convert_image_formats.ipynb

Kernel: Python 3

!pip install pdf2image

Collecting pdf2image
  Downloading pdf2image-1.16.0-py3-none-any.whl (10 kB)
Requirement already satisfied: pillow in /usr/local/lib/python3.7/dist-packages (from pdf2image) (7.1.2)
Installing collected packages: pdf2image
Successfully installed pdf2image-1.16.0

import pdf2image

!sudo apt-get install poppler-utils

Reading package lists... Done
Building dependency tree       
Reading state information... Done
The following package was automatically installed and is no longer required:
  libnvidia-common-460
Use 'sudo apt autoremove' to remove it.
The following NEW packages will be installed:
  poppler-utils
0 upgraded, 1 newly installed, 0 to remove and 40 not upgraded.
Need to get 154 kB of archives.
After this operation, 613 kB of additional disk space will be used.
Get:1 http://archive.ubuntu.com/ubuntu bionic-updates/main amd64 poppler-utils amd64 0.62.0-2ubuntu2.12 [154 kB]
Fetched 154 kB in 1s (271 kB/s)
debconf: unable to initialize frontend: Dialog
debconf: (No usable dialog-like program is installed, so the dialog based frontend cannot be used. at /usr/share/perl5/Debconf/FrontEnd/Dialog.pm line 76, <> line 1.)
debconf: falling back to frontend: Readline
debconf: unable to initialize frontend: Readline
debconf: (This frontend requires a controlling tty.)
debconf: falling back to frontend: Teletype
dpkg-preconfigure: unable to re-open stdin: 
Selecting previously unselected package poppler-utils.
(Reading database ... 148489 files and directories currently installed.)
Preparing to unpack .../poppler-utils_0.62.0-2ubuntu2.12_amd64.deb ...
Unpacking poppler-utils (0.62.0-2ubuntu2.12) ...
Setting up poppler-utils (0.62.0-2ubuntu2.12) ...
Processing triggers for man-db (2.8.3-2ubuntu0.1) ...

from google.colab import drive

drive.mount("/content/drive", force_remount=True)

Mounted at /content/drive

# pip install pdf2image
# pip install --upgrade pillow

import os
import shutil
from pdf2image import convert_from_path
from PIL import Image
from PIL import ImageCms
import argparse
from glob import glob
from tqdm import tqdm
import functools
import multiprocessing
import concurrent.futures


def split_file_name(input_path):
    base_name, dir_name = os.path.basename(input_path), os.path.dirname(input_path)
    file_name, ext = os.path.splitext(os.path.basename(base_name))
    return base_name, dir_name, file_name, ext


def convert(
    input_path,
    output_path,
    color_space="CMYK",
    input_profile_path=None,
    output_profile_path=None,
    quality=100,
    verbose=False,
    overwrite=False,
):
    """converts an image or pdf into a color space of choice
    for CMYK the default output format is JPG
    Keyword arguments:
    input_path -- the input path of the file
    output_path -- the output path for the result to be written.
    color_space -- the color space to convert to , default value is CMYK
    input_profile_path -- the path to the input profile
    output_profile_path -- the path to the output profile
    """
    try:
        if not overwrite and os.path.exists(output_path):
            return True

        if input_path.endswith(".pdf") or input_path.endswith(".PDF"):
            # _, dir_name, file_name, _ =split_file_name(output_path)
            _, dir_name, file_name, _ = split_file_name(input_path)
            temp_file_name = "temp" + file_name
            temp_file_path = os.path.join(dir_name, temp_file_name)
            print("converting ", input_path, " to ", temp_file_path)
            convert_from_path(
                input_path,
                output_file=temp_file_path,
                fmt="png",
                use_pdftocairo=True,
                single_file=True,
                use_cropbox=True,
            )
            temp_file_path += ".png"
            print("converting ", temp_file_path, " to ", output_path)
            _convert_profiles(
                temp_file_path,
                output_path,
                color_space=color_space,
                input_profile_path=input_profile_path,
                output_profile_path=output_profile_path,
                quality=quality,
            )
            os.remove(temp_file_path)
            return True
        elif (
            input_path.endswith(".png")
            or input_path.endswith(".PNG")
            or input_path.endswith(".jpg")
            or input_path.endswith(".JPG")
            or input_path.endswith(".jpeg")
            or input_path.endswith(".JPEG")
        ):
            return _convert_profiles(
                input_path,
                output_path,
                color_space=color_space,
                input_profile_path=input_profile_path,
                output_profile_path=output_profile_path,
                quality=quality,
            )
        else:
            print(f"{input_path} is not a valid image file, copying it instead to {output_path}.")
            shutil.copy(input_path, output_path)
            return False
    except Exception as e:
        if verbose:
            print(f"Error in file: {input_path}\n", e)
        return False


