can't use SIFT on a numpy array image

Question

I open a Brain MRI scan:

path = os.path.join(r"C:\temp\generated_noisy\patients", 
lst_of_filenames[index_of_filename])
scan = nib.load(path)
data = scan.get_data()
data = data[14:168,18:205,98]

so now "data" is a numpy array (the elemnts in it are "numpy.float64"):

print(pic.shape)
print(type(pic))
print(type(pic[50,50]))

(154, 187)
<class 'numpy.ndarray'>
<class 'numpy.float64'>

Allright, now I try to use SIFT: kp,des = sift.detectAndCompute(pic,None)

and I get this error:

OpenCV(3.4.2) C:\projects\opencv- 
python\opencv_contrib\modules\xfeatures2d\src\sift.cpp:1121: error: (-5:Bad 
argument) image is empty or has incorrect depth (!=CV_8U) in function 
'cv::xfeatures2d::SIFT_Impl::detectAndCompute'

Anyone know how to solve this? I can't seem to find any help online...

A Kruger · Accepted Answer · 2018-11-30 21:33:21Z

1

SIFT expects the data type to be uint8. However, you won't be able to just convert the array to uint8 if the values are not in the correct range, such as being greater than 255 or in the range 0-1. Before converting, you could scale the values so the values are in the range 0-255. For example, the following will subtract the minimum value so the minimum is zero, then scale so the max is 255. Then you can convert to type uint8 without corrupting the image.

data = (data - np.min(data)) * 255 / np.max(data)
data = data.astype(np.uint8)

edited Nov 30, 2018 at 21:33

answered Nov 30, 2018 at 20:38

A Kruger

2,42917 silver badges18 bronze badges

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

Collectives™ on Stack Overflow

can't use SIFT on a numpy array image

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related