开源的识别图片中文字-526互联

https://blog.csdn.net/weixin_38809962/article/details/126966598


import logging
from aip import AipOcr
import easyocr
import cv2
import numpy as np

logger = logging.getLogger(__name__)


def cv2_readimg(filename, mode):
    # 把图片文件存入内存
    img_date = np.fromfile(filename, dtype=np.uint8)
    # 从内存数据读入图片
    img = cv2.imdecode(img_date, mode)
    return img


def baiduOCR(picfile):  # picfile:图片文件名
    # APP_ID = '24656072'
    # API_KEY = '30HUE1gfGA7Qoqkgn5tAXHKm-1'
    # SECRET_KEY = 'hOk41TfUk4nmlboFBMilTxoP67WDwMIA-2'
    # client = AipOcr(APP_ID, API_KEY, SECRET_KEY)
    # i = open(picfile, 'rb')
    # img = i.read()
    # # message = client.basicAccurate(img)  # """ 调用通用文字识别（高精度版）,每天800次免费 """
    # # message = client.basicGeneral(img)  # 通用文字识别，每天 50 000 次免费
    # message = client.accurate(img)
    # # message = client.general(img)
    # # message = client.general(img)
    # # message = client.enhancedGeneral(img)
    # # while message is None:
    # #     message = client.basicAccurate(img)
    # #     if message is None:
    # #         message = client.general(img)
    # #         if message is None:
    # #             message = client.basicGeneral(img)
    # #         else:
    # #             break
    # #     else:
    # #         break
    # i.close()
    # # 输出文本内容
    # info_list = []
    # for text in message.get('words_result'):
    #     info_list.append(text.get('words'))
    # logger.info("图像识别返回的信息列表为：{}".format(info_list))
    # return str(info_list)

    # 以下为easyocr检测，只需要运行一次就可以将模型加载到内存中
    reader = easyocr.Reader(['ch_sim', 'en'], gpu=False)  # 简体中文、英文检测
    picfile = cv2_readimg(picfile, cv2.IMREAD_GRAYSCALE)
    origin_list = reader.readtext(picfile)
    result_list = []
    for text in origin_list:
        result_list.append(text[1])
    result_str = str(result_list)
    logger.info("识别出的文字为：{}".format(result_str))
    return result_str

文字

图片