使用python从opencv中的findcontours检测到的单个轮廓中提取坐标

Question

Here my code to store (x,y) in two found by cv2.findcontours in OpenCV python.

I have selected a random contour

c=contour[6]

Now I want x an y values detected in a separate array to perform some operation

numpy array are stored in this fashion.

[[[ 746  997]]
 [[ 744  998]]
 [[ 742  999]]
 [[ 740 1000]]]

I tried using this to extract x values

x = c[:,[0]]

but I'm getting the same array back.

So I try to extract using this loop

    for a in c:
      for b in a:
        s_x = np.append(s_x, b[0])
        s_y = np.append(s_y, b[1])

Is there a simple way to select x coordinates, than going through the loop and not having this error at all?

Answer 1

看来您的数组有额外的维度，因此您可以删除它，然后建立索引。

x = c.squeeze()[:, 0]

Answer 2

cnt = cnts[1]             # choose one 
cnt = cnt.reshape(-1,2)   # change the shape 
xs  = cnt[:,0]            # get xs

Answer 3

Your array has three dimensions. The details can be seen with the shape attribute.

import numpy as np

c = np.array([[[ 746,  997]],
              [[ 744,  998]],
              [[ 742,  999]],
              [[ 740, 1000]]])
print(c.shape)

(4,1,2)

Once you have the shape, slice the array.

x = c[:,0,0]
print(x)

[746 744 742 740]