Vehicle Type Recognition and Vehicle Color Recognition
Help was taken from https://github.com/hoanhle/Vehicle-Type-Detection for this part. All due credits to the original owner. The vehicle image is classified into one of the following:
- Ambulance
- Barge
- Bicycle
- Boat
- Bus
- Car
- Cart
- Caterpillar
- Helicopter
- Limousine
- Motorcycle
- Segway
- Snowmobile
- Tank
- Taxi
- Truck
- Van
Make sure all the requirements are installed as specified in requirements.txt file. Then run api_server.py file to run the server. You can make a POST request at the specified server as shown in the terminal when you run api_server.py file. Attach the image whose vehicle type needs to be recognized in the POST request with the key image. The response is a json text with key result containing the prediction.
The trained models can be found in vehicle_type_recognition folder at https://drive.google.com/drive/folders/1iBAn9IwWXY8Ur4JA89ZkIOP4MSjtDea0?usp=sharing and these models (namely densenet.h5, inception_v3.h5 and mobilenet_v2.h5) should be downloaded and put inside vehicle_type_recognition/models folder. We can change the path of the folder which contains the models by changing the app.config["MODELS_PATH"] configuration of the flask app (currently, this configuration is set inside api_server.py file).
Visit the VehicleColorRecognition repository for details of Vehicle Color Recognition part.