cv2.imwrite doesn't save result image but initial only

Question

I have a code:

import cv2
import numpy

background = numpy.zeros((1080, 1920, 3))
img = numpy.ones((255, 465, 3))
offset = numpy.array((12, 12))

x_offset = offset[0]
y_offset = offset[1]

img_w = img.shape[1]
img_h = img.shape[0]

background_w = background.shape[1]
background_h = background.shape[0]

x = x_offset
y = y_offset
for i in range(0, 16):
    background[y:y + img.shape[0], x:x + img.shape[1]] = img
    x += img_w + x_offset
    if x > background_w - img_w:
        x = x_offset
        y += img_h + y_offset

cv2.imshow("test", background)
cv2.imwrite("background.jpg", background)
cv2.waitKey(0)
cv2.destroyAllWindows()

that generates grid like this one:

So cv2.imshow shows the grid but cv2.imwrite writes only initial black background not the grid for some reason. How to fix that?

I have a strange suspicion, could you move cv2.waitKey(0) one line up? so between the imshow and imwrite? and also make sure that the file you check is really the file you saved (check if the timestamp on the file is changed) — Nullman
– Nullman, Commented Mar 6, 2022 at 9:00
Does this answer your question? why my imshow differs from imwrite — Christoph Rackwitz
– Christoph Rackwitz, Commented Mar 6, 2022 at 12:57

Rabbid76 · Accepted Answer · 2022-03-06 10:22:53Z

4

You need to scale the color channels:

~~cv2.imwrite("background.jpg", background)~~

cv2.imwrite("background.jpg", background * 255)

Alternatively you can create a "white" image with type uint8:

~~img = numpy.ones((255, 465, 3))~~

img = numpy.ones((255, 465, 3), dtype = numpy.uint8) * 255

edited Mar 6, 2022 at 10:22

answered Mar 6, 2022 at 9:50

Rabbid76

212k30 gold badges163 silver badges205 bronze badges

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

Collectives™ on Stack Overflow

cv2.imwrite doesn't save result image but initial only

1 Answer 1

Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related