def main():
pdfToImage('assets/pdfs/form.pdf')
image = "assets/images/page.jpg"
crop(image, (95, 634, 1557, 708), 'assets/images/date.jpg')
crop(image, (99, 776, 1557, 1008), 'assets/images/comments.jpg')
print(ocr('assets/images/date.jpg'))
print(ocr('assets/images/comments.jpg'))
def obtener_traducciones(mapa):
img = detectar_area_variantes(mapa)['tra']['im']
img_mejorada = cvr.preprocesar_texto_otsu(img,127,255,(3,7),3)
text_eng = ocr(img_mejorada,'eng')
text_fra = ocr(img_mejorada,'fra')
traducciones = {}
traducciones['frances']=text_fra[0]
traducciones['ingles']=text_eng[1]
return(traducciones)
def main():
image = cv2.imread("pic2.jpg")
edged = edgeDetection(image)
screenCnt = findContour(edged)
scannedImage = scan(screenCnt,image)
### OCR
processedImg = preprocess(scannedImage)
ocr(processedImg)
cv2.imshow("Scanned", imutils.resize(scannedImage, height = 650))
cv2.waitKey(0)
cv2.destroyAllWindows()
def menu():
if request.method == 'POST':
uploaded_file = request.files['img']
filename = uploaded_file.filename
if filename != '':
file_ext = os.path.splitext(filename)[1]
if (file_ext not in app.config['UPLOAD_EXTENSIONS']):
abort(400)
uploaded_file.save(
os.path.join(app.config['UPLOAD_PATH'], filename))
path = "static/uploads/" + str(filename)
try:
content = ocr(path)
content = json.dumps(content)
newFile = FileContents(name=filename, path=path, content=content)
db.session.add(newFile)
db.session.commit()
data = FileContents.query.order_by(FileContents.date_created).all()
return render_template('menu.html', data=data)
except:
return '<h1>There was a problem processing the receipt, make sure you uploaded the correct file</h1>'
else:
data = FileContents.query.order_by(FileContents.date_created).all()
return render_template('menu.html', data=data)
def test_numero_cinco(self):
numero_cinco = """
__
|__
__|
"""
self.assertEquals(5, ocr(numero_cinco))
def test_numero_nove(self):
numero_nove = """
__
|__|
__|
"""
self.assertEquals(9, ocr(numero_nove))
def test_numero_zero(self):
numero_zero = """
__
| |
|__|
"""
self.assertEquals(0, ocr(numero_zero))
def gameResult(self):
print('=' * 20 + 'game reslut:')
if self.imgbg is None:
self.updateImgbg()
img_royale_red = cv2.imread(self.game_params.img['royale_red'])
is_red, red_royale = self.inHere(img_royale_red, self.imgbg, 0.97)
if is_red:
print('red crown:', len(red_royale), red_royale)
img_royale_blue = cv2.imread(self.game_params.img['royale_blue'])
is_blue, blue_royale = self.inHere(img_royale_blue, self.imgbg, 0.97)
if is_blue:
print('blue crown:', len(blue_royale), blue_royale)
sign = ''
img_win_blue = cv2.imread(self.game_params.img['winner_blue'])
win_blue, _ = self.inHere(img_win_blue, self.imgbg, 0.95)
if win_blue:
print('blue win!', end=' ')
sign = '+'
img_win_red = cv2.imread(self.game_params.img['winner_red'])
win_red, _ = self.inHere(img_win_red, self.imgbg, 0.95)
if win_red:
print('red win!', end=' ')
sign = '-'
cups_img = self.grabScreen(self.game_params.game_area_left + 310,
self.game_params.game_area_top + 410,
self.game_params.game_area_left + 310 + 24,
self.game_params.game_area_top + 410 + 15)
# cups_img.show()
cups_str = ocr(cups_img, type='digits')
print(cups_str)
print('=' * 20 + 'game over')
def ocr_image(cache, url, codelang):
