CropTure

An application that classifies crop diseases and give details about the disease and solutions on how to prevent and cure them.

Things to take note

Install Docker, Start Docker, Then Run the Following:

docker run -t -rm -p <port>:<port> -v <directory>:<docker directory> tensorflow/serving --rest_api_port=<port> --model_config_file=<models.config directory>

Example:

sudo docker run -t --rm -p 8501:8501 -v /home/carl/Documents/Development/CropTure:/CropTure tensorflow/serving --rest_api_port=8501 --model_config_file=/CropTure/models.config

Make sure to check models.config and rename the name column base on the code.

After doing all the process make a postman check using the URL below:

http://localhost:8001/predict

Make sure to get a new request using POST method.

Then go to Body > Pick form-data > add a Key named "file", then in the value select an image based on dataset then press send.

Response should look something like this.

If everything works fine then you are good to go.

But if not, DEEEEBUUUUUG!!!

PS: Don't forget that the endpoint in /api/main-tf-serving.py is connected to models.config's 'name'