A package for the signature detection
Project description
Signature Detection
A simple tool to detect if there are signatures in an image or a PDF file.
The full documentation is presented at the Github Repository.
Extra Dependencies
This tool uses Wand
to convert a PDF file into images.
Wand
is a ctypes-based simple ImageMagick
binding for Python. You may need to install ImageMagick
on your local machine.
More details are available here.
Usage
Signature_detect contains 4 classes.
- Loader
- Extractor
- Cropper
- Judger
Loader
Loader can read an image or a PDF file page by page.
It returns a list of the masks. Each mask is a numpy 2 dimensions array. Its element's value is 0
or 255
.
from signature_detect.loader import Loader
loader = Loader(
low_threshold=(0, 0, 250),
high_threshold=(255, 255, 255))
masks = loader.get_masks(file_path)
Extractor
Extractor reads a mask, labels the regions in the mask, and removes both small and big regions. We consider that the signature is a region of middle size.
from signature_detect.extractor import Extractor
extractor = Extractor(
outlier_weight=3,
outlier_bias=100,
amplfier=10,
min_area_size=10)
labeled_mask = extractor.extract(mask)
Cropper
Cropper crops the regions in the labeled mask.
from signature_detect.cropper import Cropper
cropper = Cropper(
min_region_size=10000,
border_ratio=0.1)
results = cropper.run(labeled_mask)
Judger
Judger decides whether a region is a signature.
from signature_detect.judger import Judger
judger = Judger(
size_ratio=[1, 4],
pixel_ratio=[0.01, 1])
is_signed = judger.judge(result["cropped_mask"])
Dev version
If you would like to develop this package and run the tests, you can download the code and install dev environment locally.
pip install -e .[dev]
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for signature_detect-0.1.4-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 52c2775ff0415619243afc0b943acbfcdc86be3f9999582cd6adae3d80d7dc2a |
|
MD5 | 5854fe55aa40f93b5f71edcfd07fcd65 |
|
BLAKE2b-256 | 8f7996c9b3f4eef8d33bf18bf44e76f1f35d8ca6115ad54b71237ae580eee11e |