| import streamlit as st | |
| import numpy as np | |
| from PIL import Image | |
| import cv2 | |
| import tensorflow as tf | |
| st.title("Hello Parimal") | |
| def data_preprocessing(img): | |
| img = cv2.cvtColor(img, cv2.COLOR_RGB2GRAY) | |
| img = np.resize(img, [1, 28, 28]) | |
| img = img/255 | |
| return img | |
| image = st.file_uploader("Upload files", type=["jpeg", "png", "jpg", "webp"]) | |
| model = tf.keras.models.load_model("mnist.h5") | |
| if image is not None: | |
| img = Image.open(image) | |
| img = np.array(img) | |
| st.image(img, caption="Uploaded Image", use_column_width=True) | |
| images = data_preprocessing(img) | |
| predictions = model.predict(images) | |
| predictions = np.argmax(predictions) | |
| st.write(f"The Predicted Value: {predictions}") | |