app.py

import streamlit as st
import tensorflow as tf
import cv2
import numpy as np
from PIL import Image, ImageOps
#from io import BytesIO

@st.cache_resource()
def load_model():
    model=tf.keras.models.load_model('./hip_impant_model.h5')
    return model
with st.spinner('Model is being loaded..'):
    model=load_model()

st.write("""
         # Image Classification
         """
         )

file = st.file_uploader("Upload an X-ray image", type= ['png', 'jpg'])


def model_prediction(image, model):
    resize = tf.image.resize(image, (256,256))
    yhat = model.predict(np.expand_dims(resize/255, 0))
    if(yhat>0.5):
        result = "Prediction is loose"
    else:
        result = "Prediction is control"
    return result

def model_prediction(image, model):
    img = cv2.imread(image)
    plt.imshow(img)
    plt.show()
    resize = tf.image.resize(img, (256,256))
    yhat = model.predict(np.expand_dims(resize/255, 0))
    if(yhat>0.5):
        result = "Prediction is loose"
    else:
        result = "Prediction is control"
    return result


def on_click():
    if file is None:
        st.text("Please upload an image file")
    else:
        image = Image.open(file)
        st.image(image, use_column_width=True)
        predictions = model_prediction(image, model)
        st.write(prediction)
        print(
            "This image most likely belongs to {}."
            .format(prediction)
)

st.button('Predict', on_click=on_click)