Update app.py

app.py

@@ -164,7 +164,7 @@ def preprocess_image(image_data):
         image = Image.open(io.BytesIO(image_data)).convert("RGB")
         
         # Crop (x1=450, y1=400, x2=1090, y2=1060)
-        crop_box = (450, 140, 1090, 800)  # (left, top, right, bottom)
+        crop_box = (450, 400, 1090, 1060)  
         image = image.crop(crop_box)
 
         # Resize to model input size
@@ -186,8 +186,7 @@ def preprocess_image(image_data):
 
 def preprocess_image_test(image_data):
     """
-    Preprocess image to match training pipeline:
-    - Crop ROI from ESP32 frame (400x296)
+    Preprocess image WITHOUT cropping for testing:
     - Resize to 224x224
     - Convert to numpy array, add batch dim
     """
@@ -195,7 +194,7 @@ def preprocess_image_test(image_data):
         # Convert bytes to PIL Image
         image = Image.open(io.BytesIO(image_data)).convert("RGB")
         
-        # Resize to model input size
+        # Resize to model input size (NO CROPPING)
         image = image.resize((224, 224), Image.Resampling.LANCZOS)
 
         # Normalize and expand dims
@@ -204,7 +203,6 @@ def preprocess_image_test(image_data):
         # Add batch dimension
         image = np.expand_dims(image, axis=0)
 
-
         # Model has Rescaling(1./255) layer, so no manual normalization
         return image
 
@@ -212,6 +210,7 @@ def preprocess_image_test(image_data):
         logger.error(f"Image preprocessing error: {e}")
         raise HTTPException(status_code=400, detail=f"Image preprocessing failed: {e}")
 
+
 @app.get("/health")
 async def health_check():
     """Health check endpoint"""
@@ -240,6 +239,7 @@ async def health_check():
             "model_loaded": model is not None,
             "classes": class_labels
         }
+        
 @app.on_event("startup")
 async def startup_event():
     """Load model on startup"""
@@ -314,6 +314,57 @@ async def classify_image(file: UploadFile = File(...)):
             content={"error": f"Classification failed: {str(e)}"}
         )
 
+@app.post("/classify_test")
+async def classify_image_test(file: UploadFile = File(...)):  # CHANGED FUNCTION NAME
+    """
+    Test classification endpoint WITHOUT cropping
+    
+    Expected usage:
+    curl -X POST -F "file=@image.jpg" https://your-space-url.hf.space/classify_test
+    
+    Returns:
+        JSON: {"label": "plastic"} or {"error": "message"}
+    """
+    try:
+        # Validate file type
+        if not file.content_type or not file.content_type.startswith('image/'):
+            logger.warning(f"Invalid file type: {file.content_type}")
+            raise HTTPException(status_code=400, detail="File must be an image")
+        
+        # Read image data
+        image_data = await file.read()
+        if len(image_data) == 0:
+            raise HTTPException(status_code=400, detail="Empty image file")
+            
+        logger.info(f"TEST MODE - Processing image: {file.filename}, size: {len(image_data)} bytes")
+        
+        # Preprocess image WITHOUT cropping
+        processed_image = preprocess_image_test(image_data)
+        
+        # Make prediction
+        predictions = model.predict(processed_image, verbose=0)
+        predicted_class_index = np.argmax(predictions[0])
+        predicted_class = class_labels[predicted_class_index]
+        confidence = float(predictions[0][predicted_class_index])
+        
+        logger.info(f"TEST MODE - Prediction: {predicted_class} (confidence: {confidence:.3f})")
+        
+        # Return simple response for ESP32 - match your ESP32 expectation exactly
+        return {
+            "label": predicted_class.capitalize(),
+            "confidence": round(confidence, 3),
+            "mode": "test_no_crop"
+        }
+        
+    except HTTPException:
+        raise
+    except Exception as e:
+        logger.error(f"Classification error: {str(e)}")
+        return JSONResponse(
+            status_code=500, 
+            content={"error": f"Classification failed: {str(e)}"}
+        )
+
 @app.post("/classify/detailed")
 async def classify_detailed(file: UploadFile = File(...)):
     """
@@ -356,53 +407,6 @@ async def classify_detailed(file: UploadFile = File(...)):
         logger.error(f"Detailed classification error: {str(e)}")
         raise HTTPException(status_code=500, detail=f"Classification failed: {str(e)}")
 
-@app.post("/classify_test")
-async def classify_image(file: UploadFile = File(...)):
-    """
-    Main classification endpoint for ESP32
-    
-    Expected usage:
-    curl -X POST -F "file=@image.jpg" https://your-space-url.hf.space/classify_test
-    
-    Returns:
-        JSON: {"label": "plastic"} or {"error": "message"}
-    """
-    try:
-        # Validate file type
-        if not file.content_type or not file.content_type.startswith('image/'):
-            logger.warning(f"Invalid file type: {file.content_type}")
-            raise HTTPException(status_code=400, detail="File must be an image")
-        
-        # Read image data
-        image_data = await file.read()
-        if len(image_data) == 0:
-            raise HTTPException(status_code=400, detail="Empty image file")
-            
-        logger.info(f"Processing image: {file.filename}, size: {len(image_data)} bytes")
-        
-        # Preprocess image
-        processed_image = preprocess_image_test(image_data)
-        
-        # Make prediction
-        predictions = model.predict(processed_image, verbose=0)
-        predicted_class_index = np.argmax(predictions[0])
-        predicted_class = class_labels[predicted_class_index]
-        confidence = float(predictions[0][predicted_class_index])
-        
-        logger.info(f"Prediction: {predicted_class} (confidence: {confidence:.3f})")
-        
-        # Return simple response for ESP32 - match your ESP32 expectation exactly
-        return {"label": predicted_class.capitalize()}  # Capitalize to match your ESP32 labels
-        
-    except HTTPException:
-        raise
-    except Exception as e:
-        logger.error(f"Classification error: {str(e)}")
-        return JSONResponse(
-            status_code=500, 
-            content={"error": f"Classification failed: {str(e)}"}
-        )
-
 @app.get("/info")
 async def get_info():
     """API information endpoint"""