如何改进圆检测

Question

I am trying to detect certain circles in this image:

This is the best result I have succeeded:

You can see there are 4 circles that it detected that I wasn't trying to detect and 1 circle that it missed.

Here is the code i have used:

def draw_circles(img, circles):
    cimg = cv2.cvtColor(img,cv2.COLOR_GRAY2BGR)
    for i in circles[0,:]:
    # draw the outer circle
        cv2.circle(cimg,(i[0],i[1]),i[2],(0,255,0),2)
        # draw the center of the circle
        cv2.circle(cimg,(i[0],i[1]),2,(0,0,255),3)
        cv2.putText(cimg,str(i[0])+str(',')+str(i[1]), (i[0],i[1]), cv2.FONT_HERSHEY_SIMPLEX, 0.4, 255)
    return cimg

def detect_circles(image_path):
    gray = cv2.imread(image_path, 0)
    gray_blur = cv2.medianBlur(gray, 13)  # Remove noise before laplacian
    gray_lap = cv2.Laplacian(gray_blur, cv2.CV_8UC1, ksize=5)
    dilate_lap = cv2.dilate(gray_lap, (3, 3))  # Fill in gaps from blurring. This helps to detect circles with broken edges.
    # Furture remove noise introduced by laplacian. This removes false pos in space between the two groups of circles.
    lap_blur = cv2.bilateralFilter(dilate_lap, 5, 9, 9)
    # Fix the resolution to 16. This helps it find more circles. Also, set distance between circles to 55 by measuring dist in image.
    # Minimum radius and max radius are also set by examining the image.
    circles = cv2.HoughCircles(lap_blur, cv2.HOUGH_GRADIENT, 16, 80, param2=450, minRadius=20, maxRadius=40)
    cimg = draw_circles(gray, circles)
    print("{} circles detected.".format(circles[0].shape[0]))
    # There are some false positives left in the regions containing the numbers.
    # They can be filtered out based on their y-coordinates if your images are aligned to a canonical axis.
    # I'll leave that to you.
    return cimg

plt.imshow(detect_circles('test.jpeg'))

I have tried playing with the min and max radius params but without real success, any suggestions would be helpful.

Answer 1

I don't see why you would need all this pre-processing. You have perfect circles in your image, you know their exact radius...

All your pre-processing is doing is to render good input into poor input. You remove the nice circle outlines you have and then try to recreate them from what is left after medianing your image to death :)

def detect_circles(image_path):
    gray = cv2.imread(image_path, 0)
    circles = cv2.HoughCircles(gray, cv2.HOUGH_GRADIENT, 2, 60, param2=100, minRadius=20, maxRadius=40)
    cimg = draw_circles(gray, circles)
    print("{} circles detected.".format(circles[0].shape[0]))
    return cimg

I think you should really research how those algorithms work. You're code looks like you blindly applied stuff you found online.

Answer 2

maybe this solution will help.

def draw_circles(img, circles):
    cimg = img.copy()
    for i in circles[0,:]:
    # draw the outer circle
        cv2.circle(cimg,(i[0],i[1]),i[2],(0,255,0),2)
        # draw the center of the circle
        cv2.circle(cimg,(i[0],i[1]),2,(0,0,255),3)
        cv2.putText(cimg,str(i[0])+str(',')+str(i[1]), (i[0],i[1]), cv2.FONT_HERSHEY_SIMPLEX, 0.4, 255)
    return cimg

def detect_circles(image_path):
    # open color image
    image = cv2.imread(image_path)
    image_blur = cv2.medianBlur(image, 3)
    # find edges
    edges = cv2.Canny(image_blur, 5, 50)
    # let's clean the neighboring pixels
    edges = cv2.medianBlur(edges, 3)

    circles = cv2.HoughCircles(edges, cv2.HOUGH_GRADIENT, 2, 80, param1=100, param2=45, minRadius=20, maxRadius=40)
    cimg = draw_circles(image, circles)
    print("{} circles detected.".format(circles[0].shape[0]))
    return cimg

The result of the code above