调用百度文字识别接口实现文字识别

目标：

1.学习使用百度AI开放平台进行语音识别与语音合成

       百度AI有两种开发方式：REST, SDK;    本文使用的是SDK python进行开发

        文档地址：https://ai.baidu.com/docs#/OCR-Python-SDK/53932383

2.将学到的内容：多线程，字符串转变为字典的方法

3.生成exe格式可执行文件

  pyinstaller -F xxx.py 参考文档 https://blog.csdn.net/qq_35203425/article/details/78568141

4.python-Tkinter图形界面开发

  1.单选按钮控件 Radiobutton  2.多选框按钮Checkbutton

问题点：

1.界面上的模块不会随窗口放大产生变化

2.未添加异常捕获

界面：

代码：

# 导包
from aip import AipOcr       #百度文字识别api
from tkinter import *         #python界面开发
from tkinter import ttk
from tkinter.filedialog import askopenfilename,askdirectory
import time
import json

class Text_Recognition():
    """ SDK开发"""
    def __init__(self):
        # 参数初始化化
        """ 你的 APPID AK SK """
        self.APP_ID = '17574626'
        self.API_KEY = 'nITPuSyAT6fwoRs029RR7zMl'
        self.SECRET_KEY = 'BjknUD067cBoEmv30GE7TXrFpdceHLgP'
        self.client = AipOcr(self.APP_ID, self.API_KEY, self.SECRET_KEY)
        self.myWindow = Tk()
        self.image_path=StringVar() #保存图片路径
        self.var_1=StringVar() #精度变量
        self.var_2 = StringVar()  # 位置信息
        self.var_3 = StringVar()  # 检测方向
        self.var_4= StringVar()   # 卡证识别类型
        self.card_list=['身份证正面','身份证背面','银行卡','营业执照','驾驶证','行驶证','车牌']
    def format(self,result,direction='false'):
        # print(result)
        string='文字识别结果'+'n'+'-'*25+'n'
        if direction=='true':
            string=string+'方向:{}'.format(result['direction'])+'n'
        for item in result['words_result']:
            string_1='(L:{},T:{},W:{},H:{}):{}'.format(item['location']["left"],item['location']["top"],item['location']["width"],item['location']["height"],item[ 'words'])
            string=string+string_1+'n'
        return string
    def image_open(self):
        """
        打开图片
        :return:
        """
        # 路径搜索
        self.name = askopenfilename()
        self.image_path.set(self.name)
        with open(self.name, 'rb') as fp:
            self.image= fp.read()
        # print(self.image)

    def basicgeneral(self):
        """
        通用图片识别
        :return:
        """
        self.text_21.delete(1.0,'end')
        string=''
        """ 如果有可选参数 """
        options = {}
        # options["recognize_granularity"] = "big"
        # options["language_type"] = "CHN_ENG"                    #识别语言类型
        options["detect_direction"] = str(self.var_3.get())      #是否检测图像朝向
        # options["detect_language"] = "true"                    #是否检测语言
        options["vertexes_location"] = str(self.var_2.get())     #是否返回文字外接多边形顶点位置
        # options["probability"] = "true"                         #每一行的置信度
        direction=str(self.var_3.get())
        """ 带参数调用通用文字识别, 图片参数为本地图片 """
        if self.var_1.get()=='0':
            result=self.client.general(self.image, options)       # 普通精度识别(含位置信息)
            # self.client.basicGeneral(self.image, options)  # 普通精度识别(不含位置信息)
            # print(result['words_result'])
            string=self.format(result,direction)
        elif self.var_1.get()=='1':
            result=self.client.accurate(self.image, options)   #高精度识别
            string=self.format(result, direction)
        self.text_21.insert('insert', string)
    def webimage(self):
        """
        网络图片识别
        :return:
        """
        """ 如果有可选参数 """
        self.text_31.delete(1.0,'end')
        options = {}
        options["detect_direction"] = "true"
        options["detect_language"] = "true"
        """ 带参数调用网络图片文字识别, 图片参数为本地图片 """
        result=self.client.webImage(self.image, options)
        self.text_31.insert('insert',str(result))
    def card_identification(self):
        """
        卡证识别
        :return:
        """
        self.text_41.delete(1.0,'end')
        string=''
        ident_id=int(self.var_4.get())  #返回的默认是字符串
        print(ident_id)
        if ident_id==0:
            idCardSide = "front"  #含照片面
            """ 调用身份证识别 """
            result=self.client.idcard(self.image, idCardSide)
            for item in result["words_result"].items():
                string_1 = '{}:{}n'.format(item[0], item[1]['words'])
                string = string + string_1
        elif ident_id==1:
            idCardSide = "back"  # 含国徽面
            """ 调用身份证识别 """
            result=self.client.idcard(self.image, idCardSide)
            for item in result["words_result"].items():
                string_1 = '{}:{}n'.format(item[0], item[1]['words'])
                string = string + string_1
        elif ident_id==2:
            """ 调用银行卡识别 """
            result=self.client.bankcard(self.image)
            string='银行卡卡号：{}n有效期：{}n银行名称{}'.format(result['result']['bank_card_number'],result['result']['valid_date'],result['result']['bank_name'])
        elif ident_id==3:
            """ 调用营业执照识别 """
            result=self.client.businessLicense(self.image)
            for item in result["words_result"].items():
                string_1 = '{}:{}n'.format(item[0], item[1]['words'])
                string = string + string_1
        elif ident_id==4:
            """ 如果有可选参数 """
            options = {}
            options["detect_direction"] = "true"

