A Novel Method for Space Circular Target Detection in Machine Vision

Area of Science:

• Computer Vision
• 3D Reconstruction
• Robotics

Background:

• Space circular target detection is crucial for visual measurement and object detection.
• Traditional monocular vision methods for 3D circular target localization are sensitive to image ellipse parameters and prone to interference.
• Existing methods struggle with noise resistance in practical applications.

Purpose of the Study:

• To develop an optimized method for fitting circular targets in 3D space.
• To improve the anti-interference ability and noise resistance of 3D circular target detection.
• To provide a more robust and valuable method for applications in visual measurement and object detection.

Main Methods:

• Projecting the image ellipse of a space circular target back into 3D space.
• Developing an optimization method for fitting the 3D circular target.
• Detecting the center coordinate and normal vector of the space circular target in 3D.

Main Results:

• The proposed method demonstrates enhanced noise resistance compared to traditional approaches.
• The method is less sensitive to variations in image ellipse parameters.
• Both simulation and practical experiments validated the feasibility and effectiveness of the new technique.

Conclusions:

• The developed optimization method offers a more robust solution for 3D circular target detection.
• This approach significantly improves upon the limitations of traditional methods, particularly in noisy environments.
• The technique holds notable application value in fields requiring precise visual measurement and object detection.