def ocr_api():
image = request.files['img']
image.save(os.path.join(app.config['UPLOAD_FOLDER'], "down.jpg"))
img = cv2.imread('down.jpg')
text = run_ocr(img)
# pos = text.find('EXP')
# target = text[pos: pos+10]
# pos = target.find("/")
# date = target[pos-2: pos] + target[pos+1: pos+3]
print(text)
date = ""
try:
date = re.search(r'[0-9]+/[0-9]+', text).group()
text2 = text[text.find(date) + 3:]
try:
date = re.search(r'[0-9]+/[0-9]+', text2).group()
except:
pass
except:
return json.dumps("sorry can't read date please retake image")
return json.dumps(date)
def creator(files, printer):
printer.sprint('Initializing OCR...', 0, 0)
if threading.currentThread().getName() != 'MainThread':
pythoncom.CoInitialize()
#socket.emit()
docs = [
'Bank Statement', 'Bill Statement', 'Loan Repayment Schedule',
'CAT Template'
]
template_loc = "static/template"
temp_loc = "static/temp"
output_loc = "static/output"
upload_loc = "static/uploads"
wd = os.getcwd().replace("\\", "/") + "/"
template = template_loc + '/template.xlsm'
output = output_loc + '/output.xlsm'
shutil.copy(template, output)
def adder(file_path):
xl = win32.gencache.EnsureDispatch('Excel.Application')
xl.Visible = False # change to False after development
xl.DisplayAlerts = False
xl.AskToUpdateLinks = False
output_xl = xl.Workbooks.Open(wd + output_loc + '/output.xlsm')
file_xl = xl.Workbooks.Open(file_path)
file_xl.Sheets(1).Activate()
for j in xl.ActiveWorkbook.Sheets:
xl.DisplayAlerts = False
j.Copy(After=output_xl.Sheets(output_xl.Sheets.Count))
xl.DisplayAlerts = False
file_xl.Close(SaveChanges=False)
output_xl.Sheets(1).Activate()
for k in xl.ActiveWorkbook.Sheets:
if k.Name not in ["Introduction", "Summary", "User Metric"]:
k.Activate()
k.Visible = False
output_xl.Save()
# def adder(file_path):
# xl = win32.gencache.EnsureDispatch('Excel.Application')
# xl.Visible = False # change to False after development
# xl.DisplayAlerts = False
# output_xl = xl.Workbooks.Open(wd + output_loc + '/output.xlsm')
# file_xl = xl.Workbooks.Open(file_path)
# file_xl.Sheets(1).Activate()
#
# for j in xl.ActiveWorkbook.Sheets:
# j.Activate()
# j.Visible = 1
# j.Copy(After=output_xl.Sheets(output_xl.Sheets.Count))
# file_xl.Close()
#
# output_xl.Sheets(1).Activate()
# for k in xl.ActiveWorkbook.Sheets:
# if k.Name != "Introduction":
# k.Activate()
# k.Visible = False
#
# output_xl.Save()
# xl.Quit()
# def adder(file_path):
# xl = win32.gencache.EnsureDispatch('Excel.Application')
# xl.Visible = False # change to False after development
# xl.DisplayAlerts = False
# xl.AskToUpdateLinks = False
# output_xl = xl.Workbooks.Open(wd + output_loc + '/output.xlsm', UpdateLinks=False)
# file_xl = xl.Workbooks.Open(file_path, UpdateLinks=False)
# file_xl.Sheets(1).Activate()
#
# for j in xl.ActiveWorkbook.Sheets:
# j.Activate()
# j.Visible = 1
# j.Copy(After=output_xl.Sheets(output_xl.Sheets.Count))
# file_xl.Close()
#
# output_xl.Sheets(1).Activate()
# for k in xl.ActiveWorkbook.Sheets:
# if k.Name != "Introduction":
# k.Activate()
# k.Visible = False
#
# output_xl.Save()
# xl.Quit()
def identifier(file):
if 'bank' in file.lower():
return 'BaS', docs[0]
elif 'bill' in file.lower():
return 'BiS', docs[1]
elif 'facility' in file.lower():
return 'LR', docs[2]
elif 'cat' in file.lower():
return 'CAT', docs[3]
else:
return "Not a compatible document type."
