Research on the quantification and automatic classification method of Chinese cabbage plant type based on point cloud...

Area of Science:

• Agricultural Science
• Computer Science
• Plant Biology

Background:

• Accurate plant type quantification is vital for crop improvement and management.
• Current Chinese cabbage classification relies on subjective manual observation, lacking standardized metrics.
• Automated classification methods are needed for efficient crop management and breeding.

Purpose of the Study:

• To develop a rapid and accurate method for quantifying and classifying Chinese cabbage plant types.
• To establish a scientific basis for Chinese cabbage variety improvement and management.
• To overcome the limitations of manual classification in agricultural production.

Main Methods:

• Utilized point-cloud data processing and the deep learning algorithm PointNet++.
• Quantified plant type traits based on Chinese cabbage growth characteristics.
• Employed K-medoids clustering for unsupervised classification and optimized PointNet++ for supervised classification.

Main Results:

• Achieved up to 92.4% accuracy in classifying Chinese cabbage plant types.
• Obtained an average recall of 92.5% and an average F1 score of 92.3%.
• Demonstrated the effectiveness of the proposed method in automated plant type classification.

Conclusions:

• The developed method provides a scientific and unified standard for Chinese cabbage plant type classification.
• Automated classification significantly enhances crop management and breeding efficiency.
• Point-cloud data combined with deep learning offers a promising approach for plant phenotyping.