Fix 'tuple indices must be integers' error in ELA analysis

- perform_ela_analysis returns tuple (is_suspicious, confidence, details)
- Updated to properly unpack all 3 return values
- Convert diff_image numpy array to PIL Image for display
- Handle case where ELA only works on JPEG (returns error for PNG)
- Fixes runtime error when analyzing images in Detect tab

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

app.py

@@ -128,8 +128,8 @@ def detect_hidden_data(image, sensitivity):
         # Perform analysis (returns: is_suspicious, confidence, details)
         is_suspicious, confidence, details = rat_finder.analyze_image(image_path, sensitivity=sensitivity_str)
 
-        # Generate ELA visualization
-        ela_result = rat_finder.perform_ela_analysis(image_path)
+        # Generate ELA visualization (returns: is_suspicious, confidence, details)
+        ela_is_suspicious, ela_confidence, ela_details = rat_finder.perform_ela_analysis(image_path)
 
         # Clean up
         os.unlink(image_path)
@@ -183,11 +183,14 @@ def detect_hidden_data(image, sensitivity):
 Use the "Extract Data" tab if you suspect LSB steganography!
 """
 
-        # Return ELA plot if available
-        if ela_result['success'] and ela_result['ela_image']:
-            return ela_result['ela_image'], result_text
+        # Return ELA plot if available (only for JPEG images)
+        ela_image = None
+        if isinstance(ela_details, dict) and 'diff_image' in ela_details and 'error' not in ela_details:
+            # Convert numpy array to PIL Image for display
+            diff_array = ela_details['diff_image']
+            ela_image = Image.fromarray(diff_array)
 
-        return None, result_text
+        return ela_image, result_text
 
     except Exception as e:
         if 'image_path' in locals() and os.path.exists(image_path):