printer.sprint('Identifying documents...', 4, 0)
#frames = {'BiS': None, 'LR': None, 'BaS': None, 'CAT': None}
frames = {}
for file in files:
doc_type, doc_name = identifier(file)
if doc_type == 'CAT':
printer.sprint('Reading CAT Template...', 7, 0)
adder(file)
continue
data, coor, conf = run_ocr(file, doc_type, printer).multi_page_reader()
frames[doc_type] = {
'filename': file,
'docname': doc_name,
'data': data,
'coor': coor,
'conf': conf
}
printer.sprint('Storing Data...', 11, 0)
with pd.ExcelWriter(temp_loc + '/temp.xlsx') as writer:
for doc_type in frames:
if doc_type == 'CAT':
continue
doc_name = frames[doc_type]['docname']
coor = '_coor' + doc_type
conf = '_conf' + doc_type
frames[doc_type]['data'].to_excel(writer, sheet_name=doc_name)
frames[doc_type]['coor'].to_excel(writer, sheet_name=coor)
frames[doc_type]['conf'].to_excel(writer, sheet_name=conf)
frames[doc_type]['sheets'] = [doc_name, coor, conf]
adder(wd + temp_loc + '/temp.xlsx')
xl = win32.gencache.EnsureDispatch('Excel.Application')
xl.Visible = False # change to False after development
xl.DisplayAlerts = False
xl.AskToUpdateLinks = False
output_xl = xl.Workbooks.Open(wd + output_loc + '/output.xlsm',
UpdateLinks=False)
#output_xl.Sheets("Sheet1").Delete()
output_xl.Sheets(1).Activate()
output_xl.VBProject.VBComponents('Sheet1').CodeModule.AddFromString(
'Option Explicit')
# try:
# xl.Application.Run("output.xlsm!Module3.InsertPivotTable")
# except:
# pass
_sheets = {}
for i, j in enumerate(xl.ActiveWorkbook.Sheets):
j.Activate()
_sheets[j.Name] = [i + 1, j.CodeName]
printer.sprint('Creating Excel file...', 10, 0)
for i in frames:
if i == "CAT":
continue
data_sheet, coor_sheet, conf_sheet = frames[i]['sheets']
print(data_sheet, coor_sheet, conf_sheet)
# output_xl.Sheets(data_sheet).Visible = False
# output_xl.Sheets(coor_sheet).Visible = False
# output_xl.Sheets(conf_sheet).Visible = False
# output_xl.Sheets(data_sheet).Activate()
data_ws = output_xl.Sheets(data_sheet)
data_ws.Visible = True
data_ws.Activate()
data = data_ws.Range("B:Z")
data.Select()
data.FormatConditions.Add(
2,
Formula1='=if({0}!B1 <> "", {0}!B1<85, False)'.format(conf_sheet))
#data.FormatConditions(data.FormatConditions.Count).SetFirstPriority()
data.FormatConditions(
data.FormatConditions.Count).Interior.ColorIndex = 22
output_xl.Save()
pagecol = frames[i]['coor'].columns.get_loc("Page Number")
col = chr(pagecol + 98).capitalize()
maxpg = frames[i]['coor']["Page Number"].max()
pagels = {}
pageurl = {}
for j in range(int(maxpg)):
name = wd + 'sample{}/{}.jpg'.format(i, j + 1)
img = cv2.imread(name, cv2.IMREAD_UNCHANGED)
pagels[j + 1] = img.shape
pageurl[j + 1] = name
w_factor = 600 / pagels[1][1]
selector_code = '''Option Explicit
Private Sub Worksheet_SelectionChange(ByVal Target As Range)
On Error Resume Next
'Dim words As Collection
'Set words = CollectionCreator()
Dim Range As Range
Dim rCell As Range
If Selection.Count > 0 And Selection.Count < 3 Then
On Error Resume Next
Call DeleteAllShapes
For Each rCell In Target.Cells
Dim rng As String
Dim val As String
Dim i As Integer
Dim msgString As String
Dim coordinates() As String
rng = rCell.Address(RowAbsolute:=False, ColumnAbsolute:=False)
val = Worksheets("{0}").Range(rng).Value
If Not val = "" Then
coordinates = Split(val, " ")
Dim vArr() As String
Dim row_num As String
Dim pn_address As String
Dim pn As String
Dim name As String
Dim top As Double
Dim shp as Object
vArr = Split(rCell.Address(True, False), "$")
row_num = vArr(1)
pn_address = "{1}" & row_num
pn = Worksheets("{0}").Range(pn_address).Value
name = "page" & pn
ActiveSheet.Shapes(name).Visible = True
ActiveSheet.Shapes(name).Width = 600
top = ActiveSheet.Shapes(name).top
For Each shp In ActiveSheet.Shapes
If shp.name <> name Then shp.Visible = False
Next shp
'If Exists(words, range) Then
Dim x As Double: x = coordinates(0)*{2}
