Related Experiment Video
Updated: Jul 1, 2025
04:48
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
2.7K
Design of smart citrus picking model based on Mask RCNN and adaptive threshold segmentation
Ziwei Guo1, Yuanwu Shi2, Ibrar Ahmad3
1Hubei University of Technology, Wuhan, China.
Peerj. Computer Science
|March 14, 2024
Summary
This study introduces an advanced deep learning model for automated citrus harvesting, achieving over 95% accuracy in identifying fruit. This innovation supports the future of smart agriculture and intelligent fruit collection.
Area of Science:
- Agricultural Technology
- Computer Vision
- Deep Learning
Background:
- Smart agriculture increasingly relies on automation and deep learning for enhanced efficiency.
- Computer vision and deep learning enable real-time agricultural monitoring, plant growth detection, and ripeness assessment.
Purpose of the Study:
- To develop an automated citrus harvesting system using a novel deep learning model.
- To improve the accuracy and efficiency of citrus fruit detection and identification in smart agriculture.
Main Methods:
- Developed an Attention-based Mask Region-based Convolutional Neural Network (ATT-MRCNN) model integrating channel and spatial attention mechanisms.
- Applied Mask Region-based Convolutional Neural Networks (Mask R-CNN) to diverse citrus image classifications.
- Utilized transfer learning for model training to enhance data performance and optimize efficiency.
Main Results:
- The ATT-MRCNN model achieved a recognition rate exceeding 95% across three sensory recognition tasks.
- The integrated attention mechanisms significantly enhanced the efficacy of citrus image detection.
Conclusions:
- The proposed ATT-MRCNN model offers robust algorithmic support for automated citrus harvesting.
- This research provides essential guidance for the advancement of intelligent harvesting systems in smart agriculture.
