Update app.py

app.py

@@ -1,59 +1,26 @@
-
 import streamlit as st
-from tensorflow.keras.models import load_model, model_from_json
+from tensorflow.keras.models import load_model
 from tensorflow.keras.preprocessing import image
 import numpy as np
 from PIL import Image
-import requests
-import zipfile
-import os
 
-# Download and extract the filter model files
+# Hugging Face model path for the filter model
+FILTER_MODEL_PATH = "https://huggingface.co/your-repo-name/path/to/filter_model_folder/"
+
+# Function to load the filter model directly
 @st.cache_resource
 def load_filter_model():
-    # URL of the zipped filter model folder
-    filter_model_url = "https://huggingface.co/spaces/Ten7/Maize_and_Cassava_Crop_Disease_Classification/blob/main/model_checkpoint.zip"
-
-    # Download the zipped folder
-    zip_path = "model_checkpoint.zip"
-    with open(zip_path, "wb") as f:
-        f.write(requests.get(filter_model_url).content)
-
-    # Extract the zipped folder
-    extract_dir = "filter_model"
-    with zipfile.ZipFile(zip_path, 'r') as zip_ref:
-        zip_ref.extractall(extract_dir)
-
-    # Load the filter model
-    config_path = os.path.join(extract_dir, "config.json")
-    weights_path = os.path.join(extract_dir, "model.weights.h5")
-
-    with open(config_path, "r") as f:
-        config_json = f.read()
-    filter_model = model_from_json(config_json)
-    filter_model.load_weights(weights_path)
-    return filter_model
+    model = load_model(FILTER_MODEL_PATH)
+    return model
 
+# Load the filter model
 filter_model = load_filter_model()
 
-# Load the disease models
-@st.cache_resource
-def load_disease_models():
-    corn_model = load_model('corn_model_mobilenetv2.h5')
-    cassava_model = load_model('cassava_disease_model.h5')
-    return corn_model, cassava_model
-
-corn_model, cassava_model = load_disease_models()
+# Load maize and cassava models
+corn_model = load_model('corn_model_mobilenetv2.h5')
+cassava_model = load_model('cassava_disease_model.h5')
 
-# Function to predict using the filter model
-def filter_prediction(img, model):
-    img = image.img_to_array(img)
-    img = np.expand_dims(img, axis=0)
-    img = img / 255.0  # Normalize image
-    prediction = model.predict(img)
-    return np.argmax(prediction)
-
-# Function to predict crop disease
+# Function to predict an uploaded image
 def predict_image(img, model, class_names):
     img = image.img_to_array(img)
     img = np.expand_dims(img, axis=0)
@@ -68,73 +35,83 @@ def home_page():
     st.image("1093908406.jpg", use_column_width=True)
     st.write("""
         Welcome to the Crop Disease Identification App. This tool helps farmers identify diseases in maize and cassava crops by analyzing images of leaves.
+        
         Use the navigation menu to explore the app:
         - **Home:** Overview of the app
         - **About:** Information about the app and its purpose
-        - **Prediction:** Upload an image and get a disease diagnosis along with recommendations
+        - **Prediction:** Upload an image and get a disease diagnosis along with recommendations.
     """)
 
 def about_page():
     st.title("About")
     st.image("IMG_20240727_143208_444.jpg", use_column_width=True)
     st.write("""
-        This application assists farmers in diagnosing common diseases in maize and cassava crops.
-        It uses deep learning models trained on leaf images to identify diseases like blight, common rust, gray leaf spot (maize),
-        bacterial blight, green mottle, brown streak, and mosaic (cassava).
-        If you have questions or feedback, reach out to the developer at:
-        E-mail: olorunnisholato7@gmail.com
+        This application helps diagnose common diseases in maize and cassava crops. It uses deep learning models trained on thousands of leaf images to accurately identify diseases.
+        
+        Diseases include:
+        - Maize: Blight, Common Rust, Gray Leaf Spot
+        - Cassava: Bacterial Blight, Green Mottle, Brown Streak, Mosaic
+        
+        For questions or feedback, email: olorunnisholato7@gmail.com
     """)
 
 def prediction_page():
     st.title("Crop Disease Prediction")
-
+    crop_type = st.selectbox("Select the crop:", ["Maize", "Cassava"])
     uploaded_file = st.file_uploader("Upload an image of the leaf...", type=["jpg", "jpeg", "png"])
 
