身份证信息简单识别并保存

csx

因为工作原因经常需要输入身份证信息到excel
基于百度文字识别功能简单写了个身份证图片识别
有需要可以参考下

[Python] 纯文本查看 复制代码

from aip import AipOcr
import glob,re, time
import openpyxl

APPID=
APIKEY=
SECRETKET=
client=AipOcr(APPID,APIKEY,SECRETKEY)

def get_path(path):#获取图片路径
      return glob.iglob(r".\%s\**\*[jpg,png]"%(path),recursiwe=True)

def open file(filePath) :#读取图片
     with open(filePath,'rb') as fp:
            return fp.read()

def get text (fileName] :#转换文本
      image = open_file(fileName)
      dic_result = client.basicGeneral(image)

      #time.sleep(1)
      res = dic_result['words_result’]
      result=''
      for m in res:
           result=result+str (m['words’])
      return result

parse_text (aaa):#解析文本
      list1=[]
      list2=[]
      for i in aaa:
            ls_dict={}
            text=get_text(i)
            if re.search('认民身份号码',text):
                  try:
                        xm=re.search(r"名(.*?)性别",text).group(1)
                        sf=re.search(r"号码([\d|X|x]*)",text).group(1)
                        ls_dict['xm']=xm
                        ls_dict['sf']=sf
                        list1.append(ls_dict)
                 except:
                        continue
           else:
                 list2.append(text)
      return list1

def write_excel(bbb):#写入excel
      wb = openpyxl.Workbook()
      sheet = wb.active
      sheet['A1']='姓名'
      sheet['B1']='身份证号码'
      row=2
      for i in bbb:
            sheet.cell(row,1,i['xm'])
            sheet.cell(row,2,i['sf'])
            row += 1
            print(f"            {i['xm']}已录入…")
      wb.save("./人员明细.xlsx")

def main():
      path='aaa'
      aaa=get_path(path)
      aaa=list(aaa)

      print('正在解析图片信息....')
      print(f'大约需要{int(len(aaa)*2)}秒!'}
      bbb=parse_text(aaa)
      write_excel(bbb)
      print('录入完毕!!!')
      time.sleep(2)
if __name__=='__main__':
      main()