# This is checked in bot_listening but must be redone here, so if
# the ocr for the same page is asked multiple time, we will do the ocr
# only once.
text = get_from_cache(cache, url, codelang)
if text:
return ret_val(0, text)
url = url.encode('utf-8')
cache_key = image_key(url)
lang = ocr.tesseract_languages.get(codelang, 'eng')
basename = os.path.expanduser('~/tmp') + '/tesseract/image_%s' % cache_key
image_filename = basename + ".jpg"
utils.copy_file_from_url(url, image_filename)
if not os.path.exists(image_filename):
return ret_val(1, "could not download url: %s" % url)
text = ocr.ocr(image_filename, basename, lang)
if text == None:
return ret_val(2, "ocr failed")
os.remove(image_filename)
if os.path.exists(basename + ".txt"):
os.remove(basename + ".txt")
cache.set(cache_key, text)
return ret_val(0, text)
def test_numero_seis(self):
numero_seis = """
__
|__
|__|
"""
self.assertEquals(6, ocr(numero_seis))
def test2():
body = request.json
#pprint.pprint(body)
image_url = body['action']['detailParams']['namecard']['origin']
#image_url =image_url.startswith('http://dn-m.talk.kakao.com/talkm')
#print(image_url)
with urllib.request.urlopen(image_url) as input:
with open('./image.jpeg', 'wb') as output:
output.write(input.read())
result_file = ocr('./image.jpeg')
#print(result_file)
pytesseract.pytesseract.tesseract_cmd = r'C:\Program Files\Tesseract-OCR\tesseract'
result_string = pytesseract.image_to_string(result_file, lang='kor+eng')
result_string = result_string.strip().replace(' ', '')
print(result_string)
return {
"version": "2.0",
"template": {
"outputs": [{
"simpleText": {
"text": result_string
}
}]
}
}
def build(self):
if self.cells is None:
self.find_cells()
ocr_data = ocr(self.image)
table_data = []
for row in self.cells:
table_data.append([])
for _ in row:
table_data[-1].append([])
for i, data in ocr_data.iterrows():
centroid = get_centroid(data['left'], data['left'] + data['width'],
data['top'], data['top'] + data['height'])
cell = self.find_cell_for_point(centroid)
if cell is not None:
table_data[cell[0]][cell[1]].append(data['text'])
for i, row in enumerate(table_data):
for j, cell in enumerate(row):
if len(cell) == 0:
table_data[i][j] = None
else:
table_data[i][j] = ' '.join(cell)
self.data = table_data
def do_POST(self):
formData = cgi.FieldStorage(self.rfile,
headers=self.headers,
environ={'REQUEST_METHOD': 'POST'})
content = formData.getvalue('content')
img = Image.open(BytesIO(content))
base_name = str(int(time.time()))
image_name = base_name + '.bmp'
img.save(image_name)
ocr_image = ocr.ocr(image_name, base_name)
ocr_name = base_name + '.png'
Image.fromarray(ocr_image).save(ocr_name)
blob = storage.upload(CONFIG.project_id,
CONFIG.buket,
ocr_name,
service_account_file='service_account.json')
signed_url = storage.signed_url_for_png(blob)
shutil.os.remove(image_name)
shutil.os.remove(base_name + '.tsv')
shutil.os.remove(ocr_name)
self.send_response(HTTPStatus.OK)
self.send_header('Content-type', 'application/json')
self.end_headers()
responseBody = json.dumps({'url': signed_url})
self.wfile.write(responseBody.encode('utf-8'))
def test_numero_tres(self):
numero_tres="""
__
__|
__|
"""
self.assertEquals(3, ocr(numero_tres))
def ocr_image(cache, url, codelang):
