how to get only contour area color range

Question

I want to get contour area color range and do some condition for example this is input image: Below is the code to find contours:

import cv2

img = cv2.imread('D:/original.png', cv2.IMREAD_UNCHANGED)

#convert img to grey
img_grey = cv2.cvtColor(img,cv2.COLOR_BGR2GRAY)
#set a thresh
thresh = 100
#get threshold image
ret,thresh_img = cv2.threshold(img_grey, thresh, 255, cv2.THRESH_BINARY)
#find contours
contours, hierarchy = cv2.findContours(thresh_img, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)

Now I am trying to get each contour color and write condition for example:

if contours[0] in color range ((100,100,100),(200,200,200)) then drawContour

All the things I'm trying to do are: get each contour area and check if selected contour is in specific color range or not.

Answer 1

In order to extract single contours differentiated by color, the logical way to approach this problem would be to use the different colors and not convert the image to grayscale .

You can than work on single channels. For instance, for the the blue channel:

thresh = 100

ret,thresh_img = cv2.threshold(b, thresh, 255, cv2.THRESH_BINARY)

Then with a combination of bit_wise operations, you can extract a specific contour.

Another approach would be to replace the threshold operator with the Canny operator.

thresh1 = 40
thresh2 = 120
#get canny image
thresh_img = cv2.Canny(img[:,:,0], thresh1,thresh2)
#find contours
contours, hierarchy = cv2.findContours(thresh_img, cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)

Which yields the following contours:

The use of Canny as preprocessing for contour is suggested by the OpenCV documentation .

Answer 2

We may start with cv2.kmeans for performing color clustering - kind of what described in the following tutorial .

The result is a list of labels.
Each label (label 0, label 1,...) represents all the pixels that belongs to a specific color cluster.

Example of applying K-Means for colors clustering:

# Reshape the image into a 2D array with one row per pixel and three columns for the color channels.
data = image.reshape((cols * rows, 3))

# Perform K-Means clustering on the image.
criteria = (cv2.TERM_CRITERIA_EPS + cv2.TERM_CRITERIA_MAX_ITER, 10, 1.0)
_, labels, centroids = cv2.kmeans(data, K, None, criteria, 10, cv2.KMEANS_RANDOM_CENTERS)

---

Iterate the labels, and create a mask with 255 where labels == k :

mask = np.zeros((rows*cols, 1), np.uint8)  # Create a zerod mask in the size of image.
mask[labels == k] = 255  # Place 255 where labels == k (all labels equals 0 are going to be 255 then all labels equals 1...)
mask = mask.reshape((rows, cols))  # Reshape the mask back to the size of the image.

---

For each mask apply the following stages:

• Find Contours.
• For each contour, find contour area, and ignore relatively small contours.
• Get the first coordinate in the contour: x, y = tuple(c[0][0]) .
• Get the color of the pixel in that coordinate: color = original_image[y, x] .
• Draw the contour with that color (mainly for testing): cv2.drawContours(colored_mask, [c], 0, color.tolist(), -1) .

---

Complete code sample:

import cv2
import numpy as np

K = 16  # Number of color clusters (16 is a bit larger than the accrual number of colors).

# Load the image and convert it float32 (kmeans requires float32 type).
original_image = cv2.imread('original.png')
image = original_image.astype(np.float32)

cols, rows = image.shape[1], image.shape[0]

# Reshape the image into a 2D array with one row per pixel and three columns for the color channels.
data = image.reshape((cols * rows, 3))

# Perform K-Means clustering on the image.
criteria = (cv2.TERM_CRITERIA_EPS + cv2.TERM_CRITERIA_MAX_ITER, 10, 1.0)
_, labels, centroids = cv2.kmeans(data, K, None, criteria, 10, cv2.KMEANS_RANDOM_CENTERS)

# Convert the labels back into an image (for testing).
quantized_image = centroids[labels].reshape(image.shape).astype(np.uint8)

# Save the quantized_image image (for testing).
cv2.imwrite('quantized_image.png', quantized_image)

for k in range(K):
    mask = np.zeros((rows*cols, 1), np.uint8)  # Create a zeroed mask in the size of image.
    mask[labels == k] = 255  # Place 255 where labels == k (all labels equals 0 are going to be 255 then all labels equals 1...)
    mask = mask.reshape((rows, cols))  # Reshape the mask back to the size of the image.
    #cv2.imshow(f'mask {k}', mask)  # Show mask for testing

    cnts = cv2.findContours(mask, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_NONE)[0]  # Find contours
    for c in cnts:
        area_tresh = 500
        area = cv2.contourArea(c)
        if area > area_tresh:  # Ignore relatively small contours
            colored_mask = np.zeros_like(original_image)  # Initialize colored_mask with zeros
            x, y = tuple(c[0][0])  # First coordinate in the contour
            color = original_image[y, x]  # Get the color of the pixel in that coordinate
            cv2.drawContours(colored_mask, [c], 0, color.tolist(), -1)  # Draw contour with the specific color
            cv2.imshow(f'colored_mask {k}', colored_mask)  # Show colored_mask for testing
            cv2.imwrite(f'colored_mask_{k}.png', colored_mask)  # Save as PNG for testing
            continue  # Assume only the first large contour is relevant - continue to next iteration

cv2.waitKey()
cv2.destroyAllWindows()

---

Few output samples: