Related Experiment Video
Updated: Jun 23, 2026
07:12
High-Throughput, In-Field Screening of Photosynthetic Efficiency in Crop Plants Using an Autonomous Robot
Published on: January 9, 2026
CFPR-YOLO: chili flower pose estimation for robotic pollination in unstructured environments.
Minqiu Kuang1,2, Yushi Wang1, Xiaojian Li1
1College of Electrical and Mechanical Engineering, Hunan Agricultural University, Changsha, China.
Frontiers in Plant Science
|June 22, 2026
Summary
This study introduces CFPR-YOLO, a new AI vision system for detecting chili flowers. It improves precision pollination and fruit management in agriculture by enabling real-time, accurate flower recognition in challenging conditions.
Area of Science:
- Agricultural Engineering
- Computer Vision
- Horticultural Science
Background:
- Intelligent agricultural systems require accurate visual perception of crops.
- Chili flower detection is crucial for tasks like monitoring and precision pollination.
- Challenges include small size, occlusion, and variable lighting.
Purpose of the Study:
- To develop a robust and lightweight edge vision framework for chili flower detection and pose-aware perception.
- To address the limitations of existing methods in complex agricultural environments.
Main Methods:
- Proposed CFPR-YOLO framework based on YOLOv11n architecture.
- Incorporated EfficientFormerV2 for feature extraction and C3k2_EMA for target localization.
- Utilized Poly-Scale Convolution (PSConv) and a lightweight attention mechanism.
Main Results:
- Achieved 92.6% precision, 86.8% recall, and 92.1% mAP50 with 7.26 M parameters.
- Demonstrated real-time inference (39.5 FPS) on edge devices.
- Showcased robustness and generalization across chili varieties.
Conclusions:
- CFPR-YOLO offers an effective visual perception solution for agricultural engineering informatics.
- The framework has practical potential for precision pollination and intelligent fruit-set management.