# This is checked in bot_listening but must be redone here, so if
# the ocr for the same page is asked multiple time, we will do the ocr
# only once.
text = get_from_cache(cache, url, codelang)
if text:
return ret_val(0, text)
url = url.encode('utf-8')
cache_key = image_key(url)
lang = ocr.tesseract_languages.get(codelang, 'eng')
basename = os.path.expanduser('~/tmp') + '/tesseract/image_%s' % cache_key
image_filename = basename + ".jpg"
utils.copy_file_from_url(url, image_filename)
if not os.path.exists(image_filename):
return ret_val(1, "could not download url: %s" % url)
text = ocr.ocr(image_filename, basename, lang)
if text == None:
return ret_val(2, "ocr failed")
os.remove(image_filename)
if os.path.exists(basename + ".txt"):
os.remove(basename + ".txt")
cache.set(cache_key, text)
return ret_val(0, text)
def solve(path):
eq = ocr.ocr(path)
if (detect.isArith(eq)):
if (detect.isArithEq(eq)):
return (f"\t\t{arith.arithEq(eq)}\n")
else:
if (arith.arithIneq(eq)):
return ("\t\tTrue\n")
else:
return ("\t\tFalse\n")
else:
nterm, var = alg.checknovar(eq)
if (nterm == 1):
pow = alg.checkdegree(eq, var[0])
if (pow == 1):
return (linear(eq, var[0]))
elif (pow == 2):
return (quadratic(eq, var[0]))
elif (pow == 3):
return ("Cubic equation\n")
else:
return ("Can solve up to 3rd degree polynomial only!!\n")
elif (nterm == 2):
return (simult(eq, var, nterm))
elif (nterm == 3):
print("Simultaneous equation three variables\n")
else:
print("Can solve equations upto 3 vaiables only!!!\n")
def test_numero_sete(self):
numero_sete = """
__
|
|
"""
self.assertEquals(7, ocr(numero_sete))
def test_numero_oito(self):
numero_oito = """
__
|__|
|__|
"""
self.assertEquals(8, ocr(numero_oito))
def do_one_page(opt, page_nr, filename):
tiff_name = extract_image(opt, page_nr, filename)
if not tiff_name:
return
out_filename = opt.out_dir + "page_%04d" % page_nr
if opt.config == 'hocr':
out_filename += '.html'
else:
out_filename += '.txt'
ocr.ocr(tiff_name, opt.out_dir + 'page_%04d' % page_nr, opt.lang, opt.config)
if opt.compress:
utils.compress_file(out_filename, out_filename, opt.compress)
os.remove(out_filename)
os.remove(tiff_name)
def test_numero_dez(self):
numero_dez = """
__
| | |
| |__|
"""
print repr(numero_dez)
self.assertEquals(10, ocr(numero_dez))
def test_numero_um(self):
numero_um = """
|
|
"""
self.assertEquals(1, ocr(numero_um))
def test_numero_quatro(self):
numero_quatro="""
|__|
|
"""
self.assertEquals(4, ocr(numero_quatro))
def obtener_titulo(mapa):
img = detectar_area_variantes(mapa)['tit']['im']
img_mejorada = cvr.preprocesar_texto_adapta(img, 255,71,30)
text = ocr(img_mejorada,'spa')
titulo = {}
titulo['numero']=text[0].replace(' ','')
titulo['ententrada']=[' '.join(text[1:])][0]
return(titulo)
def test_numero_dois(self):
numero_dois = """
__
__|
|__
"""
self.assertEquals(2, ocr(numero_dois))
def do_one_page(opt, page_nr, filename):
tiff_name = extract_image(opt, page_nr, filename)
if not tiff_name:
return
out_filename = opt.out_dir + "page_%04d" % page_nr
if opt.config == 'hocr':
out_filename += '.hocr'
else:
out_filename += '.txt'
ocr.ocr(tiff_name, opt.out_dir + 'page_%04d' % page_nr, opt.lang, opt.config)
if opt.compress:
utils.compress_file(out_filename, out_filename, opt.compress)
os.remove(out_filename)
os.remove(tiff_name)
def getyzm():
print("getyzm")
fp=io.BytesIO(a.get("https://isisn.nsfc.gov.cn/egrantindex/validatecode.jpg",result=False,o=True).content)
image=Image.open(fp)
yzm = ocr(image)
try:
return 'success' in a.post("https://isisn.nsfc.gov.cn/egrantindex/funcindex/validate-checkcode","checkCode="+yzm).text, yzm
except:
return False,''
def getPokemonInfo():
toolkit.takeScreenshot()
img = ds.getScreenshot()
pokemonName = ocr.ocr('pokemon', img).lower()
x = ds.getShinyListX(pokemonName)
y = ds.getShinyListY(pokemonName)
shinyColor = ds.getShinyListColor(pokemonName)
