main.py

import os
import numpy as np
import tensorflow as tf
from fastapi import FastAPI, File, UploadFile
from fastapi.responses import JSONResponse
from io import BytesIO
from PIL import Image
from tensorflow.keras.preprocessing.image import img_to_array
import tensorflow_addons as tfa
import uvicorn
import requests

# Initialize FastAPI app
app = FastAPI()

# Register the custom object
custom_objects = {
    "Addons>CohenKappa": tfa.metrics.CohenKappa,
}

# Model and class information
model_path = "model.h5"
class_labels = {
    0: "Apple___Apple_scab",
    1: "Apple___Black_rot",
    2: "Apple___Cedar_apple_rust",
    3: "Apple___healthy",
    4: "Background_without_leaves",
    5: "Blueberry___healthy",
    6: "Cherry___Powdery_mildew",
    7: "Cherry___healthy",
    8: "Corn___Cercospora_leaf_spot Gray_leaf_spot",
    9: "Corn___Common_rust_",
    10: "Corn___Northern_Leaf_Blight",
    11: "Corn___healthy",
    12: "Grape___Black_rot",
    13: "Grape___Esca_(Black_Measles)",
    14: "Grape___Leaf_blight_(Isariopsis_Leaf_Spot)",
    15: "Grape___healthy",
    16: "Orange___Haunglongbing_(Citrus_greening)",
    17: "Peach___Bacterial_spot",
    18: "Peach___healthy",
    19: "Pepper,_bell___Bacterial_spot",
    20: "Pepper,_bell___healthy",
    21: "Potato___Early_blight",
    22: "Potato___Late_blight",
    23: "Potato___healthy",
    24: "Raspberry___healthy",
    25: "Soybean___healthy",
    26: "Squash___Powdery_mildew",
    27: "Strawberry___Leaf_scorch",
    28: "Strawberry___healthy",
    29: "Tomato___Bacterial_spot",
    30: "Tomato___Early_blight",
    31: "Tomato___Late_blight",
    32: "Tomato___Leaf_Mold",
    33: "Tomato___Septoria_leaf_spot",
    34: "Tomato___Spider_mites Two-spotted_spider_mite",
    35: "Tomato___Target_Spot",
    36: "Tomato___Tomato_Yellow_Leaf_Curl_Virus",
    37: "Tomato___Tomato_mosaic_virus",
    38: "Tomato___healthy"
}

# Load the model if it exists
if os.path.exists(model_path):
    model = tf.keras.models.load_model(model_path, custom_objects=custom_objects)
    print("Model loaded successfully.")
else:
    print(f"Model file not found at {model_path}. Please upload the model.")

# Function to preprocess input image
def preprocess_image(image_data, img_size=224):
    img = Image.open(BytesIO(image_data))
    img = img.resize((img_size, img_size))
    img_array = img_to_array(img)
    img_array = img_array / 255.0
    img_array = np.expand_dims(img_array, axis=0)
    return img_array

# Predict function
def predict_image(image_data):
    preprocessed_image = preprocess_image(image_data)
    predictions = model.predict(preprocessed_image)
    class_idx = int(np.argmax(predictions, axis=1)[0])  # Convert to int for JSON serialization
    confidence = float(predictions[0][class_idx])      # Convert to float for JSON serialization
    class_label = class_labels.get(class_idx, "Unknown")

    # Fetch additional data from external API
    try:
        response = requests.get(f"https://navpan2-sarva-ai-back.hf.space/kotlinback/{class_label}")
        external_data = response.json() if response.status_code == 200 else {"error": "Failed to fetch external data"}
    except Exception as e:
        external_data = {"error": str(e)}

    return external_data
    

# Route for health check
@app.get("/health")
async def api_health_check():
    return JSONResponse(content={"status": "Service is running"})

# Route for prediction using image via API
@app.post("/predict")
async def api_predict_image(file: UploadFile = File(...)):
    try:
        image_data = await file.read()
        prediction = predict_image(image_data)
        return JSONResponse(content={"prediction": prediction})
    except Exception as e:
        return JSONResponse(content={"error": str(e)})

# Run the FastAPI app
if __name__ == "__main__":
    uvicorn.run(app, host="0.0.0.0", port=7860)