Minimum external trapezoidal correction of perspective distortion circle center position based on machine vision

Aiming at the phenomenon that the circular perspective distortion in machine vision inspection leads to the deviation of the center of the circle, a minimum external trapezoidal correction method is proposed in machine vision. The position of the distortion circle is determined by the coordinates of the smallest external trapezoidal diagonal intersection. The coordinates of the inner hole positions with the Hough transformation method and the minimum outer rectangular method were compared. The experimental results show that the use of the minimum external trapezoid than the minimum external rectangle and hoff transform determine the relative error is significantly reduced, and the correction error using the minimum external trapezoid is stable at 1%-2%, which has higher accuracy and stability, the detection of the center position and circular perspective distortion correction provides an effective method.