def __ocr_check(self, file_path, save_path, option=None, change_image=True):
"""
选择百度ocr识别，还是tesseract识别；以及图片二值化是否开启，
:param file_path: ocr识别图片路径
:param save_path: ocr结果保存路径，建议使用txt跟tesseract同一
:param option: 对tesseract的命令传递
:param change_image:是否进行二值化，默认使用二值化
:return:
"""
global enable_api
if change_image:
binarization_image(file_path)
if enable_api:
try:
ocr(file_path, save_path + ".txt")
except ConnectionError:
self.shell_color.failure_text("[!] 百度API无法连接")
enable_api = False
self.shell_color.helper_text("继续使用tesseract")
if option is not None:
option = " " + option
os.popen(
"tesseract {} {}{}".format(file_path, save_path, option) + self.__rebase_to_null
)
self.__wait(3)
else:
if option is not None:
option = " " + option
os.popen(
"tesseract {} {}{}".format(file_path, save_path, option) + self.__rebase_to_null
)
self.__wait(3)
def __ocr_check(self,
file_path, # 输入文件路径
save_path, # 输出文件路径
option=None, # 附加选项
change_image=True): # 是否预处理图片
self.shell_log.debug_text("base.__ocr_check")
global enable_baidu_api
if change_image:
if enable_baidu_api:
if enable_help_baidu:
binarization_image(filepath=file_path, save_backup=True)
else:
self.shell_log.info_text("不对百度ocr进行图像处理")
else:
binarization_image(filepath=file_path, save_backup=True)
if enable_baidu_api:
try:
ocr(file_path, save_path + ".txt")
except ConnectionError:
self.shell_log.failure_text("[!] 百度API无法连接")
enable_baidu_api = False
self.shell_log.helper_text("继续使用tesseract")
if not enable_baidu_api:
self.shell_log.debug_text("使用 tesseract OCR")
if option is not None:
option = " " + option
os.popen(
'tesseract "{}" "{}" {}'.format(file_path, save_path, option) + self.__rebase_to_null
)
self.__wait(3)
else:
self.shell_log.debug_text("使用 baidu api")
def __ocr_check(self,
file_path,
save_path,
option=None,
change_image=True):
# 对图形进行二值化操作
if change_image:
binarization_image(file_path)
if config.Enable_api:
try:
ocr(file_path, save_path + ".txt")
# Baidu_OCR.ocr(file_path, save_path+".txt")
except ConnectionError:
self.shell_color.failure_text("[!] 百度API无法连接")
config.Enable_api = False
self.shell_color.helper_text("继续使用tesseract")
if option is not None:
option = " " + option
os.popen(
"tesseract {} {}{}".format(file_path, save_path, option) +
self.__rebase_to_null)
self.__wait(3)
else:
if option is not None:
option = " " + option
os.popen("tesseract {} {}{}".format(file_path, save_path, option) +
self.__rebase_to_null)
self.__wait(3)
def get_loacation(filepath, option="makebox", change_image=False):
if change_image:
# image = Image.open(filepath)
# 对图片进行阈值过滤，然后保存
# image = image.point(lambda x: 0 if x < 143 else 255)
# image.save(filepath)
binarization_image(filepath)
os.popen('tesseract "{}" "{}" {}'.format(filepath, filepath, option))