Dim y As Double: y = coordinates(1)*{2} + top
Dim w As Double: w = coordinates(2)*{2}
Dim h As Double: h = coordinates(3)*{2}
'Dim name As String: name = coordinates(4)
Call BoxCreator(x, y, w, h)
'Else
'Call DeleteAllShapes
End If
Next rCell
End If
End Sub'''.format(coor_sheet, col, w_factor)
# print(_sheets)
# for i in output_xl.VBProject.VBComponents:
# print(i)
# print(i.Name)
# print(i.Properties)
output_xl.VBProject.VBComponents(str(
_sheets[data_sheet][1])).CodeModule.AddFromString(selector_code)
data_ws.Visible = True
data_ws.Activate()
xl.Range("B1").Select() # add the () at the end here
xl.ActiveWindow.FreezePanes = True
data_ws.Columns("A").ColumnWidth = 106
data_ws.Columns("A").Font.ColorIndex = 2
#xl.Selection.ClearContents()
# print(pageurl)
for l in pageurl:
# data_ws.Activate()
entry_index = frames[i]['coor'].index[
frames[i]['coor']["Page Number"] == l].tolist()[0] + 1
top = xl.Range("A" + str(entry_index)).Top
link = pageurl[l].replace("/", "\\")
pic = data_ws.Shapes.AddPicture(link,
LinkToFile=False,
SaveWithDocument=True,
Left=0,
Top=0,
Width=-1,
Height=-1)
pic.Name = 'page' + str(l)
pic.Left = 0
pic.Top = top
pic.Width = 600
if l == 1:
pic.Visible = True
else:
pic.Visible = False
output_xl.Sheets(1).Activate()
xl.DisplayAlerts = False
output_xl.Save()
xl.Quit()
from model import ProductClassifier
from ocr import run_ocr
from config import *
import argparse
import time
PC = ProductClassifier()
PC.load(MODEL_PATH)
if __name__ == '__main__':
parser = argparse.ArgumentParser(description='Process images.')
parser.add_argument('-i',
'--image',
type=str,
help='images of products to identify')
args = parser.parse_args()
output = list()
start = time.time()
ocr_result = run_ocr(args.image)
print('Ocr time: {0} sec.'.format(round(time.time() - start, 2)))
product_name, class_probabilities = PC.predict(ocr_result)
print('Full time: {0} sec.'.format(round(time.time() - start, 2)))
print(product_name, class_probabilities.max())
from ocr import run_ocr
from excel import excel_coor
import pandas as pd
#data, coor, conf = run_ocr('Obligor_5_Facility_Letter_Dah_Sing.pdf', 'LR').multi_page_reader()
data, coor, conf = run_ocr('Obligor_3_DBS_Bills_Statements_Apr.pdf', 'BiS').multi_page_reader()
data.to_excel("static/temp/output_info.xlsx")
coor.to_excel("static/temp/cor_info.xlsx")
excel_coor()
capture = cv2.VideoCapture(
f'http://{creds.username}:{creds.password}@{creds.ip}/video')
# .jpg image file name
image_file = r'outfiles\image.jpg'
while True:
ret, frame = capture.read()
img = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY) # grayscaling the image
_, th = cv2.threshold(img, 0, 255, cv2.THRESH_BINARY +
cv2.THRESH_OTSU) # applying otsu binarization
cv2.imshow('frame', th)
k = cv2.waitKey(1)
if k == 32: # RETURN Key is pressed
r = requests.get(url_af)
with open(image_file, 'wb') as f:
f.write(r.content)
ocr.filter_image(image_file, False)
text = ocr.run_ocr(image_file)
print(text)
with open(r'outfiles\infile.txt', 'w') as f:
f.write(text)
if k == 27: # ESC Key is pressed
break
help="path to input images to be OCR'd",
default="images")
ap.add_argument("-p",
"--preprocess",
type=str,
help="type of preprocessing to be done",
default="thresh")
ap.add_argument("--framepath",
help="path to OCR output of frames",
default="output")
args = vars(ap.parse_args())
start = time.perf_counter()
frame_extractor.extract_frames(args["video"])
frame_extract_time = time.perf_counter() - start
start = time.perf_counter()
ocr.run_ocr(args["image_path"], args["preprocess"])
ocr_time = time.perf_counter() - start
start = time.perf_counter()
match_code.match(args["codefile"], args["framepath"])
code_match_time = time.perf_counter() - start
print("\n")
print("Frame extraction completed in : ", frame_extract_time)
print("\n")
print("OCR completed in : ", ocr_time)
print("\n")
print("Code matching completed in : ", code_match_time)