Two-Stage Flask OCR

This is asimple python optical character recognition server. It's model consists of two parts - detection(CRAFT) and recognition(deep-text-regonition)

Warning! This is a commercial application by Huawei Turkey (Copyright 2021)

Getting started

Install dependencies;

• numpy
• Pillow
• opencv-python-headless
• scikit-image
• werkzeug
• flask-restplus == 0.12.1
• flask == 1.0.3
• pyarmor
• cryptography

pip install -r requirements.txt

Run Server

Open the terminal on the project path and then run the following command.

python3 app.py

Note : Import OCR_REST-API.postman_collection.json file to postman collections for easy demo

Swagger Usage

Swagger will be available http://{server_ip}:{server_port}

Please, look at the wiki page for swagger usage details

Docker Build & Run

First of all, download Ascend-cann-nnrt_5.0.2_linux-x86_64.run inference engine software package from hiascend website to project path and then build the docker image by running the following code on bash.

sudo docker build -t pyacl_ocr_api:1.0 \
                --build-arg NNRT_PKG=Ascend-cann-nnrt_5.0.2_linux-x86_64.run .

Finally, you can run the docker image with the following code on bash.

sudo docker_run.sh 8500 8500 pyacl_ocr_api:1.0

File Encryption

Please, use utils/file_encryption.py script to encrypt files like *.om and *.cfg files as follows

python3 file_encryption.py --model <file_name>

And then put the generated code in the app.py script.

Code Obfuscation & Licensing

First, creat an expired license and then obfuscate the code with that license as like below

$ pyarmor licenses --expired YYYY-MM-DD r001
$ pyarmor obfuscate --with-license licenses/r001/license.lic --recursive --advanced 2 app.py