Image Edge Detection with python

Question

I'd like to recognize 6 circles in the image below.

When I tried using cv2.canny and HoughCircles, it didn't recognize the circle well.

Do I need to modify the parameters? Or is there any other good way?

The results I want are as follows.

Answer 1

As others mentioned, the usage of the HoughCircles is to be taken with a grain of salt. The shapes mentioned look more like ellipsoids than circles.

With that being said, here is what I got by going through a standard procedure:

im = cv2.imread("test_circles.jpg")
im = cv2.cvtColor(im, cv2.COLOR_BGR2RGB) # convert to RGB, I use this so I can plot with matplotlib, not very important, just make sure to stick with one colourspace
gray = cv2.cvtColor(im, cv2.COLOR_BGR2GRAY) # convert to grayscale
blurred = cv2.GaussianBlur(gray, (3, 3), 2) # reduce noise a bit
circles = cv2.HoughCircles(blurred, cv2.HOUGH_GRADIENT, dp=1, minDist=20, 
                           param1=50, param2=30, minRadius=10, maxRadius=50) # get the circles
if circles is not None:
    circles = np.uint16(np.around(circles))
    for i in circles[0, :]:
        if 5 < i[2] < 20: # check the radius of the circle, ajust as needed
            # circle itself
            cv2.circle(im, (i[0], i[1]), i[2], (0, 255, 0), 2)
            # center of the image...
            cv2.circle(im, (i[0], i[1]), 2, (0, 0, 255), 3)

plt.imshow(im)
plt.show()

The results are like this:

Though not the best of approaches, since I needed to cut one more circle that was detected to the right of the mentioned ones. Try it yourself and adjust the parameters as needed.