About

This is a fork from https://github.com/hikopensource/DAVAR-Lab-OCR with modified setup script and added docker web service for LGPMA table extraction model inference.

Web service setup

Download model maskrcnn-lgpma-pub-e12-pub.pth from Google Drive and place it in models directory.

Running service locally

• Run setup.sh locally and it should install and build all required components.
• Then you can run web service using uvicorn uvicorn src.web:app --host 0.0.0.0 --port 8000

Running service from docker container

• Run docker build -f docker/Dockerfile -t <tag> .
• Then you can start the container with web service with docker run -d -p 8000:8000 -t <tag>

Usage

When the web service is running, you can use http protocol to process table images or set processing device.

Service endpoints

• /extract_table_structure/ - returns json with the table data obtained from model.
• /get_device/ - returns the name of the device used by model, either cpu or cuda.
• /set_device/{device}/ - tries to set processing device as device and returns json with currently used device and method output - either success or failure

Example

import requests

filename = ...
with open(filename, "rb") as f:
    response = requests.post(url="http://0.0.0.0:8000/extract_table_structure", files={"input_file": f})