Related Experiment Video
Updated: May 12, 2026
06:00
Electroantennography-based Bio-hybrid Odor-detecting Drone using Silkmoth Antennae for Odor Source Localization
Published on: August 27, 2021
5.2K
YOLOv8s-Longan: a lightweight detection method for the longan fruit-picking UAV
Jun Li1,2,3, Kaixuan Wu1, Meiqi Zhang1
1College of Engineering, South China Agricultural University, Guangzhou, China.
Frontiers in Plant Science
|February 6, 2025
Summary
A new lightweight deep learning algorithm, YOLOv8s-Longan, enhances fruit detection for unmanned aerial vehicles (UAVs). This AI model improves accuracy by 3.9% while reducing parameters by 20.3%, enabling faster and more precise fruit picking.
Area of Science:
- Computer Vision
- Artificial Intelligence
- Robotics
Background:
- Unmanned aerial vehicles (UAVs) require efficient algorithms for rapid fruit detection due to limited onboard computing power and high flight speeds.
- Accurate fruit localization is crucial for automated fruit-picking operations.
Purpose of the Study:
- To develop a lightweight deep learning algorithm (YOLOv8s-Longan) for enhanced fruit detection in UAV applications.
- To improve detection accuracy and reduce model parameters for real-time fruit picking.
Main Methods:
- Integration of an Average and Max pooling attention (AMA) module into DenseAMA and C2f-Faster-AMA modules for network lightweighting and generalization.
- Implementation of a VOVGSCSPC module for multiscale feature fusion to enhance image understanding.
- Adoption of a novel Inner-SIoU loss function for improved target bounding box regression.
Main Results:
- The YOLOv8s-Longan algorithm achieved a mean Average Precision (mAP@0.5) of 84.3% for detecting densely packed and occluded longan fruit in complex environments.
- Demonstrated a 3.9% improvement in mAP@0.5 compared to other YOLOv8 models.
- Achieved a 20.3% reduction in model parameters, contributing to faster processing.
Conclusions:
- The YOLOv8s-Longan algorithm meets the high accuracy and speed requirements for fruit detection in UAV-based picking systems.
- The proposed lightweight model offers a practical solution for real-time fruit identification in agricultural robotics.
Related Concept Videos
02:16Light Acquisition
In order to produce glucose, plants need to capture sufficient light energy. Many modern plants have evolved leaves specialized for light acquisition. Leaves can be only millimeters in width or tens of meters wide, depending on the environment. Due to competition for sunlight, evolution has driven the evolution of increasingly larger leaves and taller plants, to avoid shading by their neighbors with contaminant elaboration of root architecture and mechanisms to transport water and nutrients.
01:29Application of Linearization and Approximation
A drone flying through complex terrain often relies on more than one sensing method to estimate small changes in altitude. Along with direct measurements, air pressure provides a useful indirect indicator of vertical movement. Atmospheric pressure decreases as altitude increases, and this relationship is commonly described using an exponential model. Although accurate, converting pressure measurements into altitude values requires calculations that are too complex to perform repeatedly during...