1.準備：

1)Python開發環境, 筆者用的是3.7; 工具用的是Pycharm

2)百度云后臺創建文字識別的應用, 獲取AppID, API key, Secret Key

百度云后臺創建文字識別的應用

3) 百度模塊

pip install baidu-aip

安裝百度模塊

4) 要保存成csv需要用到pandas模塊

pip Install pandas

安裝pandas成功

2.上路：

1)初始化百度客戶端, 用來發送圖片信息

初始化百度客戶端

2)調用通用文字接口

調用通用文字接口并返回數據

這邊我們斷點查看一下返回來的數據:

返回結果

3)保存成CSV

這里不顯示返回數據的進一步處理過程，我把處理后的數據保存到全局變量Company_Data中。具體數據處理過程可以參考本文源碼或者在《Python使用騰訊Ocr識別文字》中的方法.下面把Company_Data中的數據保存成CSV.

保存成CSV

3.結果：

讀取圖片內容

讀取6張圖片

輸出結果:

輸出識別結果

4.完整代碼：

# -*- coding: utf-8 -*-

from aipimport AipOcr

import os

import pandas

# create a new AipOcr

APP_ID ="16921559"

API_KEY ="HfpMM13vAnDlTRWabQVDKnk8"

SECRET_KEY ="EQpdKCeICwfHLWazx0vsIpRqoRkVX6pM"

IMG_EXT = ['.png', '.jpg', '.jpeg', '.bmp']

Company_Data = []

# initialize

client = AipOcr(APP_ID, API_KEY, SECRET_KEY)

#1 get text content

def imageReader(file_path):

with open(file_path,'rb')as f:

content = f.read()

# general text API

api_result = client.basicGeneral(content)

# text content

words_result = []

for iin api_result['words_result']:

words_result.append(i['words'])

return words_result

#2 write the content into file

def saveData(file_path):

# separate the file name and extend type

filename, ext = os.path.splitext(file_path)

if extin IMG_EXT:

#? ? ? ? new_path = file_path + '.txt'

print(" reading the following image %s" % file_path)

result = imageReader(file_path)

#with open(new_path, 'w', encoding='utf-8') as f:

#? ? f.write(result)

# 電話和傳真前面加上'\t'，可以防止excel打開csv內容的時候自動計算

data = {}

for textin result:

if ':' in text:

itemname, value = text.split(":")

if '展位號' in itemname:

data['展位號'] = value

elif '地址' in itemname:

data['地址'] = value

elif '郵編' in itemname:

data['郵編'] = value

elif '電話' in itemname:

data['電話'] ='\t'+ value

elif '傳真' in itemname:

data['傳真'] ='\t'+ value

elif '聯系人' in itemname:

data['聯系人'] = value

elif '職務' in itemname:

data['職務'] = value

elif '電郵' in itemname:

data['電郵'] = value

elif '網址' in itemname:

data['網址'] = value

elif '業務性質' in itemname:

data['業務性質'] = value

elif '產品類型' in itemname:

data['產品類型'] = value

print(" Saving the data of %s" % file_path)

print(data)

Company_Data.append(data)

#3 get all .jpg under the path

def each_path(dir_path):

# get file names of current directory

file_name = os.listdir(dir_path)

for namein file_name:

if '.jpg' in name:

image_path = dir_path +'\\' + name

saveData(image_path)

def save2csv():

file_name ='test1.csv'

save = pandas.DataFrame(data=Company_Data, dtype='object')

try:

save.to_csv(file_name,quoting=1,float_format='str', mode='a+')

except UnicodeDecodeError:

print("編碼錯誤，該數據無法寫到文件中，直接忽略該數據")

if __name__ =='__main__':

file_path =r"F:\我的堅果云\其他\Study\Python\Demo\BaiduOcr\picture"

each_path(file_path)

save2csv()