Automatic hierarchy classification in venation networks using directional morphological filtering for hierarchical...

Area of Science:

• Botany
• Evolutionary Biology
• Image Analysis

Background:

• Extracting vein traits from plant venation networks is vital for research, but traditional methods focus on reticulate structure traits (ReSTs), insufficient for distinguishing vein orders.
• Existing tools for hierarchical structure traits (HiSTs) often require human assistance, limiting their scalability for large-scale analysis.

Purpose of the Study:

• To develop an automated method for classifying vein orders within plant venation networks.
• To enable the extraction of ecologically significant traits from complex, dense venation patterns.

Main Methods:

• A novel vein hierarchy classification method utilizing directional morphological filtering was developed.
• The method automatically classifies vein orders, distinguishing it from traditional approaches.

Main Results:

• The proposed method demonstrates high performance in classifying vein orders on a soybean dataset (STID).
• Achieved an average deviation of less than 5 pixels for major veins and 54.28% average completeness for second-order veins.
• This represents the first known attempt at automatic vein hierarchy classification.

Conclusions:

• The developed method offers an effective solution for automated vein hierarchy classification.
• It facilitates the extraction of valuable traits from dense venation networks for ecological and evolutionary studies.