Spaces:

OneFi
/

hf-similarity-check

Sleeping

File size: 8,010 Bytes

import streamlit as st
import similarity_check as sc
import cv2
from PIL import Image
import numpy as np
import demo
import streamlit as st
import request_json.sbt_request_generator as sbt
import check_hkid_validity as chv
import av
from streamlit_webrtc import webrtc_streamer, VideoTransformerBase, RTCConfiguration, WebRtcMode
import search_engine as se
import get_bank_statement as bs

# def init():
#     face_locations = []
#     # face_encodings = []
#     face_names = []
#     process_this_frame = True

#     score = []

#     faces = 0

# def video_frame_callback(frame: av.VideoFrame) -> av.VideoFrame:
#     image = frame.to_ndarray(format="bgr24")
    

def main():

    # st.title("SBT Web Application")
    # today's date = get_today_date

    # global data
    html_temp = """
        <body style="background-color:red;">
        <div style="background-color:teal ;padding:10px">
        <h2 style="color:white;text-align:center;">SBT Web Application</h2>
        </div>
        </body>
        """
    st.markdown(html_temp, unsafe_allow_html=True)
    
    if 'hkid_image_validity' not in st.session_state:
        st.session_state.hkid_image_validity = False

    if 'data' not in st.session_state:
        st.session_state['data'] = {}

    st.header("I. Similarity Check")
    image_file = st.file_uploader("Upload Image", type=['jpg', 'png', 'jpeg', 'pdf'], accept_multiple_files=True)
    if len(image_file) == 1:
        image1 = Image.open(image_file[0])
        st.text("HKID card")
        st.image(image1)
        image1.save('image/hkid.jpg', 'JPEG')
        if chv.check_hkid('image/hkid.jpg'):
            st.text("Valid HKID card.")
            st.session_state.hkid_image_validity = True
        else:
            st.text("Invalid HKID card. Please upload again!")
            st.session_state.hkid_image_validity = False
    elif len(image_file) == 2:
        image1 = Image.open(image_file[0])
        st.text("HKID card")
        st.image(image1)
        image2 = Image.open(image_file[1])
        # image2 = image_file[1]
        # image2.save('image/hkid.jpg', 'JPEG')
        # file_name = image_file[1].name
        st.text("Bank statement")
        st.image(image2)

    print(f"the id is: {st.session_state.hkid_image_validity}")
    # if image_file2 is not None:
    #     image2 = Image.open(image_file)
    #     st.text("Bank statement")
    #     st.image(image2)

    # path1 = 'IMG_4495.jpg'
    # path2 = 'hangseng_page-0001.jpg'    
    # image1 = save_image(image1)
    # image2 = save_image(image2)

    data = {}
    if st.button("Recognise"):
        with st.spinner('Wait for it...'):
            # global data
            data = sc.get_data(image1, image2)
            # se.get_data_link(data['chi_name_id'], data["name_on_id"], data["address"])
        if 'data' in st.session_state:
            data["nationality"] = 'N/A' # for hkid
            st.session_state['data'] = data
            st.session_state['verified'] = "True"
        st.success('Done!')
        score = int(st.session_state['data']['similarity_score'])
        st.text(f'score: {score}')
        if (score>85):
            st.text(f'matched')
        else:
            st.text(f'unmatched')
        
        data = st.session_state['data']
        st.header("Ia. HKID Data Extraction")
        st.text(f'English Name: {data["name_on_id"]}') # name is without space
        st.text(f'Chinese Name: {data["chi_name_id"]}') # name is without space
        st.text(f'HKID: {data["hkid"]} and validity: {data["validity"]}')
        st.text(f'Date of issue: {data["issue_date"]}')
        st.text(f'Date of birth: {data["dateofbirth"]}')
        st.text(f'nationality: {data["nationality"]}')
    
        st.header("Ib. Bank Statement Data Extraction")
        st.text(f'Name: {data["nameStatement"]}')
        st.text(f'Address: {data["address"]}')
        st.text(f'Bank: {data["bank"]}')
        st.text(f'Date: {data["statementDate"]}')
        st.text(f'Asset: {data["totalAsset"]} hkd')
        st.text(f'Liabilities: {data["totalLiability"]} hkd')

    if 'data' in st.session_state:
        tempout = st.session_state['data']
        print(f'data: {tempout}')
    

