Batch recognition of tabular data - Python calls Baidu Cloud API

Recognition effect display

original form

 Identify the processed form, you can see that the effect is very good

 

Get KEY

Baidu Smart Cloud (baidu.com)

Products - Text Recognition - General Scenarios

to collect

 General scene OCR - all - 0 yuan to receive

 to create

Fill in the application name - select all - the application description at the bottom - create immediately - return to the application list

These two keys are copied to client_id and client_secret in main_GUI respectively

 

The first step we get access_token

import requests
import json
import base64
import time

def get_access_token():
    client_id = '你的apikey'
    client_secret = '你的Secret Key'
    # client_id 为官网获取的AK， client_secret 为官网获取的SK
    host = 'https://aip.baidubce.com/oauth/2.0/token?grant_type=client_credentials&client_id={}&client_secret={}'.format(
        client_id, client_secret)
    response = requests.get(host).text
    data = json.loads(response)
    access_token = data['access_token']
    return access_token
# 获取返回信息

In the second step, we obtain the json data of whether the recognition is successful

If it is successfully recognized, it will give you a request_id, use this to get the excel table, or you can choose to get the json data (note that the picture must be converted to base64 encoding, support PNG, JPG, JPEG, BMP, TIFF, PNM, WebP)

#在这里我把文件路径放在了函数里面，可以放在外面，也可以做一个循环，识别一个文件夹里面的所有图片。
def get_info(access_token):

    request_url = "https://aip.baidubce.com/rest/2.0/solution/v1/form_ocr/request"
    # 二进制方式打开图片文件
    f = open('你需要识别的文件', 'rb')
    img = base64.b64encode(f.read())#base64编码
    params = {"image": img}
    request_url = request_url + "?access_token=" + access_token
    headers = {'content-type': 'application/x-www-form-urlencoded'}
    response = requests.post(request_url, data=params, headers=headers)
    # if response:
    #     print(response.json())
    data_1 = response.json()
    return data_1

In the third step, we get the excel sheet.

If your result_type is excel, it will return you an excel web page connection, if it is json, it will return json data

def get_excel(requests_id, access_token):
    headers = {'content-type': 'application/x-www-form-urlencoded'}
    pargams = {
        'request_id': requests_id,
        'result_type': 'excel'
    }
    url = 'https://aip.baidubce.com/rest/2.0/solution/v1/form_ocr/get_request_result'
    url_all = url + "?access_token=" + access_token
    res = requests.post(url_all, headers=headers, params=pargams)#访问链接获取excel下载页
    info_1 = res.json()['result']['ret_msg']
    excel_url=res.json()['result']['result_data']
    excel_1=requests.get(excel_url).content
    with open('识别结果.xls','wb+') as f:
        f.write(excel_1)
    print(info_1)

full code

The following is the code that can process images in batches and has a GUI interface

 

import os
import requests
import json
import base64
import time
import tkinter as tk
from tkinter import filedialog
from tkinter import ttk
# 获取access_token
def get_access_token():
    client_id = '你的apikey'
    client_secret = '你的apikey'
    # client_id 为官网获取的AK， client_secret 为官网获取的SK
    host = 'https://aip.baidubce.com/oauth/2.0/token?grant_type=client_credentials&client_id={}&client_secret={}'.format(
        client_id, client_secret)
    response = requests.get(host).text
    data = json.loads(response)
    access_token = data['access_token']
    return access_token

# 获取识别结果
def get_info(access_token, image):
    request_url = "https://aip.baidubce.com/rest/2.0/solution/v1/form_ocr/request"
    # 二进制方式打开图片文件
    img = base64.b64encode(image)  # base64编码
    params = {"image": img}
    request_url = request_url + "?access_token=" + access_token
    headers = {'content-type': 'application/x-www-form-urlencoded'}
    response = requests.post(request_url, data=params, headers=headers)
    data_1 = response.json()
    return data_1

# 获取excel
def get_excel(requests_id, access_token, image_file, output_dir):
    headers = {'content-type': 'application/x-www-form-urlencoded'}
    pargams = {
        'request_id': requests_id,
        'result_type': 'excel'
    }
    url = 'https://aip.baidubce.com/rest/2.0/solution/v1/form_ocr/get_request_result'
    url_all = url + "?access_token=" + access_token
    res = requests.post(url_all, headers=headers, params=pargams)  # 访问链接获取excel下载页
    info_1 = res.json()['result']['ret_msg']
    excel_url = res.json()['result']['result_data']
    excel_1 = requests.get(excel_url).content
    excel_name = '第{}.xlsx'.format(image_file.split('.')[0])
    excel_path = os.path.join(output_dir, excel_name)
    with open(excel_path, 'wb+') as f:
        f.write(excel_1)
    print('图片 {} 处理完成'.format(image_file))
    print(info_1)

def process_images(input_dir, output_dir, progress_bar):
    access_token = get_access_token()
    image_files = os.listdir(input_dir)  # 获取文件夹中的所有图片文件
    total_images = len(image_files)
    for i, image_file in enumerate(image_files):
        image_path = os.path.join(input_dir, image_file)
        with open(image_path, 'rb') as f:
            image = f.read()
        data_1 = get_info(access_token, image)
        try:
            requests_id = data_1['result'][0]['request_id']
            if requests_id != '':
                print('识别完成')
        except:
            print('识别错误')
        print('正在获取excel')
        time.sleep(10)  # 延时十秒让网页图片转excel完毕，excel量多的话，转化会慢，可以延时长一点
        get_excel(requests_id, access_token, image_file, output_dir)
        progress = (i + 1) / total_images * 100
        progress_bar['value'] = progress
        progress_bar.update()

def select_input_dir():
    input_dir = filedialog.askdirectory()
    input_dir_entry.delete(0, tk.END)
    input_dir_entry.insert(tk.END, input_dir)

def select_output_dir():
    output_dir = filedialog.askdirectory()
    output_dir_entry.delete(0, tk.END)
    output_dir_entry.insert(tk.END, output_dir)

def start_processing():
    input_dir = input_dir_entry.get()
    output_dir = output_dir_entry.get()
    progress_bar['value'] = 0
    progress_bar.update()
    process_images(input_dir, output_dir, progress_bar)
    progress_bar['value'] = 100
    progress_bar.update()

# 创建GUI窗口
window = tk.Tk()
window.title("图片处理")
window.geometry("400x250")

# 输入地址
input_dir_label = tk.Label(window, text="输入地址:")
input_dir_label.pack()
input_dir_entry = tk.Entry(window, width=40)
input_dir_entry.pack()
input_dir_button = tk.Button(window, text="选择文件夹", command=select_input_dir)
input_dir_button.pack()

# 输出地址
output_dir_label = tk.Label(window, text="输出地址:")
output_dir_label.pack()
output_dir_entry = tk.Entry(window, width=40)
output_dir_entry.pack()
output_dir_button = tk.Button(window, text="选择文件夹", command=select_output_dir)
output_dir_button.pack()

# 进度条
progress_bar = ttk.Progressbar(window, orient='horizontal', length=300, mode='determinate')
progress_bar.pack()

# 开始处理按钮
start_button = tk.Button(window, text="开始处理", command=start_processing)
start_button.pack()

# 运行GUI窗口
window.mainloop()

reference:

(1 message) Python calls Baidu Cloud API to identify form recognition_Call Baidu interface to identify form_Bai Pang Pang Song Bangbang's Blog-CSDN Blog