return x, y, shinyColor
def test_image(self):
for name, expected in cases.items():
image = os.path.join("screenshots", name)
out, err = ocr(image)
if "err" in expected:
assert err == expected["err"]
else:
assert err == b""
assert expected["ocr_out"] in out
# ocrout = os.path.join("screenshots", name) + ".txt"
# with open(ocrout, encoding='utf-8') as fh:
# expectedoutput = fh.read()
# assert out == expectedoutput
image = os.path.join("screenshots", "view.jpg")
out, err = ocr(image)
assert err == b"Warning. Invalid resolution 0 dpi. Using 70 instead.\n"
assert b' \n\n' == out
def workerThreadFunc():
self.ongoing = True
print('==========')
print('光学字符识别')
print('启动时间:' + time.asctime(time.localtime(time.time())))
print('共找到文件数:' + str(int(len(ocr.file_list)) - 1))
Counts = (int(len(ocr.file_list)) - 1)
time.sleep(0.5)
self.ui.progressBar.setRange(0, Counts)
self.ui.progressBar.setValue(0)
print('==========')
ocr.textlist.append('==========')
ocr.textlist.append('光学字符识别')
ocr.textlist.append('启动时间:' +
time.asctime(time.localtime(time.time())))
ocr.textlist.append('共找到文件数:' + str(int(len(ocr.file_list) - 1)))
ocr.textlist.append('==========')
count = 1
for ocr.path in ocr.file_list:
print(ocr.file_list)
Counts = int(Counts)
if ocr.path == "C:\\MonianHello\\list.txt":
continue
try:
# ocr.transimg(ocr.path)
time.sleep(0.5)
self.ui.progressBar.setValue(count)
print('{0} / {1}'.format(str(count), str(Counts)))
ocr.ocr(ocr.path)
count += 1
except:
print('出现内部错误')
for i in ocr.textlist:
fl.write(i)
fl.write('\n')
fl.write('结束时间:' + time.asctime(time.localtime(time.time())))
fl.close()
print('写入成功，已将文件写入' + str(ocr.outputtext))
print('结束时间:' + time.asctime(time.localtime(time.time())))
print('识别完成，现在将结果写入文件...')
time.sleep(5)
self.ongoing = False
def upload_file():
if request.method == 'POST':
f = request.files['file']
file_name = f.filename
f.save(file_name)
text = str(ocr(str(file_name)))
text = re.findall(r'\w+', text)
text = ' '.join(text)
df = cosine_similarity_T(10, text).head(6)
return render_template('index.html', frame=df)
def __init__(self, image, caption=None):
self.image = image
self.caption = caption
# Photo OCR if we don't have an input caption
if not self.caption:
self.caption = ocr.ocr(self.image)
# Make image dimensions uniform
if len(self.image) != dimX or len(self.image[0]) != dimY:
self.image = alignDimensions.alignDimensions(
self.image, dimX, dimY)
def __init__(self):
"""
初始化
:return: null
"""
# 超类初始化
super().__init__()
# UI初始化
self.ui = Ui_mainWidget()
self.ui.setupUi(self)
self.grabKeyboard()
self.setMouseTracking(True)
self.setWindowFlags(Qt.FramelessWindowHint)
self.setWindowIcon(QIcon('OCR.ico'))
# 初始化相机
self.camera = QCamera()
self.imageCapture = QCameraImageCapture(self.camera)
self.viewsetting = QCameraViewfinderSettings()
self.initimplement()
# 初始化标题栏
self.initTitleBar()
# 初始化系统托盘
self.tray = QSystemTrayIcon()
self.tray.setIcon(QIcon('OCR.ico'))
self.initTray()
# OCR识别部分
self.OCR = ocr()
self.OCR.setappid('1257206643')
self.OCR.setsecretid('AKIDFTddWEg9Ncsz0sE7oOpBNOExdDdeCUJ3')
self.OCR.setsecretkey('FQitsgUND8yfrZK0RrBMOJB5tWhCm5Ol')
# 初始化登录部分
self.logWidget = QWidget()
self.logui = Ui_Form()
self.logui.setupUi(self.logWidget)
self.logWidget.setWindowFlags(Qt.FramelessWindowHint)
self.logWidget.setWindowModality(Qt.ApplicationModal)
self.logui.close_btn.clicked.connect(self.logWidget.close)
# 初始化变量
self.mousePressd = False
self.mousePoint = None
self.result = {}
self.isFirst = False
self.ocrType = ocrType.ocr_general # 默认为印刷体识别
# 初始化字定义信号连接
self.processFinished.connect(self.updateOCRInfo)
self.ui.btn_login.clicked.connect(self.logWidget.show)
self.ui.comboBox_choose.currentIndexChanged.connect(self.changeocrType)
def getCups(self):
cups_img = self.grabScreen(self.game_params.game_area_left + 230,
self.game_params.game_area_top + 55,
