huvimal/Vietnamese-document-intelligence-platform

1import streamlit as st
2from PIL import Image
3import numpy as np
4import easyocr
5
6st.set_page_config(
7    page_title="Vietnamese Document Intelligence",
8    layout="wide"
9)
10
11st.title("📄 Vietnamese Document Intelligence Platform")
12
13st.write("Upload an image to extract text.")
14
15@st.cache_resource
16def load_reader():
17    reader = easyocr.Reader(
18        ['en'],
19        gpu=False
20    )
21    return reader
22
23try:
24    reader = load_reader()
25
26    uploaded_file = st.file_uploader(
27        "Upload Image",
28        type=["png", "jpg", "jpeg"]
29    )
30
31    if uploaded_file:
32
33        image = Image.open(uploaded_file)
34
35        st.image(
36            image,
37            caption="Uploaded Image",
38            use_container_width=True
39        )
40
41        image_np = np.array(image)
42
43        with st.spinner("Extracting text..."):
44            results = reader.readtext(image_np)
45
46        extracted_text = "\n".join(
47            [item[1] for item in results]
48        )
49
50        st.subheader("OCR Result")
51
52        st.text_area(
53            "Extracted Text",
54            extracted_text,
55            height=300
56        )
57
58except Exception as e:
59    st.error(f"Error: {str(e)}")