h

app.py

@@ -0,0 +1,202 @@
+import json
+import requests
+import streamlit as st
+import pandas as pd
+import numpy as np
+from urllib.request import urlretrieve ,urlopen
+import os
+import torch
+from distutils.dir_util import copy_tree
+import tempfile
+import zipfile
+import subprocess
+import mmcv
+from mmcv import Config
+import sys
+from shapely.geometry import Polygon 
+
+
+import shutil
+import ssl
+
+import urllib.request
+from pathlib import Path
+
+import cv2
+from PIL import Image
+##########
+@st.cache(allow_output_mutation=True)
+def loading_resources():
+  dl_url = os.environ['dl_url']
+  if 'utils2.py' not in os.listdir('./'):
+    urlretrieve(dl_url, 'appdata.zip')
+    with zipfile.ZipFile("appdata.zip","r") as zip_ref:
+      zip_ref.extractall("./")
+loading_resources()
+
+import utilss
+from utils2 import *
+if 'model_inference' not in sys.modules:
+  from mmocr.apis import init_detector ,model_inference
+else:
+  model_inference = sys.modules['model_inference']
+  init_detector = sys.modules['init_detector']
+
+########## utils
+
+def image_resize(image, width = None, height = None, inter = cv2.INTER_AREA):
+    dim = None
+    (h, w) = image.shape[:2]
+    if width is None and height is None:
+        return image
+
+    if width is None:
+        r = height / float(h)
+        dim = (int(w * r), height)
+    else:
+        r = width / float(w)
+        dim = (width, int(h * r))
+    resized = cv2.resize(image, dim, interpolation = inter)
+    return resized
+def download_file(url, local_filename):
+    with requests.get(url, stream=True) as r:
+        with open(local_filename, 'wb') as f:
+            shutil.copyfileobj(r.raw, f)
+
+    return local_filename
+def load_and_preprocess_img(img_path, bbox=None):
+  img = Image.open(img_path).convert('RGB')
+  img = np.array(img).astype(np.uint8)
+  return img
+
+
+# @st.cache()
+def predictim(im , model):
+  if 'model_inference' not in sys.modules:
+    # if not model_inference:
+    from mmocr.apis import model_inference
+  result = model_inference(model,im )
+  # result[]
+  return result
+  # return result
+# minfer=None
+# minit=None
+@st.cache(allow_output_mutation=True)
+def loading_model():
+  # global minfer
+  # global minit
+  
+  from mmdet.apis import set_random_seed
+  set_random_seed(0, deterministic=False)
+  cfg = Config.fromfile('./cfgsn.py')
+  checkpoint = "./200s.pth"
+  model1 = init_detector(cfg, checkpoint, device="cpu")
+  if model1.cfg.data.test['type'] == 'ConcatDataset':
+    model1.cfg.data.test.pipeline = model1.cfg.data.test['datasets'][0].pipeline
+  return model1
+model = loading_model()
+
+
+
+def main():
+  st.sidebar.info('Images are deleted instantly')
+  form0 =st.sidebar.form("my_form0")
+
+  f = form0.file_uploader("Upload an Image", type=['png', 'jpg', 'jpeg', 'tiff', 'gif'])
+  btnpredicrupload = form0.form_submit_button("PREDICT")
+
+
+  st.sidebar.write(" ------ ")
+  form =st.sidebar.form("my_form")
+  photos = ['1.jpg','2.jpg','3.jpg','4.jpg','5.jpg','6.jpg',]
+  option = form.selectbox('Or choose a sample image', photos)
+  submitted = form.form_submit_button("PREDICT")
+
+  st.sidebar.write(" ------ ")
+  st.sidebar.info('Options')
+  fcolor = st.sidebar.selectbox('boundary color', ['red','green','blue'])
+  extra_postprocess = st.sidebar.checkbox('perform extra postprocessing',value=True)
+  # do_thresh_box = st.sidebar.checkbox('adaptive threshold')
+
+  if btnpredicrupload and f is not None:
+    tfile = tempfile.NamedTemporaryFile(delete=True)
+    tfile.write(f.read())
+    imgpath=tfile.name
+    run_app(imgpath ,fcolor,extra_post=extra_postprocess)
+  if submitted:
+    st.empty()
+    directory ='./imgs/'
+    pic = os.path.join(directory, option)
+    imgpath=directory+option
+    run_app(imgpath ,fcolor,extra_post=extra_postprocess)
+  # inp = st.text_input('t2','')
+  # if st.button('run'):
+  #   subprocess.run(inp.split(' '))
+
+
+# def run_app(imgpath,do_thresh,fcolor):
+def run_app(imgpath,fcolor,**kwargs):
+  d0 =dict(red=[255,0,0] ,green=[0,255,0] ,blue=[0,0,255]  )
+  fcolor = d0[fcolor]
+  # st.sidebar.write(imgpath)
+  # r = np.random.randint(1e3,1e7)
+  tfile = tempfile.NamedTemporaryFile(delete=True,suffix='.jpg')
+  dst = tfile.name
+  # tfile = open()
+  img = load_and_preprocess_img(imgpath)
+  # img = utilss.resize_with_pad(img,800,800)
+  # if do_thresh:
+  #   gray = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
+  #   threshim = cv2.adaptiveThreshold(gray, 255, cv2.ADAPTIVE_THRESH_GAUSSIAN_C,cv2.THRESH_BINARY, 51, 20)
+  #   img= cv2.cvtColor(threshim, cv2.COLOR_GRAY2RGB)
+  
+  cv2.imwrite(dst,cv2.cvtColor(img, cv2.COLOR_BGR2RGB))
+  # imgo = image_resize(img , width = 640)
+  st.write('input:')
+  st.image(img, caption = "Selected Input" , width =640)
+
+  # if st.button('predict2'):
+  
+  # try:
+  # except: print('errr--------------')
+  ''
+  ##################
+  preds,img_metas,downsample_ratio =  model_inference(model,dst )
+  ff= fixer(img,preds,img_metas,downsample_ratio)
+  polys= ff.polysfin3
+  if kwargs['extra_post']:
+    polys =find_polys4(ff) 
+  
+
+  # with open('./0.json','r') as f:
+  #   res = json.load(f)
+  ################
+  mm = utilss.draw_conts(img,[poly2cont(p) for p in polys],fcolor)
+  ## mm=image_resize(mm, width = 640)
+  #######
+  # mm=preds[1]*preds[0]
+  # mm=cv2.cvtColor(mm, cv2.COLOR_GRAY2RGB)
+  st.image(mm,caption='output image' , width = 640)
+  # del polys
+  del ff
+  del mm 
+  del preds
+  del polys
+  del img
+  del img_metas
+  # os.remove(dst)
+  # x = predictim(img,model)
+  # st.write(x)
+
+main()
+
+
+
+
+
+
+
+
+
+
+  

requirements.txt

@@ -0,0 +1,26 @@
+--find-links https://download.pytorch.org/whl/torch_stable.html
+torch==1.12.0+cu113
+torchvision==0.13.0+cu113
+
+#--find-links https://download.openmmlab.com/mmcv/dist/cpu/torch1.12.0/index.html
+--find-links https://download.openmmlab.com/mmcv/dist/cu113/torch1.12.0/index.html
+mmcv-full==1.6.0
+mmdet
+
+numpy
+pyclipper
+
+opencv-python-headless==4.6.0.66
+pycocotools
+rapidfuzz
+scikit-image
+six
+terminaltables
+matplotlib
+numba>=0.45.1
+lanms-neo==1.0.2
+imgaug
+lmdb
+matplotlib
+easydict
+sklearn