Related Experiment Video
Updated: Jun 7, 2025
08:47
Author Spotlight: UAV Remote Sensing for Efficient Invasive Plant Biomass Estimation
Published on: February 9, 2024
1.3K
Estimation of Strawberry Canopy Volume in Unmanned Aerial Vehicle RGB Imagery Using an Object Detection-Based
Min-Seok Gang1,2, Thanyachanok Sutthanonkul3, Won Suk Lee3
1Department of Biosystems Engineering, College of Agriculture and Life Sciences, Seoul National University, Seoul 08826, Republic of Korea.
Sensors (Basel, Switzerland)
|November 9, 2024
Summary
Accurately estimating strawberry canopy volume using a convolutional neural network (CNN) with unmanned aerial vehicle (UAV) imagery improves yield prediction. This advanced method offers significant gains over traditional approximations for precision agriculture.
Area of Science:
- Agricultural Engineering
- Computer Vision
- Remote Sensing
Background:
- Accurate strawberry canopy volume estimation is vital for yield prediction and precision agriculture.
- Manual unmanned aerial vehicle (UAV) flights with RGB cameras are increasingly used for crop monitoring.
- Existing methods for canopy volume estimation often lack precision and struggle with spatial variability.
Purpose of the Study:
- To develop and evaluate a convolutional neural network (CNN) model for precise strawberry canopy volume estimation.
- To assess the spatial variability of strawberry canopy volumes within a field.
- To improve upon conventional methods for canopy volume approximation.
Main Methods:
- Utilized a ResNet50V2-based CNN model trained on RGB images from UAV flights.
- Implemented a You Only Look Once v8 Nano (YOLOv8n) object detection model for image distortion correction.
- Trained the CNN model using actual canopy volumes measured with expanded polystyrene (EPS) balls to account for internal plant structure.
Main Results:
- The CNN model, incorporating flight altitude compensation, demonstrated a strong correlation (R²=0.98) with measured canopy volumes.
- Achieved a 84% improvement in accuracy compared to the conventional paraboloid shape approximation.
- Generated a canopy volume map revealing significant spatial variability crucial for site-specific management.
Conclusions:
- The developed CNN model provides a highly accurate and reliable method for estimating strawberry canopy volumes.
- The approach effectively addresses image distortions and internal plant structures for improved volume estimation.
- The generated canopy volume maps enable targeted agricultural management strategies for enhanced strawberry production.
