python使用paddleocr实现识别pdf和图片_pycharm paddleocr pdf

这周周末又整理了一下代码,支持了pdf识别

直接上代码

方便自己 方便大家

from flask import Flask, request
from paddleocr import PaddleOCR, draw_ocr
from pdf2image import convert_from_path
import json
import pandas as pd
import sys
import datetime
import numpy as np
import numpy_financial as npf
from dateutil.relativedelta import relativedelta
import math
import warnings
 
import uuid
 
import os
 
 
warnings.filterwarnings('ignore')
 
 
pd.set_option('display.max_rows', None)
pd.set_option('display.max_columns', None)
np.set_printoptions(edgeitems=5, precision=3, suppress=True)
 
# Paddleocr目前支持的多语言语种可以通过修改lang参数进行切换
# 例如`ch`, `en`, `fr`, `german`, `korean`, `japan`
ocr = PaddleOCR(use_angle_cls=True, lang="ch")  # need to run only once to download and load model into memory
 
 
app = Flask(__name__)
 
# 支持的文件类型
supportFileType = ['pdf','img','image','jpg','jpeg','png']
 
# 校验参数合法性
def checkParam(file,fileType):
 
    # 默认返回内容
    return_result = {'code': '200', 'message': '处理成功', 'data': False}
    if file is None:
        print("文件参数为空")
        return_result['code'] = 'E100001'
        return_result['message'] = '文件参数为空'
        return json.dumps(return_result, ensure_ascii=False)
 
    if fileType is None:
        print("文件类型为空")
        errorMessage.append(",文件类型为空")
        return_result['code'] = 'E100002'
        return_result['message'] = '文件类型为空'
        return json.dumps(return_result, ensure_ascii=False)
    if fileType not in supportFileType:
        print("不支持的文件类型")
        errorMessage.append(",不支持的文件类型")
        return_result['code'] = 'E100003'
        return_result['message'] = '不支持的文件类型'
        return json.dumps(return_result, ensure_ascii=False)
 
 
    # 判断传入的json数据是否为空
    if request.get_data() is None:
    # if not request.get_data():
        return_result['return_code'] = '5004'
        return_result['return_info'] = '请求参数为空'
        return json.dumps(return_result, ensure_ascii=False)
 
# 进行图片ocr
def doImgOcr(tmpFileAbsoultePath,ocr_content):
    #OCR
    result = ocr.ocr(tmpFileAbsoultePath, cls=True)
    for line in result:
        print("line:")
        print(line)
        for tmpTwo in line:
          print("tmpTwo:")
          print(tmpTwo)
          for tmpThree in tmpTwo:
              print("tmpThree:")
              print(tmpThree)
              finalArr = str(tmpThree).split(",")
              finalStr = finalArr[0].replace("'","").replace("(","")
              print("finalStr:"+finalStr)
              if finalStr.find("[") < 0:
                ocr_content.append(finalStr)
 
    # 显示结果
    from PIL import Image
    img_path = tmpFileAbsoultePath
    image = Image.open(img_path).convert('RGB')
    boxes = [line[0] for line in result]
    txts = [line[1][0] for line in result]
    scores = [line[1][1] for line in result]
    #im_show = draw_ocr(image, boxes, txts, scores, font_path='./fonts/simfang.ttf')
    #im_show = Image.fromarray(im_show)
    #im_show.save('result.jpg')
 
 
    print("ocr_content:")
    print(ocr_content)
    # 删除文件
    #os.remove(tmpFileAbsoultePath)
 
    return ocr_content
 
# 进行pdf ocr
def doPdfOcr(tmpFileAbsoultePath,ocr_content):
    pages = convert_from_path(tmpFileAbsoultePath)
    print(pages)
 
    imgPathArr = []
 
    for i, page in enumerate(pages):
        imgNamePrefix = str(uuid.uuid4());
        imgNamePrefix = imgNamePrefix.replace("-","")
        imgName = '/opt/tmp/'+f'page_{i+1}_'+imgNamePrefix+'.png'
        imgPathArr.append(imgName)
        page.save(imgName, 'PNG')
        doImgOcr(imgName,ocr_content)
 
    print("imgPathArr:")
    print(imgPathArr)
 
# ocr识别请求
@app.route('/ocr', methods=['POST'])
def do_ocr():
 
    # ocr识别结果文字
    ocr_content = []
    # 默认返回内容
    return_result = {'code': '200', 'message': '处理成功', 'data': ''}
 
    # 保存文件
    if request.method == 'POST':
        file = request.files['file']
        fileType = request.form.get('fileType')
        print("文件类型:"+fileType)
 
        # 参数校验
        checkParam(file,fileType)
 
 
        tmpFilenamePrefix = str(uuid.uuid4());
        print("tmpFilenamePrefix:"+tmpFilenamePrefix)
        tmpFilenamePrefix = tmpFilenamePrefix.replace("-","")
        tmpFileAbsoultePath = "/opt/" +tmpFilenamePrefix + "." + fileType
        print("tmpFileAbsoultePath:"+tmpFileAbsoultePath)
        file.save(tmpFileAbsoultePath)
 
        if fileType == 'pdf':
            # 文件为pdf
            doPdfOcr(tmpFileAbsoultePath,ocr_content)
        else: 
            # 文件为图片
            doImgOcr(tmpFileAbsoultePath,ocr_content);
 
 
        return_result = {'code': '200', 'message': '处理成功', 'data': ocr_content}
        return json.dumps(return_result, ensure_ascii=False)
 
 
 
if __name__ == "__main__":
 
    app.run(host="0.0.0.0", port="5000", debug=True)