How to use OpenCV ConnectedComponents to get the images

Question

How to use Python OpenCV ConnectedComponents function to obtain the images?

From searching some past question, I have only been able to find how to shade the connected objects in different colors (Which I tested and it worked, but I have no idea how the labels work)
Reference from these previously answered questions: Stackoverflow question 48303309 and Stackoverflow question 46441893

Using this code, I can get the shaded output

import cv2
import numpy as np

img = cv2.imread('eGaIy.jpg', 0)
img = cv2.threshold(img, 127, 255, cv2.THRESH_BINARY)[1]  # ensure binary
ret, labels = cv2.connectedComponents(img)

# Map component labels to hue val
label_hue = np.uint8(179*labels/np.max(labels))
blank_ch = 255*np.ones_like(label_hue)
labeled_img = cv2.merge([label_hue, blank_ch, blank_ch])

# cvt to BGR for display
labeled_img = cv2.cvtColor(labeled_img, cv2.COLOR_HSV2BGR)

# set bg label to black
labeled_img[label_hue==0] = 0

cv2.imshow('labeled.png', labeled_img)
cv2.waitKey()

Is there any way I can get the connected objects out from the image?
So output would be multiple images from the original image

Solution

image = cv2.imread('image.png', cv2.IMREAD_UNCHANGED);
gray = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
binary = cv2.threshold(gray, 0, 255, cv2.THRESH_BINARY | cv2.THRESH_OTSU)[1]

# getting mask with connectComponents
ret, labels = cv2.connectedComponents(binary)
for label in range(1,ret):
    mask = np.array(labels, dtype=np.uint8)
    mask[labels == label] = 255
    cv2.imshow('component',mask)
    cv2.waitKey(0)

# getting ROIs with findContours
contours = cv2.findContours(binary, cv2.RETR_EXTERNAL, cv2.CHAIN_APPROX_SIMPLE)[1]
for cnt in contours:
    (x,y,w,h) = cv2.boundingRect(cnt)
    ROI = image[y:y+h,x:x+w]
    cv2.imshow('ROI', ROI)
    cv2.waitKey(0)

cv2.destroyAllWindows()

Comments

[Chopin]

Thank you! It worked well. May I ask how to merge all the ROI into just 1 big ROI?

[Chopin]

Should I keep the lowest x, highest x, lowest y, highest y during the contour loop, then crop out using image[ smally: highy - smally, smallx: highx - smallx]

[jtlz2]

The first for loop can be very slow for a large image. Any ideas on how to accelerate it??

[Scott]

You have to edit your answer: contours=cv2.findContours(...)[1] to contours=cv2.findContours(...)[0]