Implementation of Simple Hough Circle Detection Algorithm in Python.
This is based on paper Use of the Hough Transformation To Detect Lines and Curves in Pictures by Richard O. Duda and Peter E. Hart.
This is an extension of the Hough Transform to detect circles using the equation,
r^2 = ( x - a )^2 + ( y - b )^2 in parameter space rho = ( a, b, r)
Please refer to Hough Line Detection python implementation at the following git repository for more information.
''' python find_hough_circles.py ./images/ex1.png --r_min 10 --r_max 200 --delta_r 1 --num_thetas 100 --bin_threshold 0.4 --min_edge_threshold 100 --max_edge_threshold 200 '''
The script requires one positional argument and few optional parameters:
- image_path - Complete path to the image file for circle detection.
- r_min - Min radius circle to detect. Default is 10.
- r_max - Max radius circle to detect. Default is 200.
- delta_r - Delta change in radius from r_min to r_max. Default is 1.
- num_thetas - Number of steps for theta from 0 to 2PI. Default is 100.
- bin_threshold - Thresholding value in percentage to shortlist candidate for circle. Default is 0.4 i.e. 40%.
- min_edge_threshold - Minimum threshold value for edge detection. Default 100.
- max_edge_threshold - Maximum threshold value for edge detection. Default 200.
The output of the script would be two files:
- circles.txt - List of circles in format (x,y,r,votes)
- circle_img.png - Image with the Circles drawn in Green color.
| Sample Input Image | Sample Output Image | Post-Processed Output Image |
|---|---|---|
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |
 |