| import tensorflow as tf |
| from tensorflow import keras |
| from PIL import Image |
| import numpy as np |
| import gradio as gr |
|
|
| model_path = './cnn_neumonia.keras' |
| IMG_HEIGHT = 64 |
| IMG_WIDTH = 64 |
| |
| try: |
| model = keras.models.load_model(model_path) |
| print("Modelo cargado exitosamente.") |
| except Exception as e: |
| print(f"Error al cargar el modelo: {e}") |
| |
| exit() |
| |
| def preprocess_image_for_prediction(image: Image.Image, target_size=(IMG_HEIGHT, IMG_WIDTH)): |
| img = image.convert('RGB') |
| img = img.resize(target_size) |
| img_array = np.array(img) |
| img_array = np.expand_dims(img_array, axis=0) |
| img_array = img_array / 255.0 |
| return img_array |
|
|
| def predict_pneumonia(image_path_or_object): |
| if isinstance(image_path_or_object, str): |
| |
| image = Image.open(image_path_or_object) |
| else: |
| |
| image = image_path_or_object |
|
|
| |
| processed_image = preprocess_image_for_prediction(image, target_size=(IMG_HEIGHT, IMG_WIDTH)) |
|
|
| |
| prediction = model.predict(processed_image, verbose=0) |
|
|
| |
| probability = prediction[0][0] |
| if probability > 0.5: |
| message = f"隆Tiene Neumon铆a! (Probabilidad: {probability:.2f})" |
| else: |
| message = f"No tiene Neumon铆a (Probabilidad: {probability:.2f})" |
|
|
| return message |
|
|
| iface = gr.Interface( |
| fn=predict_pneumonia, |
| title="Predicci贸n de Neumon铆a por Radiograf铆a", |
| description="Carga una imagen de radiograf铆a y predice si contiene neumon铆a.", |
| inputs=gr.Image(type="pil",label="Sube una imagen de rayos X de t贸rax"), |
| outputs=gr.Textbox(label="Diagnostico"), |
| examples=[ |
| './normal1.jpeg', |
| './normal2.jpeg', |
| './pneumonia1.jpeg', |
| './pneumonia2.jpeg', |
| ] |
| ) |
|
|
| iface.launch(debug=True,share=True) |
