A component of a traffic sign recognition application, that publishes the recognizer model.
The traffic sign recognizer comes with a SPA client application, that was developed on an Angular 8 platform. The client supports real time recognition by using the device's camera.
The application comes with a deployed Spring Boot server application. The server mediates the communication between the client and the traffic sign detector.
The traffic sign detector is published in this repository. It contains a localizer and a classifier, both implemented with the Keras API. The detector is not deployed, due to lack of resources.
In order to get the TSR application properly working on your PC follow the following steps.
- Clone the 3 repositories
- Run the following commands in the client repository
npm install
npm run start
- Run the following commands in the detector repository
pipenv install
python runserver 8081
- Run the following commands in the server repository
mvn clean install
mvn spring-boot:run
It is also possible to start each server on a different port. By default the client server runs on 4200, the spring server runs on 8080, and the detector runs on 8081. If you want the detector to run on a different port, start it with
python runserver <port number>
and start the spring server with
mvn spring-boot:run DETECTOR_HOST=http://localhost:<port number>