pythonopencvcontouropencv3.0

unable to get contour properly

I am playing with openCV and doing simple below task.

1) Read Image

2) thresholding

3) Finding contours.

4) Drawing all contours in the blank image.

5) Drawing individual contour.

Drawing all contours on the dummy image looks good, whereas drawing single contour produces scattered contour as shown in the below images.

Original:

enter image description here

All Contours :

enter image description here

Single Contour :

enter image description here

Please find the code below. 

import  cv2
import numpy as np

#Reading Image.
srcImg = cv2.imread("./bottle.jpeg")

#Color Conversion.

grayedImg = cv2.cvtColor(srcImg,cv2.COLOR_RGB2GRAY)
__, thresholdedImg = cv2.threshold(grayedImg, 240, 255, cv2.THRESH_BINARY)

#Noice Removal
kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (3,3))
erodeImage = cv2.erode(thresholdedImg,kernel, iterations=1)
dilatedImg = cv2.dilate(erodeImage,kernel, iterations=1)
_, contours, _ = cv2.findContours(dilatedImg,cv2.RETR_TREE,cv2.CHAIN_APPROX_SIMPLE)

#draw All Contours.

dummyImg = np.zeros(grayedImg.shape, dtype=grayedImg.dtype)
cv2.drawContours(dummyImg, contours, -1, 255, 1)
cv2.imshow("All Contours", dummyImg)
cv2.imwrite("allContours.jpeg",dummyImg)

#draw Individual Contours.

mask =  np.zeros(dummyImg.shape[:2], dtype= dummyImg.dtype)
isolatedImg = cv2.drawContours(mask, contours[9], -1, 255, 1)

cv2.imshow("Indivial Contours.", isolatedImg)
cv2.imwrite("single.jpeg",isolatedImg)

cv2.waitKey(0)
Solution

  • You have to enclose another set of square brackets:

    isolatedImg = cv2.drawContours(mask, [contours[9]], -1, 255, 1)

    Expected result:

    enter image description here

    If you dig deep, cv2.findContours() returns a list of arrays. Now each array contains details on the number of points making up the contour.