     if uploaded_file is not None:
-        img = Image.open(uploaded_file).resize((128, 128))
+        img = Image.open(uploaded_file)
+
+        # Set image size for filtering and crop-specific models
+        img_resized = img.resize((128, 128))  # For the filter model
+        img_array = np.expand_dims(image.img_to_array(img_resized) / 255.0, axis=0)
+
+        # Apply the filter model
+        filter_prediction = filter_model.predict(img_array)
+        filter_class = np.argmax(filter_prediction)
+        if filter_class == 0:  # Assuming 0 is "Non-Leaf"
+            st.warning("The uploaded image does not appear to contain a leaf. Please try again.")
+            return
+
+        # Resize for crop-specific models
+        target_size = (128, 128) if crop_type == "Maize" else (224, 224)
+        img = img.resize(target_size)
+
+        # Display the uploaded image
         st.image(img, caption='Uploaded Image', use_column_width=True)
 
         if st.button('Predict'):
-            st.write("Classifying with the filter model...")
-
-            # Use the filter model to determine if the image is a leaf
-            filter_result = filter_prediction(img, filter_model)
-            if filter_result == 0:
-                st.error("The uploaded image is not a leaf. Please upload a valid leaf image.")
+            st.write("Classifying...")
+            if crop_type == "Maize":
+                class_names = ['Blight', 'Common_Rust', 'Gray_Leaf_Spot', 'Healthy']
+                predicted_class, class_name = predict_image(img, corn_model, class_names)
             else:
-                st.success("The uploaded image is a leaf. Proceeding with disease detection.")
-
-                # Select the crop type
-                crop_type = st.selectbox("Select the crop:", ["Maize", "Cassava"])
-
-                # Predict disease
-                if crop_type == "Maize":
-                    img = img.resize((128, 128))
-                    class_names = ['Blight', 'Common_Rust', 'Gray_Leaf_Spot', 'Healthy']
-                    predicted_class, class_name = predict_image(img, corn_model, class_names)
-                else:
-                    img = img.resize((224, 224))
-                    class_names = ['Cassava_bacterial_blight', 'Cassava_brown_streak_disease', 
-                                   'Cassava_green_mottle', 'Cassava_healthy', 'Cassava_mosaic_disease']
-                    predicted_class, class_name = predict_image(img, cassava_model, class_names)
-
-                # Display prediction
-                st.write(f"Predicted Class: {class_name}")
-
-                # Recommendation
-                if class_name == 'Healthy' or class_name == 'Cassava_healthy':
-                    st.success("The leaf is healthy. No action needed.")
-                else:
-                    recommendations = {
-                        'Blight': "Remove infected plants and apply fungicide.",
-                        'Common_Rust': "Use resistant varieties and fungicides.",
-                        'Gray_Leaf_Spot': "Ensure crop rotation and apply fungicides.",
-                        'Cassava_bacterial_blight': "Use disease-free planting material.",
-                        'Cassava_brown_streak_disease': "Plant resistant cassava varieties.",
-                        'Cassava_green_mottle': "Ensure good field sanitation.",
-                        'Cassava_mosaic_disease': "Use resistant varieties and good field hygiene."
-                    }
-                    st.warning(recommendations[class_name])
+                class_names = [
+                    'Cassava_bacterial_blight',
+                    'Cassava_brown_streak_disease',
+                    'Cassava_green_mottle',
+                    'Cassava_healthy',
+                    'Cassava_mosaic_disease',
+                ]
+                predicted_class, class_name = predict_image(img, cassava_model, class_names)
+
+            st.write(f"Predicted Class: {class_name}")
+
+            # Recommendations based on predictions
+            if class_name in ['Healthy', 'Cassava_healthy']:
+                st.success("The leaf is healthy. No action needed.")
+            else:
+                recommendations = {
+                    'Blight': "Remove infected plants, apply fungicides, and rotate crops.",
+                    'Common_Rust': "Use resistant varieties and apply fungicides as needed.",
+                    'Gray_Leaf_Spot': "Ensure crop rotation, balanced nitrogen, and apply fungicides.",
+                    'Cassava_bacterial_blight': "Use disease-free planting material and remove affected plants.",
+                    'Cassava_brown_streak_disease': "Plant resistant varieties and manage whiteflies.",
+                    'Cassava_green_mottle': "Ensure good sanitation and use resistant varieties.",
+                    'Cassava_mosaic_disease': "Use resistant varieties and practice good field hygiene.",
+                }
+                st.warning(recommendations[class_name])
 
 # Main app
 def main():
@@ -150,3 +127,4 @@ def main():
 
 if __name__ == "__main__":
     main()
+