Installation cloud

OS

For the demo cloud, you need to have an os with docker and docker-compose installed.

Configuration

• Clone this repo.
• Build the licence plate fake api docker image : cd LicencePlateApi then docker build -t neicureuil/plate-api:latest /LicencePlateApi/Dockerfile
• First deploy all the cloud with docker compose : docker-compose -f docker-orchrator.yaml up -d
• Open nodered : http://localhost:1880
• Import the flow orchestrator-flow.json inside nodered
• Open the database web ui : http://localhost:8080
• Login to the bdd with type : postgres, username : user, password : password, database : bfc, host: database
• Launch the script sql_init.sql from the web ui in the Import tab.

Installation raspberry

OS

For the demo, you need to install a raspbian lite (32-bit) version.

Install dependencies

• First install openvino 32-bit version on the raspberry, to do that, follow the guide on the official documentation : https://docs.openvino.ai/2022.3/openvino_docs_install_guides_installing_openvino_raspbian.html
• Add source /opt/intel/openvino_2022/setupvars.sh at the end of ~/.bashrc
• Make sur you have make, git, python3.9 and opencv installed : sudo apt udpate && sudo apt install git make python3 python3-opencv
• Install required python packages : pip3 install requests psycopg2 numpy
• Install docker : curl -fsSL https://get.docker.com -o get-docker.sh then sudo sh get-docker.sh
• Install docker compose : sudo apt install docker-compose
• Add your user to the docker group : sudo usermod -aG docker pi
• Create working directory : mkdir /home/pi/bfc && cd /home/pi/bfc
• Clone the sdk in the work folder : git clone https://github.com/DoubangoTelecom/ultimateALPR-SDK/tree/master --depth=1
• Create symbolic link : ln -s ultimateALPR-SDK-master/samples/python/recognizer app
• Add execution perms to lib : sudo chmod +x ultimateALPR-SDK-master/binaries/raspbian/armv7l/libultimate_alpr-sdk.so
• Add execution perms to setup : sudo chmod +x ultimateALPR-SDK-master/binaries/raspbian/armv7l/python_setup.sh and sudo chmod +x ultimateALPR-SDK-master/python/recognizer/run_rpi_armv7l.sh
• Build sdk : cd /home/pi/bfc/app and ./run_rpi_armv7l.sh

Install app

• To install the app copy the content of the folder Raspberry of this repository to the folder /home/pi/bfc/app
• Go to the app folder cd /home/pi/bfc/app
• Update the cloud ip in the file : recognizer2.py and sync.py
• Add rights to files : sudo chmod +x run_detect.sh run_verba.sh sync.sh
• Start the cache database : docker-compose -f docker-compose.yml up -d
• Start the first sync manually : ./sync.sh

Automatic cache sync

To sync the authorization cache automaticaly, add the sync script to the crontab :

• Open crontab editing with : crontab -e
• Add the line : * * * * * /home/pi/bfc/app/sync.sh

Start demos

Demo information sign

• Start the script run_detect.sh in /home/pi/bfc/app

Demo verbalization

• Start the script run_verba.sh in /home/pi/bfc/app