Update app.py

app.py

@@ -1,44 +1,136 @@
+"""
+Animal Type Classification App
+A robust Gradio application for classifying animals using YOLOv8
+"""
+
 import gradio as gr
 from ultralytics import YOLO
 from PIL import Image
 import numpy as np
+import logging
+import sys
+import os
+from typing import Optional, Tuple
 
-# Load your pretrained YOLOv8 model (replace with your model path if customized)
-model = YOLO("yolov8n.pt")  # Using tiny YOLOv8 pretrained weights for example
-
-# Define the class names according to your trained model classes
-class_names = ["butterfly", "chicken", "elephant", "horse", "spider", "squirrel"]
+# ============================================================================
+# LOGGING CONFIGURATION
+# ============================================================================
+logging.basicConfig(
+    level=logging.INFO,
+    format='%(asctime)s - %(name)s - %(levelname)s - %(message)s',
+    handlers=[
+        logging.FileHandler('animal_classifier.log'),
+        logging.StreamHandler(sys.stdout)
+    ]
+)
+logger = logging.getLogger(__name__)
 
-def classify_animal(image):
-    # Convert numpy array to PIL Image
-    if isinstance(image, np.ndarray):
-        image = Image.fromarray(image)
-    # YOLO expects numpy array input, convert back to np array
-    img_np = np.array(image)
+# ============================================================================
+# CONFIGURATION
+# ============================================================================
+MODEL_PATH = "best_animal_classifier.pt"
+CLASS_NAMES = ["butterfly", "chicken", "elephant", "horse", "spider", "squirrel"]
+CONFIDENCE_THRESHOLD = 0.5
+MIN_DETECTIONS = 1
 
-    # Perform prediction
-    results = model(img_np)
+# ============================================================================
+# GLOBAL MODEL VARIABLE
+# ============================================================================
+model = None
 
-    # Extract top prediction
-    if results and len(results) > 0:
-        boxes = results[0].boxes
-        if boxes is not None and len(boxes) > 0:
-            # Get the highest confidence prediction
-            best_box = boxes[0]
-            conf = best_box.conf[0].item()
-            class_id = int(best_box.cls[0].item())
-            class_name = class_names[class_id] if class_id < len(class_names) else "Unknown"
-            return f"Predicted: {class_name} (Confidence: {conf:.2f})"
-    return "No animal detected."
+# ============================================================================
+# MODEL INITIALIZATION WITH EXCEPTION HANDLING
+# ============================================================================
+def load_model() -> Optional[YOLO]:
+    """
+    Load the YOLO model with comprehensive error handling.
+    
+    Returns:
+        YOLO: Loaded model object or None if loading fails
+    """
+    global model
+    
+    try:
+        logger.info(f"Attempting to load model from: {MODEL_PATH}")
+        
+        # Check if model file exists
+        if not os.path.exists(MODEL_PATH):
+            logger.error(f"Model file not found at: {MODEL_PATH}")
+            raise FileNotFoundError(
+                f"Model file '{MODEL_PATH}' does not exist. "
+                f"Please ensure the file is in the correct location."
+            )
+        
+        # Check if file has read permissions
+        if not os.access(MODEL_PATH, os.R_OK):
+            logger.error(f"No read permission for model file: {MODEL_PATH}")
+            raise PermissionError(
+                f"No read permission for model file: {MODEL_PATH}"
+            )
+        
+        # Load the model
+        model = YOLO(MODEL_PATH)
+        logger.info("✅ Model loaded successfully!")
+        
+        return model
+    
+    except FileNotFoundError as e:
+        logger.error(f"FileNotFoundError: {e}")
+        return None
+    
+    except PermissionError as e:
+        logger.error(f"PermissionError: {e}")
+        return None
+    
+    except Exception as e:
+        logger.error(f"Unexpected error loading model: {type(e).__name__}: {e}")
+        return None
 
-# Build Gradio interface
-iface = gr.Interface(
-    fn=classify_animal, 
-    inputs=gr.Image(type="numpy", label="Upload Animal Image"),
-    outputs="text",
-    title="Real-Time Animal Type Classification",
-    description="Upload an image of an animal to classify it among butterfly, chicken, elephant, horse, spider, and squirrel."
-)
 
-if __name__ == "__main__":
-    iface.launch()
+# ============================================================================
+# CLASSIFICATION FUNCTION WITH ROBUST ERROR HANDLING
+# ============================================================================
+def classify_animal(image: Optional[np.ndarray]) -> str:
+    """
+    Classify an animal in the provided image using YOLOv8.
+    
+    Args:
+        image (Optional[np.ndarray]): Input image as numpy array or PIL Image
+    
+    Returns:
+        str: Classification result with confidence score or error message
+    """
+    
+    try:
+        # ========== INPUT VALIDATION ==========
+        if image is None:
+            logger.warning("No image provided")
+            return "❌ Error: No image provided. Please upload an image."
+        
+        logger.info("Image received for classification")
+        
+        # ========== MODEL AVAILABILITY CHECK ==========
+        if model is None:
+            logger.error("Model is not loaded")
+            return "❌ Critical Error: Model not loaded. Please restart the application."
+        
+        # ========== IMAGE TYPE CONVERSION ==========
+        try:
+            if isinstance(image, np.ndarray):
+                # Validate numpy array dimensions
+                if image.ndim not in [2, 3, 4]:
+                    logger.error(f"Invalid image dimensions: {image.ndim}")
+                    return "❌ Error: Invalid image dimensions. Expected 2D, 3D, or 4D array."
+                
+                # Validate data type
+                if not np.issubdtype(image.dtype, np.integer):
+                    logger.warning(f"Unexpected image dtype: {image.dtype}, attempting conversion")
+                    image = image.astype('uint8')
+                
+                # Convert to PIL Image
+                image_pil = Image.fromarray(image.astype('uint8'))
+                logger.debug("Converted numpy array to PIL Image")
+            
+            elif isinstance(image, Image.Image):
+                image_pil = image
+                lo