self.game_params.game_area_left + 230 + 50,
self.game_params.game_area_top + 57 + 16)
# cups_img.show()
cups_str = ocr(cups_img, type='digits')
# print(cups_str)
if cups_str == '':
return 0
return cups_str
def capture(self):
# 右下から左上に選択したときの処理
if self.startPos.x() > self.endPos.x():
self.startPos, self.endPos = self.endPos, self.startPos
# QPixmap で選択座標をコピー
pmap = self.snipScreen.copy(QRect(self.startPos, self.endPos))
# QPixmap -> Qimage -> binary
qimg = pmap.toImage()
image_format = 'PNG'
bits = QByteArray()
buffer = QBuffer(bits)
qimg.save(buffer, image_format)
# OCR
ocr(buffer.data(), self.args.debugmode, self.gcp_eng,
self.is_cut_newline)
self.close()
def ocr_endpoint():
if request.method == 'POST':
file = request.files['file']
if file.filename == '':
return 'no file selected'
filename = secure_filename(file.filename)
file_path = os.path.join(app.config['UPLOAD_FOLDER'], filename)
file.save(file_path)
text = ocr(file_path)
os.remove(file_path)
return text
else:
return "Welcome to ocr"
def make_predict():
#return jsonify(request.get_json())
#data = request.get_json(force=True)
#return '123'
#predict_request = [str(x) for x in request.form.values()]
#predict_request = np.array(predict_request)
#delay_predicted = model.predict(predict_request)
#output = predict(data['string'])
#predict_request = [str(x) for x in request.form.values()]
#predict_request = predict_request[0]
#output = predict(translate(predict_request[0],'en'))
output = predict(translate(str(request.form.get('headline')), 'en'))
#return '123'
#return str(output)
#return jsonify(str(output))
#url = predict_request[1]
url = str(request.form.get('url'))
if (request.form.get('headline') != ''):
if (output):
result = "decrease"
else:
result = "increase"
## return render_template('index.html', prediction_text='Stock Price will {}'.format(result))
elif (request.form.get("url") != ''):
toi_article = Article(url, language="en") # en for English
toi_article.download()
#toi_article.parse()
output_2 = predict(toi_article.title)
if (output_2):
result = "decrease"
else:
result = "increase"
else:
file = request.files['file']
#filename = secure_filename(file.filename)
text = ocr(file.filename)
output_3 = predict(text)
if (output_3):
result = "decrease"
else:
result = "increase"
return render_template('index.html',
prediction_text=translate(
'Stock Price will {}'.format(result), lang),
Prediction=translate('Stock Market Prediction',
lang),
Predict=translate('Predict', lang),
Plot=translate('plot', lang))
def display():
global UPLOAD_FOLDER, FILE_NAME, ocrtext
ocrtext = ocr(FILE_NAME)
f = open('test.txt', 'w+')
f.write(ocrtext)
f.close()
info = label(ocrtext)
with open('test.json', 'w+') as g:
json.dump(info, g)
return render_template('display.html', ocrtext=info)
def run(self):
if self.device.isOpened():
try:
while True:
ret, frame = self.device.read()
height, width, bytesPerComponent = frame.shape
bytesPerLine = bytesPerComponent * width
# 變換彩色空間順序
cv2.cvtColor(frame, cv2.COLOR_BGR2RGB, frame)
# 轉為QImage物件
image = QImage(frame.data, width, height, bytesPerLine,
QImage.Format_RGB888)
if self.paizhao == 1:
image.save('./origin/takephoto.png')
self.paizhao = 0
pixmap = QPixmap.fromImage(image)
pixmap = pixmap.scaled(400, 300,
QtCore.Qt.KeepAspectRatio)
self.imgLab.setPixmap(pixmap)
ocr()
outcome('./result/audio/outcome.txt')
finally:
self.device.release()
def test():
r = request
print(r.files, flush=True)
file = r.files['image']
path = os.getcwd() + '/last.jpg'
file.save(path)
result = ocr.ocr(path)
response = {'ocr_result': result}
# encode response using jsonpickle
response_pickled = jsonpickle.encode(response)
return Response(response=response_pickled,
status=200,
mimetype="application/json")
def do_ocr():
id = uuid.uuid4()
path = "/tmp/"+str(id)+".png"
nparr = np.fromstring(request.data, np.uint8) #frombuffer?
