Create app.py

app.py

@@ -0,0 +1,89 @@
+# app.py
+import streamlit as st
+import tensorflow as tf
+from PIL import Image
+import numpy as np
+
+# Treatment database (expand with your own data)
+TREATMENTS = {
+    "Tomato_Early_Blight": {
+        "summary": "Early Blight (Alternaria solani)",
+        "treatment": [
+            "Remove infected leaves immediately",
+            "Apply chlorothalonil (1 tbsp/gallon) weekly",
+            "Improve air circulation around plants"
+        ]
+    },
+    "Potato_Late_Blight": {
+        "summary": "Late Blight (Phytophthora infestans)",
+        "treatment": [
+            "Destroy infected plants",
+            "Apply copper-based fungicides every 7 days",
+            "Avoid overhead watering"
+        ]
+    }
+}
+
+# Load pre-trained model (using open-source model from TF Hub)
+@st.cache_resource
+def load_model():
+    try:
+        model = tf.keras.models.load_model('model/plant_disease_model.h5')
+    except:
+        # Fallback to pre-trained EfficientNet
+        model = tf.keras.applications.EfficientNetB3(
+            input_shape=(256,256,3),
+            weights='imagenet',
+            include_top=False
+        )
+        model.trainable = False
+        model = tf.keras.Sequential([
+            model,
+            tf.keras.layers.GlobalAveragePooling2D(),
+            tf.keras.layers.Dense(38, activation='softmax')
+        ])
+    return model
+
+def preprocess_image(image):
+    img = Image.open(image).convert('RGB')
+    img = img.resize((256, 256))
+    img_array = tf.keras.preprocessing.image.img_to_array(img)
+    img_array = tf.expand_dims(img_array, 0) / 255.0
+    return img_array
+
+def main():
+    st.title("🌱 Plant Disease Detection & Treatment System")
+    st.markdown("Upload a leaf image for disease diagnosis and treatment recommendations")
+
+    uploaded_file = st.file_uploader("Choose a leaf image...", type=["jpg", "jpeg", "png"])
+    
+    if uploaded_file is not None:
+        # Display image
+        st.image(uploaded_file, caption='Uploaded Leaf Image', width=300)
+        
+        # Preprocess and predict
+        img_array = preprocess_image(uploaded_file)
+        model = load_model()
+        predictions = model.predict(img_array)
+        predicted_class = np.argmax(predictions[0])
+        
+        # Get class name (replace with your actual class names)
+        class_names = ["Tomato_Early_Blight", "Potato_Late_Blight"]  # Add all 38 classes
+        
+        # Display results
+        st.subheader("🔍 Diagnosis Summary")
+        disease = class_names[predicted_class]
+        st.success(f"Detected Disease: **{disease.replace('_', ' ')}**")
+        st.write(f"Confidence: {predictions[0][predicted_class]*100:.2f}%")
+        
+        # Show treatment
+        st.subheader("💊 Recommended Treatment")
+        if disease in TREATMENTS:
+            st.markdown(f"**{TREATMENTS[disease]['summary']}**")
+            for step in TREATMENTS[disease]["treatment"]:
+                st.markdown(f"- {step}")
+        else:
+            st.warning("Treatment information not found for this disease")
+
+if __name__ == "__main__":
+    main()