A table reconstruction package
Project description
Table Reconstruction
table-reconstruction
is a tool used to detect table spaces and reconstruct the information in them using DL models.
To provide the above feature, Table reconstruction works based on several components as follows:
- A table detection model is developed based on Yolov5
- A line segmentation model is built based on Unet
- Additional modules are used in the information extraction process, especially a directed graph is used to extract information related to the merged cells.
Installation
Table Reconstruction is published on PyPI and can be installed from there:
pip install table-reconstruction
You can also install this package manually with the following command:
python setup.py install
Basic usage
you can easily use this library by using the following statements
import torch
from table_reconstruction import TableExtraction
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
extraction = TableExtraction(device=device)
image = ... # Accept Numpy ndarray and PIL image
tables = extraction.extract(image)
We also provide a simple Jupyter notebook which can be used to illustrate the results obtained after processing, please check it out here
Documentation
Documentation will be available soon.
Get in touch
- Report bugs, suggest features or view the source code on GitHub.
Project details
Release history Release notifications | RSS feed
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
Close
Hashes for table_reconstruction-0.0.3.tar.gz
Algorithm | Hash digest | |
---|---|---|
SHA256 | 4837cb1b769f8f2c6fb8d7678854a3e83a5a1ea146b05e35aa21963b180e9b9e |
|
MD5 | 0d10a844498d848caf2294b4354f7387 |
|
BLAKE2b-256 | 3a4588a89a40197c0da23985711aa6bb0620d490b46d6caabc4098978d001fbb |
Close
Hashes for table_reconstruction-0.0.3-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 982184d3584d23cd69e7b57b88b999e89eb0ed53bd4d58a3a66a19edea5388d9 |
|
MD5 | 46b4376d33aa8739d90727c49152df5f |
|
BLAKE2b-256 | 0683267a5eb60c7331e0d508c4f4e4aca96eb6baded179b5d7152f33483c138f |