---

Tabulo is an open source toolkit for computer vision. Currently, we support table detection, but we are aiming for much more. It is built in Python, using Luminoth, TensorFlow and Sonnet.

Table of Contents

1. Installation Instructions
   
2. Avaiable API's
   
3. Working with pretrained Models
   
4. Runnning Tabulo
   
5. Runnning Tabulo As Service
   
6. Supported models
   
7. Usage
   
8. Working with datasets
   
9. Training
   
10. LICENSE
    

1. Installation Instructions

Tabulo currently supports Python 2.7 and 3.4–3.6.

1.1 Pre-requisites

To use Tabulo, TensorFlow must be installed beforehand. If you want GPU support, you should install the GPU version of TensorFlow with pip install tensorflow-gpu, or else you can use the CPU version using pip install tensorflow.

We are using tesseract to extract data from table so you have to install tesseract also. Follow this link to install tessersact

1.2 Installing Tabulo

First, clone the repo on your machine and then install with pip:

git clone https://github.com/interviewBubble/Tabulo.git
cd tabulo
pip install -e .

1.3 Check that the installation worked

Simply run tabulo --help.

2. Avaiable API's

• localhost:5000/api/fasterrcnn/predict/ - To detect table in the image
• localhost:5000/api/fasterrcnn/extract/ - Extract table content from detected tables

3. Working with pretrained Models:

• DOWNLOAD pretrained model from Google drive
• Unzip and Copy downloaded luminoth folder inside luminoth/utils/pretrained_models folder
• Hit this command to list all check points: tabulo checkpoint list
• You will get output like this:
• Now run server using this command: tabulo server web --checkpoint 6aac7a1e8a8e

4. Runnning Tabulo

4.1 Running Tabulo as Web Server:

4.2 Example of Table Detection with Faster R-CNN By Tabulo:

4.3 Example of Table Data Extraction with tesseract By Tabulo:

5. Runnning Tabulo As Service:

5.1 Using Curl command

curl -X POST \
  http://localhost:5000/api/fasterrcnn/predict/ \
  -H 'Content-Type: application/x-www-form-urlencoded' \
  -H 'Postman-Token: 70478bd2-e1e8-442f-b0bf-ea5ecf7bf4d8' \
  -H 'cache-control: no-cache' \
  -H 'content-type: multipart/form-data; boundary=----WebKitFormBoundary7MA4YWxkTrZu0gW' \
  -F image=@/path/to/image/page_8-min.jpg

5.2 With PostMan

Header Section:

Data Section:

6. Supported models

Currently, we support the following models:

• Object Detection
  • Faster R-CNN
  • SSD

We also provide pre-trained checkpoints for the above models trained on popular datasets such as COCO and Pascal.

7. Usage

There is one main command line interface which you can use with the tabulo command. Whenever you are confused on how you are supposed to do something just type:

tabulo --help or tabulo <subcommand> --help

and a list of available options with descriptions will show up.

8. Working with datasets

DataSet to train your custom model.

9. Training

See Training your own model to learn how to train locally or in Google Cloud.

10. LICENSE

Released under the BSD 3-Clause.

---

References

• https://github.com/Sargunan/Table-Detection-using-Deep-learning
• https://github.com/tryolabs/luminoth