinning is one of the crucial crop-load management practices to achieve desired crop load, fruit quality, and return bloom. While several techniques such as chemical and mechanical thinning are available for large-scale blossom thinning, such approaches often yield unpredictable thinning results and may damage the canopy, spurs, and leaf tissue. Hence, growers still depend on laborious, labor-intensive, and expensive manual hand blossom thinning for desired thinning outcomes. This research presents a robotic solution for precision blossom thinning in apple orchards using a deep learning-based computer vision system, a six-degrees-of-freedom UR5e robotic manipulator, and an electrically actuated miniature end-effector. The integrated robotic system was evaluated in a commercial apple orchard which showed promising results for targeted and selective blossom thinning. Two thinning approaches, center and boundary thinning, were investigated to evaluate the system's ability to remove varying proportions of flowers from apple flower clusters. During boundary thinning, the end-effector was actuated around the cluster boundary, while center thinning involved end-effector actuation only at the cluster centroid for a fixed duration of 2 s. Field evaluation results showed that the boundary thinning approach thinned 67.2% of flowers from the targeted clusters with a cycle time of 9.0 s per cluster, whereas the center thinning approach thinned 59.4% of flowers with a cycle time of 7.2 s per cluster. Upon further improvement for commercial adoption, the proposed system could help address problems faced by apple growers with current hand, chemical, and mechanical blossom thinning approaches.},
note = {_eprint: https://onlinelibrary.wiley.com/doi/pdf/10.1002/rob.22330},
keywords = {agricultural automation, agricultural robotics, artificial intelligence in agriculture, blossom thinning, robotic thinning},
pubstate = {published},
tppubtype = {article}
}
The United States (US) apple industry relies heavily on semi-skilled manual labor force for essential field operations such as training, pruning, blossom and green fruitlet thinning, and harvesting. Blossom thinning is one of the crucial crop-load management practices to achieve desired crop load, fruit quality, and return bloom. While several techniques such as chemical and mechanical thinning are available for large-scale blossom thinning, such approaches often yield unpredictable thinning results and may damage the canopy, spurs, and leaf tissue. Hence, growers still depend on laborious, labor-intensive, and expensive manual hand blossom thinning for desired thinning outcomes. This research presents a robotic solution for precision blossom thinning in apple orchards using a deep learning-based computer vision system, a six-degrees-of-freedom UR5e robotic manipulator, and an electrically actuated miniature end-effector. The integrated robotic system was evaluated in a commercial apple orchard which showed promising results for targeted and selective blossom thinning. Two thinning approaches, center and boundary thinning, were investigated to evaluate the system's ability to remove varying proportions of flowers from apple flower clusters. During boundary thinning, the end-effector was actuated around the cluster boundary, while center thinning involved end-effector actuation only at the cluster centroid for a fixed duration of 2 s. Field evaluation results showed that the boundary thinning approach thinned 67.2% of flowers from the targeted clusters with a cycle time of 9.0 s per cluster, whereas the center thinning approach thinned 59.4% of flowers with a cycle time of 7.2 s per cluster. Upon further improvement for commercial adoption, the proposed system could help address problems faced by apple growers with current hand, chemical, and mechanical blossom thinning approaches.
