I'm using the Hough lines transform to detect straight lines on an edges image:
lines = cv2.HoughLinesP(edge_image, rho, theta, threshold, minLineLength, maxLineGap)
However all the parameters wildly depend on the image being used.
Is there a way to automatically select "optimal" parameters for a given image, based on its properties?
The edges were previously found using cv2.Canny.
Thanks in advance