AI Pool
Log in
Discussions>Cannot display magnitude of gradient in form of image>

Cannot display magnitude of gradient in form of image

I am trying to run this code. I took the blurred image and applied Sobel's kernel on it. kernel applied successfully like I am getting the vertical and horizontal components on the output but when I try to show the gradient it is giving an error. I have tried many ways but I am not getting a solution.

import numpy as np
import cv2 as cv
from scipy import ndimage

img=cv.imread("scarlet_guassian_blur.jpg",cv.IMREAD_UNCHANGED)

Kx = np.array([[-1, 0, 1], [-2, 0, 2], [-1, 0, 1]], np.float32)
Ky = np.array([[1, 2, 1], [0, 0, 0], [-1, -2, -1]], np.float32)

Ix = ndimage.filters.convolve(img, Kx)
Iy = ndimage.filters.convolve(img, Ky)

print("SobelX_Vertical\n",Ix)
print("SobelY_Horizontal\n",Iy)

cv.imshow("Gray_Scale",img)
cv.imshow("SobelX_Vertical",Ix)
cv.imshow("SobelY_Horizontal",Iy)

hyp=(Ix*Ix) + (Iy*Iy)
G=np.sqrt([hyp])

gradient = G / G.max() * 255
print("Gradient\n",gradient)

phase = np.arctan2(Iy, Ix)
print("Phase\n",phase)

cv.imshow("Gradient",gradient)
cv.waitKey()
cv.destroyAllWindows()

It is giving the following output in my terminal

SobelX_Vertical
 [[  0   0   0 ...   0   0   0]
 [  0   0   0 ...   0   0   0]
 [  0   0   0 ...   0   0   0]
 ...
 [  2   2 254 ... 254 255   0]
 [  3   4 254 ... 250 253   0]
 [  1   4 255 ... 248 252   0]]
SobelY_Horizontal
 [[ 0  0  0 ...  0  0  0]
 [ 0  0  0 ...  0  0  0]
 [ 0  0  0 ...  0  0  0]
 ...
 [ 6  2  0 ... 12 15 16]
 [ 1  2  2 ...  8 11 12]
 [ 1  2  1 ...  4  4  4]]
Gradient
 [[[  0.     0.     0.   ...   0.     0.     0.  ]
  [  0.     0.     0.   ...   0.     0.     0.  ]
  [  0.     0.     0.   ...   0.     0.     0.  ]
  ...
  [102.    45.6   32.25 ... 196.1  242.4    0.  ]
  [ 51.    72.1   45.6  ... 161.2  183.8  193.5 ]
  [ 22.8   72.1   22.8  ... 144.2   91.2   64.5 ]]]
Phase
 [[0.      0.      0.      ... 0.      0.      0.     ]
 [0.      0.      0.      ... 0.      0.      0.     ]
 [0.      0.      0.      ... 0.      0.      0.     ]
 ...
 [1.249   0.785   0.      ... 0.0472  0.05875 1.57   ]
 [0.3218  0.4636  0.00787 ... 0.03198 0.04346 1.57   ]
 [0.785   0.4636  0.00392 ... 0.01613 0.01587 1.57   ]]

---------------------------------------------------------------------------TypeError Traceback (most recent call last)
<ipython-input-3-84576290db15> in <module>()     23 phase = np.arctan2(Iy, Ix)     
24 print("Phase\n",phase)---> 25 cv.imshow("Gradient",gradient)     26 cv.waitKey()     
27 cv.destroyAllWindows()TypeError: Expected Ptr<cv::UMat> for argument 'mat'
Computer Vision
opencvsobel
3 votesUG3.00
1 Answers
JO246.00
3

I saw your gradient's type is float16. I don't know what is the exact problem, but when I converted it to float32 it worked end showed the result.
I just run your code on the binary mask (using copy-paste) and visualize it with matplotlib.

import matplotlib.pyplot as plt
# ....
plt.imshow(gradient[0].astype(np.float32), cmap='gray')
plt.show()

It works fine and here is the result of a binary mask example.

Reply
Couldn't find what you were looking for?and we will find an expert to answer.
How helpful was this page?