Create numpy array (contour) to be used with opencv functions

Question

I'm new to python and I don't know how to create numpy array that may be used in opencv functions. I've got two vectors defined as follows:

X=np.array(x_list)
Y=np.array(y_list)

and the result is:

[ 250.78  250.23  249.67 ...,  251.89  251.34  250.78]
[ 251.89  251.89  252.45 ...,  248.56  248.56  251.89]

I want to create opencv contour to be used in ex. cv2.contourArea(contour). I read Checking contour area in opencv using python but cannot write my contour numpy array properly. What's the best way to do that?

It looks like cv2 contours are 3-dimensional numpy arrays. If you test out contour.shape you'll be able to work out it's dimensions. If you want to write a compatible numpy array it will need to have 3 dimensions. For example numpy.zeros(1,2,3) will create a 3D array of zeros of shape 1x2x3... Try testing that out! — Sam
– Sam, Commented Jul 27, 2016 at 10:09

Sam · Accepted Answer · 2016-07-27 10:16:35Z

2

Here's some example code, that first checks the dimensions of a contour calculated from a test image, and makes a test array and has success with that as well. I hope this help you!

import cv2
import numpy as np

img = cv2.imread('output6.png',0) #read in a test image
ret,thresh = cv2.threshold(img,127,255,0)
im2,contours,hierarchy = cv2.findContours(thresh, 1, 2)

cnt = contours[0]

print cnt.shape #this contour is a 3D numpy array
print cv2.contourArea(cnt) #the function prints out the area happily

#######Below is the bit you asked about

contour = np.array([[[0,0]], [[10,0]], [[10,10]], [[5,4]]]) #make a fake array
print cv2.contourArea(contour) #also compatible with function

answered Jul 27, 2016 at 10:16

Sam

1,0624 gold badges14 silver badges33 bronze badges

Sign up to request clarification or add additional context in comments.

2 Comments

Biba Over a year ago

Thanks for your answer! It is very usefull, altough my problem is that my input is not an image but contour points of a structure read from the file. Based on that i create binary image where contour points are ones, and background is 0. Applying cv2.findContours on that binary image results in finding many contours, but i know that there only should be one. That's why i wanted to artificially create the contour from the points, but i don't know how to put them into the structure like this: contour = np.array([[[0,0]], [[10,0]], [[10,10]], [[5,4]]])

Juan Zamora Over a year ago

this does not answer the question. this finds contours from an image.

Collectives™ on Stack Overflow

Create numpy array (contour) to be used with opencv functions

1 Answer 1

2 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

2 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related