import os from fastapi import FastAPI, File, UploadFile from fastapi.responses import JSONResponse from tensorflow.keras.preprocessing.image import img_to_array from PIL import Image import tensorflow as tf import numpy as np import io import tempfile from ultralytics import YOLO # Make sure ultralytics is installed # Load YOLO model yolo_model_path = 'yolo_test.pt' if not os.path.exists(yolo_model_path): raise FileNotFoundError(f"YOLO model file not found at {yolo_model_path}") yolo_model = YOLO(yolo_model_path) # Load Keras model keras_model_path = 'Custom_layer.h5' if not os.path.exists(keras_model_path): raise FileNotFoundError(f"Keras model file not found at {keras_model_path}") keras_model = tf.keras.models.load_model( keras_model_path, custom_objects={'Precision': tf.keras.metrics.Precision, 'Recall': tf.keras.metrics.Recall} ) app = FastAPI(title="Leaf Disease Classifier API") def prepare_image(image, target_size): if image.mode != "RGB": image = image.convert("RGB") image = image.resize(target_size) image = img_to_array(image) image = np.expand_dims(image, axis=0) image = image / 255.0 # Normalize the image return image @app.get("/") async def home(): return {"message": "Welcome to the Leaf Disease Classifier API"} @app.post("/predict") async def predict(file: UploadFile = File(...)): temp_file_path = None try: contents = await file.read() image = Image.open(io.BytesIO(contents)) # Save to a unique temp file for YOLO with tempfile.NamedTemporaryFile(delete=False, suffix=".jpg") as temp_file: temp_file.write(contents) temp_file_path = temp_file.name except Exception as e: return JSONResponse(content={"error": str(e)}, status_code=400) # Run YOLO detection try: results = yolo_model.predict(source=temp_file_path, conf=0.4, verbose=False) boxes = results[0].boxes mulberry_detected = False if boxes is not None and len(boxes) > 0: for box in boxes: cls_id = int(box.cls[0]) if cls_id == 0: # Assuming class 0 is 'mulberry_leaf' mulberry_detected = True break if not mulberry_detected: os.remove(temp_file_path) return JSONResponse(content={"error": "Upload only mulberry leaf!"}, status_code=400) except Exception as e: if temp_file_path and os.path.exists(temp_file_path): os.remove(temp_file_path) return JSONResponse(content={"error": f"YOLO error: {str(e)}"}, status_code=500) # Run Keras classifier try: image = prepare_image(image, target_size=(224, 224)) prediction = keras_model.predict(image) class_idx = np.argmax(prediction) class_names = ['Disease Free leaves', 'Leaf Rust', 'Leaf spot'] # Replace with your actual class names predicted_class = class_names[class_idx] confidence = round(100 * (np.max(prediction[0])), 2) except Exception as e: if temp_file_path and os.path.exists(temp_file_path): os.remove(temp_file_path) return JSONResponse(content={"error": str(e)}, status_code=500) if temp_file_path and os.path.exists(temp_file_path): os.remove(temp_file_path) return { "predicted_class": predicted_class, "confidence": f"{confidence}%" }