Advances in Robotics and Machine Learning for Litchi Harvesting

Strategic Insights -

Researchers at Kunming University in China have developed a new method for detecting and classifying the ripeness of litchi bunches, a crucial task for robotic harvesting. The method combines deep learning, image processing, and clustering algorithms to achieve high accuracy and efficiency. The study's findings have significant implications for the development of orchard harvesting robot systems.

Key Takeaways:

  • The proposed method, called Litchi-YOLO, achieved a precision of 95.96%, recall of 95.69%, and F1-score of 95.82% in detecting litchi bunches, representing improvements of 1.25%, 6.97%, and 4.25% over YOLOv8.
  • The KGAP-DBSCAN clustering algorithm achieved homogeneity, completeness, and vmeasure scores of 0.91, 0.76, and 0.78, respectively, for clustering fruit coordinate points.
  • The ripeness grading method for individual fruits demonstrated good performance, with a precision of 94.20% and recall of 91.91%.
  • The study assessed the maturity of litchi bunches in a natural environment, assisting the orchard harvesting robot system in determining harvesting decisions.
  • The research was supported by the National Natural Science Foundation of China (NSFC) and the Guangdong Basic and Applied Basic Research Foundation.
  • The study concluded that the developed method can be used for efficient and non-destructive picking of litchi bunches.

Statistics:

  • The proposed method achieved an accuracy of 95.96% in detecting litchi bunches.
  • The KGAP-DBSCAN clustering algorithm achieved a homogeneity score of 0.91.
  • The ripeness grading method for individual fruits demonstrated a precision of 94.20%.
  • The research was supported by a grant amount of CN¥ [ amount not specified ] from the National Natural Science Foundation of China (NSFC).
  • The study assessed a total of 1,000 litchi bunches in a natural environment.

Sources:

  • Litchi Bunch Detection and Ripeness Assessment Using Deep Learning and Clustering With Image Processing Techniques. Biosystems Engineering, 2025;255.
  • NewsRx. Studies in the Area of Robotics and Machine Learning Reported from Kunming University (Litchi Bunch Detection and Ripeness Assessment Using Deep Learning and Clustering With Image Processing Techniques). Journal of Engineering. July 14, 2025; p 3959.