def _convert_profiles(
    input_path=None,
    output_path=None,
    color_space="CMYK",
    input_profile_path=None,
    output_profile_path=None,
    quality="100",
):
    try:
        with Image.open(input_path) as im:
            img_cmyk = ImageCms.profileToProfile(
                im, input_profile_path, output_profile_path, renderingIntent=0, outputMode=color_space
            )
            quality = int(quality)
            img_cmyk.save(output_path, quality=quality)

            return True
    except Exception as e:
        print(e)
        print(f"cannot convert{input_path}, copying it instead.")
        shutil.copy(input_path, output_path)
        return False


# from https://pillow.readthedocs.io/en/stable/handbook/tutorial.html?highlight=cmyk#using-the-image-class
def check_image_properties(input_path):
    try:
        with Image.open(input_path) as im:
            print(input_path, im.format, f"{im.size}x{im.mode}")
    except OSError as e:
        print("error opening the image\n", e)

from glob import glob

files = glob("/content/drive/MyDrive/MLAPA/book-images-original/*.*")
p = [print(f) for f in files]

filenames = []
for f in files:
    parts = f.split("/")
    fname = parts[-1]
    base = fname.split(".")[:-1][0]
    # filenames.append(base)
    filenames.append(fname)

print(filenames)

/content/drive/MyDrive/MLAPA/book-images-original/kernelBinaryClassifDemoRVM.pdf
/content/drive/MyDrive/MLAPA/book-images-original/kernelBinaryClassifDemologregL2.pdf
/content/drive/MyDrive/MLAPA/book-images-original/kernelBinaryClassifDemologregL1.pdf
/content/drive/MyDrive/MLAPA/book-images-original/kernelBinaryClassifDemoSVM.pdf
['kernelBinaryClassifDemoRVM.pdf', 'kernelBinaryClassifDemologregL2.pdf', 'kernelBinaryClassifDemologregL1.pdf', 'kernelBinaryClassifDemoSVM.pdf']

in_folder = "/content/drive/MyDrive/MLAPA/book-images-original"
for use_rgb in [False]:
    if use_rgb:
        out_folder = "/content/drive/MyDrive/MLAPA/book-images-rgb-80"
        color_space = "RGB"
        quality = 80
    else:
        out_folder = "/content/drive/MyDrive/MLAPA/book-images-cmyk-100"
        color_space = "CMYK"
        quality = 100

    rgb_profile = "sRGB Color Space Profile.icm"
    cmyk_profile = "USWebCoatedSWOP.icc"
    profile_folder = "/content/drive/MyDrive/MLAPA"
    input_profile_path = f"{profile_folder}/{rgb_profile}"
    if color_space == "RGB":
        output_profile_path = f"{profile_folder}/{rgb_profile}"
    else:
        output_profile_path = f"{profile_folder}/{cmyk_profile}"

    for fname in filenames:
        base = fname.split(".")[:-1][0]
        in_name = f"{in_folder}/{fname}"
        # in_name = f'{in_folder}/{fname}.pdf'
        out_name = f"{out_folder}/{base}.jpg"
        print("!converting ", in_name, " to ", out_name)
        convert(
            in_name,
            out_name,
            color_space=color_space,
            quality=quality,
            verbose=True,
            input_profile_path=input_profile_path,
            output_profile_path=output_profile_path,
        )

