main.py

import io

from PIL import Image
from fastapi import FastAPI
from fastapi.middleware.cors import CORSMiddleware
import uvicorn
import os
from pydantic import BaseModel
from fastapi import FastAPI, Form, UploadFile
from typing import Annotated

# Tensorflow
import tensorflow as tf
from tensorflow.keras.models import load_model
from tensorflow.keras.utils import get_file
from tensorflow.keras.utils import load_img
from tensorflow.keras.utils import img_to_array
import tensorflow_decision_forests
import numpy as np

from data import crop_data, leaf_disease
from gcs import GCStorage

app = FastAPI()

origins = ["*"]
methods = ["*"]
headers = ["*"]

app.add_middleware(
    CORSMiddleware,
    allow_origins=origins,
    allow_credentials=True,
    allow_methods=methods,
    allow_headers=headers,
)


@app.get("/")
async def root():
    return {"message": "Welcome !"}


class CropInput(BaseModel):
    nitrogen: int = 0
    phosphorous: int = 0
    potassium: int = 0
    temperature: int = 0
    humidity: int = 0
    ph: int = 0
    rainfall: int = 0


@app.post("/crop-recommendations")
async def get_crop_recommendations(
        nitrogen: Annotated[float, Form()],
        phosphorous: Annotated[float, Form()],
        potassium: Annotated[float, Form()],
        temperature: Annotated[float, Form()],
        humidity: Annotated[float, Form()],
        ph: Annotated[float, Form()],
        rainfall: Annotated[float, Form()]
):
    model = tf.saved_model.load('./model/crop_recommdation')

    input_data = {
        "N": np.array([nitrogen]),
        "P": np.array([phosphorous]),
        "K": np.array([potassium]),
        "temperature": np.array([temperature]),
        "humidity": np.array([humidity]),
        "ph": np.array([ph]),
        "rainfall": np.array([rainfall])
    }

    pred = model.serve(input_data)

    # Daftar nama kelas
    class_names = ['rice', 'maize', 'chickpea', 'kidneybeans', 'pigeonpeas', 'mothbeans', 'mungbean', 'blackgram',
                   'lentil', 'pomegranate', 'banana', 'mango', 'grapes', 'watermelon', 'muskmelon', 'apple',
                   'orange', 'papaya', 'coconut', 'cotton', 'jute', 'coffee']

    # Mencari indeks dengan nilai prediksi tertinggi

    predicted_label = pred[0]
    predicted_index = np.argmax(predicted_label)

    # Mengambil nama kelas yang sesuai dengan indeks
    predicted_class = class_names[predicted_index]

    return crop_data[predicted_class]


def get_predict_detail(predictions, classes):
    predicts = []
    class_names = ['bacterial_leaf_blight', 'brown_spot', 'healthy', 'leaf_blast', 'leaf_scald', 'narrow_brown_spot']
    for x in predictions.tolist():
        predicts.append({
            "disease": leaf_disease[class_names[x]],
            "percentage": classes[x],
        })
    return predicts


@app.post("/rice-disease-detection")
async def get_rice_leaf_disease_detection(image: UploadFile):

    model_dir = "./model/rice_leaf_detection.h5"
    model = load_model(model_dir)

    contents = image.file.read()
    temp_file = io.BytesIO()
    temp_file.write(contents)
    temp_file.seek(0)
    load_image = load_img(temp_file, target_size=(224, 224))

    x = img_to_array(load_image)
    x = np.expand_dims(x, axis=0)
    images = np.vstack([x])

    classes = model.predict(images)[0]
    prediction_probabilities = tf.math.top_k(classes, k=3)

    predictions = prediction_probabilities.indices.numpy()
    image_link = GCStorage().upload_file(image)

    return {
        "predictions": get_predict_detail(predictions, classes.tolist()),
        "image_url": image_link,
    }


if __name__ == "__main__":
    port = int(os.environ.get('PORT', 5000))
    uvicorn.run(app, host="0.0.0.0", port=8000)