img = cv2.imdecode(nparr, cv2.IMREAD_COLOR)
cv2.imwrite(path,img)
data = {}
text, coordinates = ocr.ocr(path)
data["text"] = text
data["coordinates"]=coordinates
response = json.dumps(data)
return Response(response=response, status=200, mimetype="application/json")
def login(username: str, password: str):
session = msession.session
res = session.get(msession.urls.cas, verify=False)
lt = re.findall(r'name="lt" value="(.*)"', res.text)
captcha_url = msession.urls.captcha
captcha_path = 'captcha.jpg'
with session.get(captcha_url) as captcha:
with open(captcha_path, mode='wb') as captcha_jpg:
captcha_jpg.write(captcha.content)
#captcha = input('验证码已生成于该目录下，请查看\n并输入验证码')
captcha = ocr(captcha_path)
login_form = {
'username': username,
'password': password,
'captcha': captcha,
'warn': 'true',
'lt': lt[0],
'execution': 'e1s1',
'_eventId': 'submit',
'submit': '登录'
}
post_res = session.post(msession.urls.cas, data=login_form)
if '账号或密码错误' in post_res.text:
print('账号或密码错误')
return
if '验证码不正确' in post_res.text:
print('验证码不正确')
return
os.remove('captcha.jpg')
session.get(msession.urls.sso, verify=False)
cookies = session.cookies
if not os.path.exists('cookies'):
os.mkdir('cookies')
if not cookies:
print('No cookies!')
else:
file_name = 'cookies' + os.sep + username
with open(file_name, mode='wb') as cookies_file:
pickle.dump(session.cookies, cookies_file)
def handler(event, context):
"""
download tar.gz, do ocr and upload it to configured cloud service
(currently AWS S3 or Google Drive)
"""
upload_type = os.environ.get('UPLOAD_TYPE', None)
if upload_type == 'gdrive':
for k in [
'GDRIVE_CLIENT_ID', 'GDRIVE_CLIENT_SECRET',
'GDRIVE_REFRESH_TOKEN'
]:
assert k in os.environ, "missing {} in environment vars".format(k)
elif upload_type == 's3':
assert 'S3_BUCKET' in os.environ
elif upload_type == 'discard':
pass
else:
raise Exception('unknown upload type {}'.format(
os.environ['UPLOAD_TYPE']))
empty_page_threshold = int(os.environ.get('EMPTY_PAGE_THRESHOLD', 200))
language = os.environ.get('TESSERACT_LANG', 'eng')
import boto3, ocr
s3 = boto3.client('s3')
for record in event['Records']:
src_bucket = record['s3']['bucket']['name']
src_file = record['s3']['object']['key']
s3.download_file(src_bucket, src_file, "/tmp/scan.tar.gz")
pdf_file = ocr.ocr("/tmp/scan.tar.gz", empty_page_threshold, language)
dest_filename = src_file.split('.')[0] + '.pdf'
if upload_type == 's3':
bucket = os.environ['S3_BUCKET']
s3.upload_file(pdf_file, bucket, dest_filename)
elif upload_type == 'gdrive':
folder = os.environ.get('GDRIVE_FOLDER', None)
client_id = os.environ['GDRIVE_CLIENT_ID']
client_secret = os.environ['GDRIVE_CLIENT_SECRET']
refresh_token = os.environ['GDRIVE_REFRESH_TOKEN']
upload_gdrive(pdf_file, dest_filename, client_id, client_secret,
refresh_token, folder)
elif upload_type == 'discard':
print('all fine, discarding file, but not deleting source file')
return
s3.delete_object(Bucket=src_bucket, Key=src_file)
os.remove(pdf_file)
def main():
program_options = processArguments()
if program_options['help']:
printUsage()
exit()
if len(program_options['image_paths']) == 0:
program_options['image_paths'].append('car.JPG')
print('Warning: Missing image path. Will use default image (' +
program_options['image_paths'][0] + ')')
printUsage()
json_output = []
for i, image_path in enumerate(program_options['image_paths']):
img = cv2.imread(image_path)
plate = detectplate(img)
if plate.size == 0:
if program_options['output'] == 'json':
output = {
'image_path': image_path,
