CropPhenoX: high-throughput automatic extraction system for wheat seedling phenotypic traits based on software and hardware collaboration
Accurately quantifying wheat seedling phenotypic traits is crucial for genetic breeding and the development of smart agriculture. However, existing phenotypic extraction methods are difficult to meet the needs of high-throughput and high-precision detection in complex scenarios. To this end, this pa...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2025-08-01
|
| Series: | Frontiers in Plant Science |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fpls.2025.1650229/full |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Accurately quantifying wheat seedling phenotypic traits is crucial for genetic breeding and the development of smart agriculture. However, existing phenotypic extraction methods are difficult to meet the needs of high-throughput and high-precision detection in complex scenarios. To this end, this paper proposes a high-throughput automated extraction system for wheat seedling phenotypic traits based on software and hardware collaboration, CropPhenoX. In terms of hardware, an architecture integrating Siemens programmable logic controller (PLC) modules is constructed to realize intelligent scheduling of crop transportation. The stability and efficiency of data acquisition are guaranteed by coordinating and controlling lighting equipment, cameras, and photoelectric switches. Modbus transmission control protocol (TCP) is used to achieve real-time data interaction and remote monitoring. In terms of software, the Wheat-RYNet model for wheat seedling detection is proposed, which combines the detection efficiency of YOLOv5, the lightweight architecture of MobileOne, and the efficient channel attention mechanism (ECA). By designing an adaptive rotation frame detection method, the challenges brought by leaf overlap and tilt are effectively overcome. In addition, a phenotypic trait extraction platform is developed to collect high-definition images in real time. The Wheat-RYNet model was used to extract wheat seedling phenotypic traits, such as leaf length, leaf width, leaf area, plant height, leaf inclination, etc. Compared with the actual measured values, the average fitting determination coefficient reached 0.9. The test results show that CropPhenoX provides an intelligent integrated solution for crop phenotyping research, breeding analysis and field management. |
|---|---|
| ISSN: | 1664-462X |