Related Experiment Video
Updated: Jul 6, 2025
08:37
Setting a Successful Sorting for Extracellular Vesicle Isolation
Published on: October 11, 2024
1.0K
A high-precision jujube disease spot detection based on SSD during the sorting process.
Zhi-Ben Yin1, Fu-Yong Liu2, Hui Geng1
1College of Information Engineering, Tarim University, Alaer, 843300, China.
Plos One
|January 5, 2024
Summary
This study introduces JujubeSSD, an automated system for precise jujube disease spot detection. The new method significantly improves accuracy and speed for agricultural applications.
Area of Science:
- Agricultural Science
- Computer Vision
- Machine Learning
Background:
- Automated grading of jujubes requires high throughput and precise disease detection.
- Current algorithms struggle with dense, varied, and poorly localized disease spots.
Purpose of the Study:
- To develop an improved automated method for accurate jujube disease spot identification.
- To enhance the precision and efficiency of disease detection in jujubes for agricultural production.
Main Methods:
- Proposed JujubeSSD method based on a single shot multi-box detector (SSD) network.
- Utilized a diverse dataset with artificial collection and data augmentation.
- Integrated deformable convolutional networks (DCNs), path aggregation feature pyramid network (PAFPN), and balanced feature pyramid (BFP) into the SSD model.
- Employed transfer learning for faster detection (0.14 s).
Main Results:
- JujubeSSD achieved a mean average precision of 97.1% (mAP@0.5).
- Demonstrated significant improvements over existing algorithms: 16.84% over YOLOv5 and 8.61% over Faster R-CNN.
- Achieved a detection time of 0.14 seconds per spot.
Conclusions:
- JujubeSSD offers superior performance for jujube disease spot detection.
- The method meets practical application requirements in agricultural production.
- The integration of advanced deep learning techniques enhances detection accuracy and speed.