'plates': [],
'error': 'Error: cannot detect plate in ' + image_path
}
json_output.append(output)
else:
print('Error: cannot detect plate in ' + image_path)
continue
chars = cropChars(plate)
if not program_options['silent']:
cv2.imshow('plate' + str(i), plate)
cv2.imshow('chars' + str(i), chars)
chars, x_cords = segmentCharacters(chars)
dic = sortingCharacters(chars, x_cords)
chars_text = []
for index, x in dic:
char_img = chars[index]
char_img = cv2.bitwise_not(char_img)
chars_text.append(ocr(char_img))
# cv2.imshow('char',char_img)
# cv2.waitKey(0)
# cv2.imwrite('./output/'+str(x)+'.png', char_img)
if program_options['output'] == 'json':
output = {'image_path': image_path, 'plates': [chars_text]}
json_output.append(output)
else:
print(' '.join(chars_text))
if program_options['output'] == 'json':
print(json.dumps(json_output, indent=4))
if not program_options['silent']:
cv2.waitKey(0)
cv2.destroyAllWindows()
def main():
#screenshot.snapscreen() #截图，保存到Steam目录
#print_time('Screenshot')
raw_location = screenshot.get_raw_location() #原图片的位置
#print(raw_location)
region.region(raw_location) #处理截图，保存到cache文件夹
#print_time('Region')
img_location = region.get_img_location() #bar图的位置
#print(img_location)
items = ocr(img_location) #物品名称列表
#print_time('OCR')
info = output.get_info(items) #获得价格列表
#print_time('Get_info')
out_GUI.popwindow(info) #输出到窗口中
def main():
for i in range(50):
image_number: str = str(i + 150).zfill(2)
target_image_path = f"movie_frames/0{image_number}.jpg"
trimed_image_path = trim_image(target_image_path)
image_texts: List[str] = ocr(trimed_image_path)
music_name = search_music_name(image_texts)
clear_rate = search_clear_rate(image_texts)
fullcombo_rate = search_fullcombo_rate(image_texts)
# print(image_texts)
print(image_number)
print(f"music_name:{music_name}")
print(f"clear_rate:{clear_rate}")
print(f"full_combo_rate:{fullcombo_rate}")
print()
#small test function; the inner loop
# tests each combination of # hidden nodes
# a 100 times, and the outerloop takes the
# average of those 100 times, and uses that
# as the general accuracy for that many hidden nodes in the ANN
def test(dm, dl, indices, nn):
avg = 0
for i in xrange(100):
rightguesses = 0
for j in indices:
test = dm[j]
prediction = nn.predict(test)
if dl[j] == prediction:
rightguesses += 1
avg += (rightguesses / float(len(indices)))
return avg / 100
#opens the data sets
dm = np.loadtxt(open('data.csv', 'rb'), delimiter = ',').tolist()
dl = np.loadtxt(open('dataLabels.csv', 'rb')).tolist()
#goes through every combination of # of hidden nodes and checks
# the accuracy of the ANN (prints it out)
for i in xrange(5, 100, 10):
nn = ocr(i, indices, dl, dm, False)
performance = test(dm, dl, indices, nn)
print "" + str(i) + " Hidden Nodes --> " + str(performance)
data = urllib2.urlopen(url).read()
soup = BeautifulSoup(data)
images = soup.findAll('img')
want = [i for i in images if i.attrs['height']=='960'][0].attrs['src']
if 'http' not in want:
want = 'http://www.klobysbbq.com/' + want
imageData = urllib2.urlopen(want).read()
imgFile = tempfile.mktemp(os.path.basename(want))
file(imgFile,'wb+').write(imageData)
ocrData = ocr.ocr(imgFile)
beerNames = []
for line in ocrData.split('\n'):
if len(line.strip()) == 0: continue
if 'Today' in line: continue
line = re.sub(percent, "", line)
line = re.sub(quoteToEnd, "", line)
line = re.sub(alphaNumSpace, "", line)
line = re.sub(lowerUpperLower, "\\1 \\2 \\3", line)
beerNames.append(line)
results = SearchForBeers(beerNames)