File size: 6,244 Bytes
5e66110 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 |
from transformers import AutoModel, AutoTokenizer,AutoProcessor
import streamlit as st
import os
from PIL import Image
import torch
from torchvision import io
import torchvision.transforms as transforms
import random
import easyocr
import numpy as np
def start():
st.session_state.start = True
def reset():
del st.session_state['start']
@st.cache_data
def get_text(image_file, _model, _tokenizer):
res = _model.chat(_tokenizer, image_file, ocr_type='ocr')
return res
@st.cache_data
def extract_text_easyocr(_image):
reader = easyocr.Reader(['hi'],gpu = False)
results = reader.readtext(np.array(_image))
# return results
return " ".join([result[1] for result in results])
@st.cache_resource
def model():
tokenizer = AutoTokenizer.from_pretrained('srimanth-d/GOT_CPU', trust_remote_code=True)
model = AutoModel.from_pretrained('srimanth-d/GOT_CPU', trust_remote_code=True, use_safetensors=True, pad_token_id=tokenizer.eos_token_id)
model = model.eval()
return model, tokenizer
@st.cache_resource
def highlight_keywords(text, keywords):
colors = generate_unique_colors(len(keywords))
highlighted_text = text
found_keywords = []
for keyword, color in zip(keywords, colors):
if keyword.lower() in text.lower():
highlighted_text = highlighted_text.replace(keyword, f'<mark style="background-color: {color};">{keyword}</mark>')
found_keywords.append(keyword)
return highlighted_text, found_keywords
def search():
st.session_state.search = True
@st.cache_data
def generate_unique_colors(n):
colors = []
for i in range(n):
color = "#{:06x}".format(random.randint(0, 0xFFFFFF))
while color in colors:
color = "#{:06x}".format(random.randint(0, 0xFFFFFF))
colors.append(color)
return colors
st.title("A Web-Based Text Extraction and Retrieval System")
language = st.selectbox("Select a language:", ["English", "Hindi"])
if language == "English":
st.subheader("You selected English!")
st.button("Let's get started", on_click=start)
if 'start' not in st.session_state:
st.session_state.start = False
if 'search' not in st.session_state:
st.session_state.search = False
if 'reset' not in st.session_state:
st.session_state.reset = False
if st.session_state.start:
uploaded_file = st.file_uploader("Upload an Image", type=["png", "jpg", "jpeg"])
if uploaded_file is not None:
st.subheader("Uploaded Image:")
st.image(uploaded_file, caption="Uploaded Image", use_column_width=True)
MODEL, TOKENIZER = model()
if not os.path.exists("images"):
os.makedirs("images")
with open(f"images/{uploaded_file.name}", "wb") as f:
f.write(uploaded_file.getbuffer())
extracted_text = get_text(f"images/{uploaded_file.name}", MODEL, TOKENIZER)
st.subheader("Extracted Text")
st.write(extracted_text)
keywords_input = st.text_input("Enter keywords to search within the extracted text (comma-separated):")
if keywords_input:
keywords = [keyword.strip() for keyword in keywords_input.split(',')]
highlighted_text, found_keywords = highlight_keywords(extracted_text, keywords)
st.button("Search", on_click=search)
if st.session_state.search:
st.subheader("Search Results:")
if found_keywords:
st.markdown(highlighted_text, unsafe_allow_html=True)
st.write(f"Found keywords: {', '.join(found_keywords)}")
else:
st.warning("No keywords were found in the extracted text.")
not_found_keywords = set(keywords) - set(found_keywords)
if not_found_keywords:
st.error(f"Keywords not found: {', '.join(not_found_keywords)}")
st.button("Reset", on_click=reset)
elif language == "Hindi":
st.subheader("You selected HINDI!")
st.button("Let's get started", on_click=start)
if 'start' not in st.session_state:
st.session_state.start = False
if 'search' not in st.session_state:
st.session_state.search = False
if 'reset' not in st.session_state:
st.session_state.reset = False
if st.session_state.start:
uploaded_file = st.file_uploader("Upload an Image", type=["png", "jpg", "jpeg"])
if uploaded_file is not None:
st.subheader("Uploaded Image:")
st.image(uploaded_file, caption="Uploaded Image", use_column_width=True)
image = Image.open(uploaded_file)
if not os.path.exists("images"):
os.makedirs("images")
with open(f"images/{uploaded_file.name}", "wb") as f:
f.write(uploaded_file.getbuffer())
extracted_text_hindi =extract_text_easyocr(image)
st.subheader("Extracted Text:")
st.write(extracted_text_hindi)
keywords_input = st.text_input("Enter keywords to search within the extracted text (comma-separated):")
if keywords_input:
keywords = [keyword.strip() for keyword in keywords_input.split(',')]
highlighted_text, found_keywords = highlight_keywords(extracted_text_hindi, keywords)
st.button("Search", on_click=search)
if st.session_state.search:
st.subheader("Search Results:")
if found_keywords:
st.markdown(highlighted_text, unsafe_allow_html=True)
st.write(f"Found keywords: {', '.join(found_keywords)}")
else:
st.warning("No keywords were found in the extracted text.")
not_found_keywords = set(keywords) - set(found_keywords)
if not_found_keywords:
st.error(f"Keywords not found: {', '.join(not_found_keywords)}")
st.button("Reset", on_click=reset)
|