!converting  /content/drive/MyDrive/MLAPA/book-images-original/kernelBinaryClassifDemoRVM.pdf  to  /content/drive/MyDrive/MLAPA/book-images-cmyk-100/kernelBinaryClassifDemoRVM.jpg
converting  /content/drive/MyDrive/MLAPA/book-images-original/kernelBinaryClassifDemoRVM.pdf  to  /content/drive/MyDrive/MLAPA/book-images-original/tempkernelBinaryClassifDemoRVM
converting  /content/drive/MyDrive/MLAPA/book-images-original/tempkernelBinaryClassifDemoRVM.png  to  /content/drive/MyDrive/MLAPA/book-images-cmyk-100/kernelBinaryClassifDemoRVM.jpg
!converting  /content/drive/MyDrive/MLAPA/book-images-original/kernelBinaryClassifDemologregL2.pdf  to  /content/drive/MyDrive/MLAPA/book-images-cmyk-100/kernelBinaryClassifDemologregL2.jpg
converting  /content/drive/MyDrive/MLAPA/book-images-original/kernelBinaryClassifDemologregL2.pdf  to  /content/drive/MyDrive/MLAPA/book-images-original/tempkernelBinaryClassifDemologregL2
converting  /content/drive/MyDrive/MLAPA/book-images-original/tempkernelBinaryClassifDemologregL2.png  to  /content/drive/MyDrive/MLAPA/book-images-cmyk-100/kernelBinaryClassifDemologregL2.jpg
!converting  /content/drive/MyDrive/MLAPA/book-images-original/kernelBinaryClassifDemologregL1.pdf  to  /content/drive/MyDrive/MLAPA/book-images-cmyk-100/kernelBinaryClassifDemologregL1.jpg
converting  /content/drive/MyDrive/MLAPA/book-images-original/kernelBinaryClassifDemologregL1.pdf  to  /content/drive/MyDrive/MLAPA/book-images-original/tempkernelBinaryClassifDemologregL1
converting  /content/drive/MyDrive/MLAPA/book-images-original/tempkernelBinaryClassifDemologregL1.png  to  /content/drive/MyDrive/MLAPA/book-images-cmyk-100/kernelBinaryClassifDemologregL1.jpg
!converting  /content/drive/MyDrive/MLAPA/book-images-original/kernelBinaryClassifDemoSVM.pdf  to  /content/drive/MyDrive/MLAPA/book-images-cmyk-100/kernelBinaryClassifDemoSVM.jpg
converting  /content/drive/MyDrive/MLAPA/book-images-original/kernelBinaryClassifDemoSVM.pdf  to  /content/drive/MyDrive/MLAPA/book-images-original/tempkernelBinaryClassifDemoSVM
converting  /content/drive/MyDrive/MLAPA/book-images-original/tempkernelBinaryClassifDemoSVM.png  to  /content/drive/MyDrive/MLAPA/book-images-cmyk-100/kernelBinaryClassifDemoSVM.jpg

!ls /content/drive/MyDrive/MLAPA/book-images-original

kernelBinaryClassifDemologregL1.pdf  kernelBinaryClassifDemoRVM.pdf
kernelBinaryClassifDemologregL2.pdf  kernelBinaryClassifDemoSVM.pdf

!ls /content/drive/MyDrive/MLAPA/book-images-rgb-80

ae_mnist_conv_20d_linear.jpg   svmCvSurf.jpg
ae_mnist_conv_20d_samples.jpg  vae_mnist_conv_20d_linear.jpg
largeMarginPrinciple2.jpg      vae_mnist_conv_20d_samples.jpg
separable-conv2d.jpg

!ls /content/drive/MyDrive/MLAPA/book-images-cmyk-100

ae_mnist_conv_20d_linear.jpg	     largeMarginPrinciple2.jpg
ae_mnist_conv_20d_samples.jpg	     separable-conv2d.jpg
kernelBinaryClassifDemologregL1.jpg  svmCvSurf.jpg
kernelBinaryClassifDemologregL2.jpg  vae_mnist_conv_20d_linear.jpg
kernelBinaryClassifDemoRVM.jpg	     vae_mnist_conv_20d_samples.jpg
kernelBinaryClassifDemoSVM.jpg