Python OCR

A proof of concept of a small OCR text recognition app using python + tesseract-ocr within Docker

Prerequisites

• docker@^24.0.3

How to use it?

1. Clone the repo executing git clone [email protected]:dmmarmol/python-ocr.git
2. Navigate into cd python-ocr
3. Run the command make build (Makefile shortcut to build the docker image)
4. Run the command make run
5. Run the command make shell to attach a shell to the running container
6. Choose between bulk process images or process text

Process images

This command will read all images inside the images/source directory and will extract the text content from each of them putting them all together in a new file inside images/output

Steps

1. Deposit any .jpg or jpeg file inside images/source directory
2. Navigate inside the container using an attached shell and from the app/ directory, run the command python3 src/process-images.py

Process text

This command will read all .txt files inside the text/source directory and will normalize the text content from each of file putting them all together in a new file inside text/output

Steps

1. Deposit any .txt file inside text/source directory
2. Navigate inside the container using an attached shell and from the app/ directory, run the command python3 src/process-text.py

Commands

Build

Build the Dockerfile image

make build

Run

Run a docker container instance of the Dockerimage

make run

Restart docker container

make stop
make remove

Lastly, repeat build and run commands