Enhancing Crop Lodging Segmentation with Deep Learning
Author Information
Author(s): Panli Zhang, Longhui Niu, Mengchen Cai, Hongxu Chen, Xiaobo Sun
Primary Institution: Northeast Agricultural University
Hypothesis
Can intelligent optimization algorithms improve the accuracy and efficiency of crop lodging segmentation using deep learning?
Conclusion
The AAUConvNeXt model significantly improves the accuracy of segmenting rice lodging areas while reducing computational costs.
Supporting Evidence
- The AAUConvNeXt model achieved the highest pixel accuracy of 96.3% among compared models.
- The model reduced the number of parameters and computational load while maintaining high accuracy.
- Optimizing the number of channels in convolutional layers improved feature extraction capabilities.
Takeaway
This study created a smart computer program that helps farmers see when their rice plants are falling over, which can help them grow better crops.
Methodology
The study used a convolutional neural network architecture optimized with intelligent algorithms to segment crop lodging from UAV imagery.
Limitations
The model may struggle with subtle differences in continuous lodging states and relies on high-quality data annotation.
Participant Demographics
The study focused on rice crops in Heilongjiang Province, China.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website