def getResult(self, src, save_image=False):
"""
@param {string} src
{bool}save_image output debug image
@return {OCRDocument} doucument
"""
pro = DataProcessor(src, ImageType.RAW, save_image=save_image)
if pro.prepare() is None:
logger.error('image error:{0}'.format(src))
return None
buffer = pro.tobinary(pro.batch())
temp_file_name = self.create_TemporyFile(buffer, True)
document = self.ocr.recognize(temp_file_name)
os.remove(temp_file_name)
output = '#' + datetime.now().strftime('%F %T.%f')[:-3] + '\n'
output += '\n'.join(document.names()) + '\n'
with Serializer.open_stream('../temp/corpus.txt', mode='a') as file:
file.write(output)
# ocr corpus data -> NaiveBayes classifier
# ranking name swap
change = self.naivebayes.predict_all(document.names())
#doucument.changeNames(change)
document.dump()
return document
def __init__(self, media, target=None):
self.media = media
self.target = target
pro = DataProcessor(self.media, ImageType.PLAN)
pro.prepare()
self.binary = pro.batch()
self.color = pro.color
self.drawClipSource = True
self.imageout_dir = '../temp/trash'
os.makedirs(self.imageout_dir, exist_ok=True)
