ghallak / jpeg-python Goto Github PK

Hi Gaith,
I noticed in your jpeg repo that dct transformation code is not right,
dct_matrix = fftpack.dct(block, norm='ortho')
since the fftpack.dct() function is only 1D DCT instead of 2D DCT as in JPEG compression.
It should be implemented as follows,

def dct2(image):
    return fftpack.dct(fftpack.dct(image.T, norm='ortho').T, norm='ortho')
dct_matrix = dct2(block)

And similar for inverse dct transform..

i want to know
"parser.add_argument("output", help="path to the output image")", the output parameter‘s format is what?
i set :
input
./原始图像.bmp
output
./yasuo.jpg"
it sent:
usage: encoder.py [-h] input output
encoder.py: error: unrecognized arguments: output ./yasuo.jpg

When I try to run this: python encoder.py -h ./image.bmp ./output.jpg

it gives me this error:

File "encoder.py", line 60 os.path.dirname(filepath))) from e ^ SyntaxError: invalid syntax

Hi @ghallak , could you add an open source license to your code? for example the MIT license? I would like to re-use the code, modify it and make it open.
Thanks

I tried
python encoder.py test.bmp output.txt
then get a 0-1 string content text document

python decoder.py output.txt
then windwos Photos App shows the 90 degrees rotated image.

is this correct?

Great code for demonstrating the inner workings of JPEG using an easily interpreted language.
Minor critique - you may wish to include DPCM in your encoding of the DC components. Given how you're already segmenting DC and AC components, the modifications required to the encoder and decoder should be minimal.

I think u need to downsample your Cb abd Cr chanell after you convert the image Y CbCr

I have a png image with 423300 bytes.
After I ran
python ./encoder.py input.png output.jpg
The output.jpg file had 1565098 bytes.

Hey! After encoding, i have broken jpeg. Its HEX contains only 0x31 and 0x30. After decoding i have a very bad image. Help)

I have some problems with read_huffman_code function in decoder. When the chars end, the function endlessly adds to the prefix '' and the cycle does not end. Can you please explain me what the problem is?

how to adjust compression ratio？and why the decode image after encode will appear some white points, why?