            """ 带参数调用驾驶证识别 """
            result=self.client.drivingLicense(self.image, options)
            for item in result["words_result"].items():
                string_1='{}:{}n'.format(item[0],item[1]['words'])
                string=string+string_1
        elif ident_id==5:
            """ 如果有可选参数 """
            options = {}
            options["detect_direction"] = "true"
            options["accuracy"] = "normal"

            """ 带参数调用行驶证识别 """
            result=self.client.vehicleLicense(self.image, options)
            for item in result["words_result"].items():
                string_1='{}:{}n'.format(item[0],item[1]['words'])
                string=string+string_1
        elif ident_id==6:
            """ 如果有可选参数 """
            options = {}
            options["multi_detect"] = "true"

            """ 带参数调用车牌识别 """
            result=self.client.licensePlate(self.image, options)
            # print(result)
            string='车牌颜色:{}n车牌号:{}'.format(result["words_result"][0]["color"],result["words_result"][0]["number"])

        self.text_41.insert('insert',string)
        # print(result_a)
    def table_identification(self):
        """
        表格文字识别
        :return:
        """
        """ 调用表格文字识别 """
        self.text_51.delete(1.0,'end')
        result=self.client.tableRecognitionAsync(self.image)
        requestId=result['result'][0]['request_id']
        """ 如果有可选参数 """
        options = {}
        options["result_type"] = "json"

        """ 带参数调用表格识别结果 """
        result_1=self.client.getTableRecognitionResult(requestId, options)
        # print(type(result_1['result']['ret_code']))
        while result_1['result']['ret_code']!=3:
            time.sleep(10)
            result_1 = self.client.getTableRecognitionResult(requestId, options)

        string=result_1['result']['result_data']
        self.text_51.insert('insert',string)
        # print(result_1['result']['ret_code'])
    def ticket_identification(self):
        """
        通用票据识别
        :return:
        """
        """ 如果有可选参数 """
        self.text_61.delete(1.0,'end')
        options = {}
        options["recognize_granularity"] = "big"
        options["probability"] = "true"
        options["accuracy"] = "normal"
        options["detect_direction"] = "true"

        """ 带参数调用通用票据识别 """
        result=self.client.receipt(self.image, options)
        string=self.format(result,"true")
        self.text_61.insert('insert',string)
        # print(result)
        # print('通用票据识别')
    def interface(self):
        """
        绘制界面
        :return:
        """
        self.myWindow.title('文字识别演示程序')  # 修改窗口标题
        self.myWindow.geometry("800x600+150+100")  # 修改窗口大小 width*height+xoffset+yoffset
        # 创建六个标签容器
        Entry(self.myWindow, textvariable=self.image_path, width=40).grid(row=0, column=0, sticky='SEW', pady=5, padx=5)
        Button(self.myWindow, text='打开图片', command=self.image_open, width=10, height=1).grid(row=0, column=1, sticky='SEW',pady=5, padx=5)

