Create app.py

app.py

@@ -0,0 +1,76 @@
+import os
+from io import BytesIO
+import gradio as gr
+import numpy as np
+import cv2
+from PIL import Image
+from dotenv import load_dotenv
+import google.generativeai as genai
+from model import classify_image  # Assuming this is a custom model for classification
+
+# Load environment variables for Google Gemini API key
+load_dotenv()
+genai.configure(api_key=os.getenv("GOOGLE_API_KEY"))
+
+# Input prompt for Gemini model
+input_prompt = """
+    "You are an expert in computer vision and agriculture who can easily predict the disease of the plant. "
+    "Analyze the following image and provide 7 short outputs in a structured format: "
+    "1. Crop: , "
+    "2. Infected or Healthy: , "
+    "3. Type of disease (if any): , "
+    "4. Confidence out of 100% whether image is healthy or infected: , "
+    "5. Reason for the disease such as whether it is happening due to fungus, bacteria, insect bite, poor nutrition, etc.: , "
+    "6. Plant Growth Stage: , "
+    "7. Pest Life Stage: ."
+"""
+
+# Function to get a response from the Google Gemini API
+def get_gemini_response(image):
+    model = genai.GenerativeModel('gemini-1.5-pro')
+
+    # Convert PIL image to bytes
+    bytes_io = BytesIO()
+    image.save(bytes_io, format='PNG')
+    bytes_data = bytes_io.getvalue()
+
+    # Send image and prompt to Gemini API
+    response = model.generate_content([input_prompt, image])
+
+    # Extract the relevant response (mock response logic for now)
+    return response.text
+
+# Image classification using custom model (e.g., for detecting specific plant diseases)
+def classify_crop_image(img):
+    img = cv2.resize(img, (224, 224))
+    img = img / 255.0
+    img = np.expand_dims(img, axis=0)
+
+    # Use the custom classification model to predict the disease
+    label, accuracy = classify_image(img)  # Assuming 'classify_image' returns label and accuracy
+    return {label: accuracy}
+
+# Function to handle the uploaded image, predict crop health, and provide a structured output
+def predict_crop_health(uploaded_image):
+    # Pass the image to the custom classifier for plant disease prediction
+    classification_result = classify_crop_image(np.array(uploaded_image))
+
+    # Pass the image to the Gemini API for detailed disease analysis
+    gemini_response = get_gemini_response(uploaded_image)
+
+    # Combine results from both models (custom and Gemini)
+    return f"Classification Result: {classification_result}\n\nGemini Response: {gemini_response}"
+
+# Define the Gradio interface: Inputs and Outputs
+inputs = gr.Image(type="pil", label="Upload Crop Image")
+outputs = gr.Textbox(label="Crop Disease Predictor", lines=10)
+
+# Launch the Gradio interface
+gr.Interface(
+    fn=predict_crop_health,
+    inputs=inputs,
+    outputs=outputs,
+    title="Crop Disease Prediction App",
+    description="Upload an image of a crop to predict its disease and get treatment suggestions.",
+    live=False
+).launch()