cv2.moments for two contours

Question

The code recognizes the two contours of the specified color, however I don't know how to save the coordinates (X, Y) of each contour individually.

Although the coordinates appear in the frame, when printing the position variables it only saves one of the two coordinates.

Does anyone know how I can use cv.moments to find the moment of each contour individually or have another idea to achieve the expected goal?

    for c_2 in contour_2:
        area = cv2.contourArea(c_2)
        # Noise reduction using area as parameter
        if area > 2000:
            M_1 = cv2.moments(c_2)
            # Centroid
            if M_1["m00"] == 0: M_1["m00", "m01"] = 1
            x_2 = int(M_1["m10"] / M_1["m00"])
            y_2 = int(M_1["m01"] / M_1["m00"])

            cv2.circle(frame, (x_2, y_2), 7, (0, 255, 0), -1)

            font = cv2.FONT_HERSHEY_SIMPLEX
            cv2.putText(frame, '{},{}'.format(x_2, y_2), (x_2 + 10, y_2), font, 0.75, (0, 255, 0), 1, cv2.LINE_AA)

            # Contour smoothing
            newContour = cv2.convexHull(c)
            cv2.drawContours(frame, [newContour], 0, (255, 0, 0), 3)´´´

enter image description here

Answer 1

The output of the findContours function is a list of contours. That means you can loop through it as you are already doing, and when doing so, save any values you want for each contour in a data structure of your choice. Your current code is only saving one set of coordinates because the coordinates are updated every time you run the loop with a contour.

One way to store all the values would be creating separate lists for each measurement for each contour:

# Initialize the empty lists for each measurement
areas = []
centers = []
moments_list = []
# loop through all contours, calculate measurements and store values in appropriate list
for c in contours:
    areas.append(cv2.contourArea(c))
    M_1= cv2.moments(c)
    moments_list.append(M_1)
    if M_1["m00"] == 0: M_1["m00", "m01"] = 1
    x = int(M_1["m10"] / M_1["m00"])
    y = int(M_1["m01"] / M_1["m00"])
    centers.append((x, y))

Then you can access the measurements for each contour using the index and use the values for subsequent operations, for example, print out the measurements for a given contour:

i = 0
print(f"Contour {i} measurements: area= areas[i], center= centers[i], moments= moments_list[i]")

Example output:

>>> Contour 0 measurements: area= 1111, center= (123, 345), moments= {...}

Once you have all the measurements you could build a dataframe using the lists, to organize the values nicely. Dictionaries are an option as well, saving the data as a combination of keys:

contours_dict = {0: {'area': 1111, 'center': (123, 345), 'moments': {...},...}, 
                 1: {'area': 2222, 'center': (234, 567), 'moments': {...},...},
                 2: {'area': 3333, 'center': (345, 678), 'moments': {...},...}

That is closer to a dataframe structure without the need to use an external library.