Stamaks / handwrite-recognition

Handwrite-recognition

Project purpose

Handwritten text recognition

Project input data

Image submitted by user via image submit form

Project output data

Recognized text

Project main componenets

1. Client-side frontend: website, point of communication with user
2. Nginx: reverse-proxy + frontend server
3. Backend:
   • Backend server, handling requests from Client-Side page
   • Messaging queue, queueing heavy image processing requests
   • Queue consumer, reading queue and requesting ML nodes, updating status on Backend Server when predict processing finished
   • ML Node, serving 'predict' requests

Usage

Launch

[e-mail]:Nosp27/handwrite-recognition.git && \
cd handwrite-recognition && \
docker-compose up --build

When launched application, open localhost in your browser and choose an image in the form. Submit the form, check out status change, displayed below.

View logs

For now logs are present for ML/Message Queue and for Backend. To view them attach to running container with help of docker exec then open file in /logs/ directory.

docker exec -it  docker exec -it handwrite_recognition_ml /bin/bash
more /logs/ml.log

Or for backend logs:

docker exec -it  docker exec -it handwrite_recognition_web /bin/bash
more /logs/backend.log

To filter logs of application itself use:

• grep root /logs/backend.log for backend container
• grep consumer /logs/ml.log for ml container