A method for recognizing positive and negative electrodes of buzzer circuit board based on machine vision

Area of Science:

• Robotics and Automation
• Computer Vision
• Electronic Manufacturing

Background:

• Automating coil soldering on buzzer circuit boards presents challenges due to the left-right symmetry and morphological similarity of positive and negative electrodes.
• Existing methods struggle to reliably distinguish these electrodes based solely on visual features.

Purpose of the Study:

• To develop and validate a machine vision method for accurate recognition and localization of positive and negative electrodes on buzzer circuit boards.
• To facilitate automated soldering processes by providing a robust electrode identification solution.

Main Methods:

• Utilized the R-G-B color difference operator to extract color feature maps.
• Employed filtering, threshold segmentation, and contour detection to extract circuit board contours and obtain rotated minimum bounding rectangles for precise locating.
• Leveraged up-down asymmetric geometric characteristics and categorized circuit board orientation into eight cases with specific position calculation formulas for electrode recognition.

Main Results:

• Achieved high recognition accuracy with errors in minimum bounding rectangle extraction within single-digit pixel values (0-10 pixels).
• Demonstrated an average relative error of less than 2% in recognizing positive and negative electrode positions.
• Confirmed robust recognition and locating performance through experimental validation.

Conclusions:

• The proposed machine vision method effectively distinguishes and locates positive and negative electrodes on buzzer circuit boards, overcoming challenges posed by symmetry and visual similarity.
• The method shows significant potential for integration into automatic soldering equipment, enabling efficient and accurate automated soldering of buzzer components.