    # st.header("II. Facial Recognition")
    # run = st.checkbox('Run')

    # webrtc_streamer(key="example")
    # 1. Web Rtc
    # webrtc_streamer(key="jhv", video_frame_callback=video_frame_callback)


    # # init the camera
    # face_locations = []
    # face_encodings = []
    # face_names = []
    # process_this_frame = True

    # score = []

    # faces = 0

    # FRAME_WINDOW = st.image([])


    # server_ip = "127.0.0.1"
    # server_port = 6666

    # camera = cv2.VideoCapture(0)
    # s = socket.socket(socket.AF_INET, socket.SOCK_DGRAM)
    # s.setsockopt(socket.SOL_SOCKET, socket.SO_SNDBUF, 1000000)
    
    # if "face_rec" not in st.session_state:
    #     st.session_state.face_rec = []

    # while run:

        # rtc_configuration = RTCConfiguration({"iceServers": [{"urls": ["stun:stun.l.google.com:19302"]}]})

        # # Capture frame-by-frame
        # # Grab a single frame of video
        # ret, frame = camera.read()
    
        # result = frame
        # # Initialize the WebRTC streaming
        # webrtc_ctx = webrtc_streamer(
        #     key="face_rec",
        #     mode=WebRtcMode.SENDRECV,
        #     rtc_configuration=rtc_configuration,
        #     # video_transformer_factory=WebcamTransformer,
        #     video_frame_callback=video_frame_callback,
        #     media_stream_constraints={"video": True, "audio": False},
        #     async_processing=True,
        # )

        # print(f'xd: look here {type(webrtc_ctx)}')

        # st.session_state.face_rec = webrtc_ctx

        # if webrtc_ctx.video_transformer:
        #     st.header("Webcam Preview")
        #     frame = webrtc_ctx.video_transformer.frame
        #     result, process_this_frame, face_locations, faces, face_names, score = demo.process_frame(frame, process_this_frame, face_locations, faces, face_names, score)
        #     st.video(result)

        # frame = cv2.cvtColor(frame, cv2.COLOR_BGR2RGB)

        # FRAME_WINDOW.image(result)

        # if ret is not None:
        #     ret, buffer = cv2.imencode(".jpg", frame, [int(cv2.IMWRITE_JPEG_QUALITY),30])

        # x_as_bytes = pickle.dumps(buffer)

        # s.sendto((x_as_bytes),(server_ip, server_port))

        # camera.release()
        # if ret:
        #     # ret, buffer = cv2.imencode(".jpg", frame, [int(cv2.IMWRITE_JPEG_QUALITY)])
        #     # result, process_this_frame, face_locations, faces, face_names, score = demo.process_frame(frame, process_this_frame, face_locations, faces, face_names, score)
        #     # Display the resulting image
        #     FRAME_WINDOW.image(frame)
        # else:
        #     print("there is no frame detected")
        #     continue

        # print(score)
        # if len(score) > 20:
        #     avg_score =  sum(score) / len(score)
        #     st.write(avg_score)
        #     # st.write(f'{demo.convert_distance_to_percentage(avg_score, 0.45)}')
        #     # camera.release()
        #     run = not run
        #     st.session_state['data']['avg_score'] = str(avg_score)


    ## unrelated

    st.header("III. Search Engine and Bank Statement")
    user_input_id  = st.text_input("Enter the user ID here", " ")
    if st.button("Search data"):
        with st.spinner('Searching data...'):
            se.get_data_link(user_input_id)
        st.success('Done!')
    if st.button("Fetch bank statement"):
        with st.spinner('getting statements...'):
            bs.get_bs(user_input_id)
        st.success('Done!')
    if st.button("Confirm"):
        st.experimental_set_query_params(
            verified=True,
        )
        with st.spinner('Sending data...'):
            print(st.session_state['data'])
            sbt.split_data(st.session_state['data'])
        st.success('Done!')
    

if __name__ == '__main__':
    main()