make pdf

Make a searchable pdf via Google Cloud Vision OCR

Compile

gcv2hocr.c : use MingW64.

gcvocr.py : use nuitka.

python -m nuitka gcvocr.py --standalone --recurse-none --remove-output --plugin-enable=qt-plugins --recurse-to=requests --recurse-to=urllib3 --recurse-to=chardet --recurse-to=certifi --recurse-to=idna --recurse-to=PIL

hocr-pdf.py : use nuitka.

python -m nuitka hocr-pdf.py --standalone --recurse-none --remove-output --plugin-enable=qt-plugins --recurse-to=reportlab --recurse-to=lxml --recurse-to=PIL

makepdfgui.py : use nuitka.

python -m nuitka makepdfgui.py --standalone --recurse-none --remove-output --plugin-enable=tk-inter

All output files put in the same folder together.

Execute

Double click makepdfgui.exe.

Set Google API Key via press "Config" button.

Set output pdf file via press "Set pdf file" button.

Select input image file directry via press "Set IMG dir" button.

Wait until "Done" is shown.

Requrements for image datas

.jpg file, image size is less than 1.5MB, and image height and width are less than 1500pixels.

Source codes

gcv2hocr.c：https://github.com/dinosauria123/gcv2hocr/blob/master/main.c

hocr-pdf.py:https://github.com/tmbdev/hocr-tools/blob/master/hocr-pdf

gcvocr.py: shown at https://www.g104robo.com/entry/google-cloud-vision-api slightly modified.

Windows binary download

https://onedrive.live.com/download.aspx/pub/makepdfGUI.zip?cid=563E2C9BCE80B408

License

Licence for makepdfgui.py is CC.

Other files follow original source code licenses.

2020.1.6 ENDO Michiaki