        #frm_1 = LabelFrame(self.myWindow, width=380, height=10)
        #frm_1.grid(row=0, column=0, ipadx=5, ipady=5, padx=10, pady=5, sticky='SNEW')
        frm_2 = LabelFrame(self.myWindow,text='通用文字识别', width=380, height=400)
        frm_2.grid(row=1, column=0, ipadx=5, ipady=5, padx=10, pady=5,rowspan=6,columnspan=2, sticky='SNEW')
        frm_3 = LabelFrame(self.myWindow, text='网络图片文字识别', width=380, height=100)
        frm_3.grid(row=7, column=0, ipadx=5, ipady=5, padx=10, pady=5, rowspan=2,columnspan=2, sticky='SNEW')
        frm_4 = LabelFrame(self.myWindow, text='卡证识别', width=380, height=180)
        frm_4.grid(row=0, column=2, ipadx=5, ipady=5, padx=10, pady=5, rowspan=5, sticky='SNEW')
        frm_5 = LabelFrame(self.myWindow, text='表格文字识别', width=380, height=105)
        frm_5.grid(row=5, column=2, ipadx=5, ipady=5, padx=10, pady=5, rowspan=2, sticky='SNEW')
        frm_6 = LabelFrame(self.myWindow, text='通用票据识别', width=380, height=105)
        frm_6.grid(row=7, column=2, ipadx=5, ipady=5, padx=10, pady=5, rowspan=2, sticky='SNEW')

        # frm_1控件编辑
        #Entry(frm_1, textvariable=self.image_path, width=40).grid(row=0, column=0, sticky='SEW', pady=5, padx=5)
        #Button(frm_1, text='打开图片', command=self.image_open, width=10, height=1).grid(row=0, column=1, sticky='EW', pady=5, padx=5)

        # frm_2控件编辑
        Radiobutton(frm_2,text='普通精度',variable=self.var_1, value='0').grid(row=0, column=0, sticky='EW', pady=5, padx=5)
        Radiobutton(frm_2, text='高精度', variable=self.var_1, value='1').grid(row=0, column=1, sticky='EW', pady=5,padx=5)
        self.var_1.set(0)  #设置Radiobutton默认选项
        Checkbutton(frm_2, text='含位置信息',variable=self.var_2, onvalue='true', offvalue='false').grid(row=1, column=0, sticky='EW', pady=5, padx=5)
        self.var_2.set('false')
        Checkbutton(frm_2, text='检查方向', variable=self.var_3, onvalue='true', offvalue='false').grid(row=1, column=1, sticky='EW',pady=5, padx=5)
        self.var_3.set('false')
        Button(frm_2, text='识别图片', command=self.basicgeneral, width=10, height=1).grid(row=1, column=2, sticky='EW',pady=5, padx=5)
        self.text_21 = Text(frm_2, width=50, height=15)
        self.text_21.grid(row=2, column=0, sticky='EW', pady=5, padx=5, columnspan=3)
        # frm_3控件编辑
        Button(frm_3, text='识别图片', command=self.webimage, width=4, height=1).grid(row=0, column=3, sticky='EW',pady=5, padx=5)
        self.text_31 = Text(frm_3, width=50, height=8)
        self.text_31.grid(row=1, column=0, sticky='EW', pady=5, padx=5, columnspan=4)
        # frm_4控件编辑
        i,j=0,0
        for a,item in enumerate(self.card_list):
            j=a % 4
            if j==0 and a!=0:
                i=i+1
            Radiobutton(frm_4, text=item, variable=self.var_4, value=a).grid(row=i, column=j, sticky='EW',pady=5, padx=5)
        self.var_4.set(1)
        Button(frm_4, text='识别图片', command=self.card_identification, width=6, height=1).grid(row=1, column=3, sticky='EW', pady=5,padx=5)
        self.text_41 = Text(frm_4, width=30, height=6)
        self.text_41.grid(row=2, column=0, sticky='EW', pady=5, padx=5, columnspan=4)
        # frm_5控件编辑
        Button(frm_5, text='识别图片', command=self.table_identification, width=4, height=1).grid(row=0, column=3, sticky='EW', pady=5,
                                                                                  padx=5)
        self.text_51 = Text(frm_5, width=50, height=6)
        self.text_51.grid(row=1, column=0, sticky='EW', pady=5, padx=5, columnspan=4)
        # frm_6控件编辑
        Button(frm_6, text='识别图片', command=self.ticket_identification, width=4, height=1).grid(row=0, column=3, sticky='EW', pady=5,
                                                                                  padx=5)
        self.text_61 = Text(frm_6, width=50, height=8)
        self.text_61.grid(row=1, column=0, sticky='EW', pady=5, padx=5, columnspan=4)

         # 启动主窗口的消息循环
        self.myWindow.mainloop()

if __name__ == '__main__':
    text=Text_Recognition()
    text.interface()

最后

以上就是粗暴滑板为你收集整理的调用百度文字识别接口实现文字识别的全部内容，希望文章能够帮你解决调用百度文字识别接口实现文字识别所遇到的程序开发问题。

如果觉得靠谱客网站的内容还不错，欢迎将靠谱客网站推荐给程序员好友。