Pose Estimation of Sweet Pepper through Symmetry Axis Detection

Area of Science:

• Robotics and Automation
• Computer Vision
• Agricultural Engineering

Background:

• Accurate fruit pose estimation is vital for automated harvesting systems.
• Sweet pepper detachment requires precise spatial orientation data.

Purpose of the Study:

• To develop and evaluate a novel method for sweet pepper pose estimation.
• To enable precise spatial orientation determination for robotic detachment.

Main Methods:

• Calculating point cloud normals and identifying candidate planes.
• Scoring candidate planes to find the symmetry plane.
• Deriving the symmetry axis to determine sweet pepper pose.

Main Results:

• Simulated tests showed an average angle error of approximately 6.5°.
• Real sweet pepper tests yielded an average error of 7.4° (peduncle removed) and 6.9° (peduncle intact).

Conclusions:

• The proposed method is effective for sweet pepper pose estimation.
• The technique is adaptable for pose estimation in other fruits and vegetables.