I am trying to read this file using Pillow and cv2 in python. However, both libraries are failing to do so.
File is here on google drive.
Pillow raises an "OSError: cannot identify image file 0.jpg" cv2 returns None
This binary file has the extension ".jpg" and can be easily viewed using Window`s Photos viewer.
However, I want to read and decode this binary file as an image in python.
Any idea?
Your file is actually a JXR file, you can see the first few bytes are:
49 49 BC
if you load it into a hex editor like http://hexed.it
The signature is IANA registered, see here.
You can read it with imagecodecs like this:
import imagecodecs
from pathlib import Path
# Slurp entire image data as binary
jxr = Path('0.jpg').read_bytes()
# Make into Numpy array
na = imagecodecs.jpegxr_decode(jxr)
print(na.shape) # prints (256, 512, 3)
Alternatively, you can convert JXR to something less Microsofty with ImageMagick in your Terminal with:
magick input.jxr output.jpg
Or, if you have old v6 ImageMagick, use:
convert input.jxr output.jpg