Ipython cv2.imwrite（）不保存图像

Question

I have written a code in python opencv. I am trying to write the processed image back to disk but the image is not getting saved and it is not showing any error(runtime and compilation) The code is

"""
Created on Wed Oct 19 18:07:34 2016

@author: Niladri
"""

import numpy as np  
import cv2  

if __name__ == '__main__':  
import sys  

img = cv2.imread('C:\Users\Niladri\Desktop\TexturesCom_LandscapeTropical0080_2_S.jpg')
if img is None:  
    print 'Failed to load image file:'
    sys.exit(1)  

gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)  
h, w = img.shape[:2]  

eigen = cv2.cornerEigenValsAndVecs(gray, 15, 3)  
eigen = eigen.reshape(h, w, 3, 2)  # [[e1, e2], v1, v2]  
#flow = eigen[:,:,2] 
iter_n = 10
sigma = 5
str_sigma = 3*sigma
blend = 0.5
img2 = img
for i in xrange(iter_n):  
    print i,  

    gray = cv2.cvtColor(img2, cv2.COLOR_BGR2GRAY)  
    eigen = cv2.cornerEigenValsAndVecs(gray, str_sigma, 3)  
    eigen = eigen.reshape(h, w, 3, 2)  # [[e1, e2], v1, v2]  
    x, y = eigen[:,:,1,0], eigen[:,:,1,1] 
    print eigen

    gxx = cv2.Sobel(gray, cv2.CV_32F, 2, 0, ksize=sigma)  
    gxy = cv2.Sobel(gray, cv2.CV_32F, 1, 1, ksize=sigma)  
    gyy = cv2.Sobel(gray, cv2.CV_32F, 0, 2, ksize=sigma)  
    gvv = x*x*gxx + 2*x*y*gxy + y*y*gyy  
    m = gvv < 0  

    ero = cv2.erode(img, None)  
    dil = cv2.dilate(img, None)  
    img1 = ero  
    img1[m] = dil[m]  
    img2 = np.uint8(img2*(1.0 - blend) + img1*blend)  
#print 'done'
cv2.imshow('dst_rt', img2)
cv2.waitKey(0)
cv2.destroyAllWindows()

#cv2.imwrite('C:\Users\Niladri\Desktop\leaf_image_shock_filtered.jpg', img2)    



for i in xrange(iter_n):  
    print i,  

    gray = cv2.cvtColor(img2, cv2.COLOR_BGR2GRAY)  
    eigen = cv2.cornerEigenValsAndVecs(gray, str_sigma, 3)  
    eigen = eigen.reshape(h, w, 3, 2)  # [[e1, e2], v1, v2]  
    x, y = eigen[:,:,1,0], eigen[:,:,1,1] 
    print eigen

    gxx = cv2.Sobel(gray, cv2.CV_32F, 2, 0, ksize=sigma)  
    gxy = cv2.Sobel(gray, cv2.CV_32F, 1, 1, ksize=sigma)  
    gyy = cv2.Sobel(gray, cv2.CV_32F, 0, 2, ksize=sigma)  
    gvv = x*x*gxx + 2*x*y*gxy + y*y*gyy  
    m = gvv < 0  

    ero = cv2.erode(img, None)  
    dil = cv2.dilate(img, None)  
    img1 = dil  
    img1[m] = ero[m]  
    img2 = np.uint8(img2*(1.0 - blend) + img1*blend)  
print 'done'

#cv2.imwrite('D:\IP\tropical_image_sig5.bmp', img2)    


cv2.imshow('dst_rt', img2)
cv2.waitKey(0)
cv2.destroyAllWindows()

#cv2.imshow('dst_rt', img2)    

cv2.imwrite('C:\Users\Niladri\Desktop\tropical_image_sig5.bmp', img2)

Can anyone please tell me why it is not working. cv2.imshow is working properly(as it is showing the correct image). Thanks and Regards Niladri

Answer 1

As a general and absolute rule, you have to protect your windows path strings (containing backslashes) with r prefix or some characters are interpreted (ex: \\n,\\b,\\v,\\x aaaaand \\t , full list here ):

so when doing this:

cv2.imwrite('C:\Users\Niladri\Desktop\tropical_image_sig5.bmp', img2)

you're trying to save to C:\\Users\\Niladri\\Desktop<TAB>ropical_image_sig5.bmp

And the annoying thing with imread and imwrite is that those functions don't throw exceptions on errors, but fail silently. imwrite returns False

>>> cv2.imread("D:/nonexisting.jpg")  # this returns None, no error
>>> s = cv2.imread("D:/sloth_book.jpg")  # this works
>>> s
array([[[250, 250, 250],
        [246, 246, 246],
        [255, 255, 255],
        ...,
>>> cv2.imwrite("inexistent_dir/file.jpg",s)  # dir doesn't exist, write fails
False

So you have to check return value of those functions.

Do this:

if not cv2.imwrite(r'C:\Users\Niladri\Desktop\tropical_image_sig5.bmp', img2):
     raise Exception("Could not write image")

Note: the read works fine because "escaped" uppercase letters have no particular meaning in python 2 ( \\U and \\N have a meaning in python 3 so it wouldn't have worked)

And if there's an error, the program now complains loudly.

Answer 2

As Jean suggested, the error is due to the \\ being interpretted as an escape sequence. It is hence always safer to use os.path.join() as it is more cross platform and you need not worry about the escape sequence problem. For instance, in your case, you further need not worry about the first few arguments, as that is your home directory

import os
cv2.imwrite(os.path.join(os.path.expanduser('~'),'Desktop','tropical_image_sig5.bmp'), img2)

os.path.expanduser('~') will directly return your home directory.