将扫描的pdf转换为文本python

Question

我有一个扫描的 pdf 文件，我尝试从中提取文本。 我尝试使用 pypdfocr 对其进行 ocr，但出现错误：

“在通常的地方找不到ghostscript”

经过搜索，我找到了这个解决方案Linking Ghostscript to pypdfocr in Windows Platform，我尝试下载GhostScript并将其放入环境变量中，但仍然出现同样的错误。

如何使用 python 在扫描的 pdf 文件中搜索文本？

谢谢。

编辑：这是我的代码示例：

import os
import sys
import re
import json
import shutil
import glob
from pypdfocr import pypdfocr_gs
from pypdfocr import pypdfocr_tesseract 
from PIL import Image

path = PATH_TO_MY_SCANNED_PDF
mainL = []
kk = {}


def new_init(self, kk):
    self.lang = 'heb'   
    self.binary = "tesseract"
    self.msgs = {
            'TS_MISSING': """ 
                Could not execute %s
                Please make sure you have Tesseract installed correctly
                """ % self.binary,
            'TS_VERSION':'Tesseract version is too old',
            'TS_img_MISSING':'Cannot find specified tiff file',
            'TS_FAILED': 'Tesseract-OCR execution failed!',
        }

pypdfocr_tesseract.PyTesseract.__init__ = new_init  

wow = pypdfocr_gs.PyGs(kk)
tt = pypdfocr_tesseract.PyTesseract(kk)


def secFile(filename,oldfilename):
    wow.make_img_from_pdf(filename)


    files = glob.glob("X:/e206333106/ocr-114/balagan/" + '*.jpg')  
    for file in files:
        im = Image.open(file)
        im.save(file + ".tiff") 

    files = glob.glob("PATH" + '*.tiff')  
    for file in files:
        tt.make_hocr_from_pnm(file)
    pdftxt = ""    
    files = glob.glob("PATH" + '*.html') 
    for file in files:
        with open(file) as myfile:
            pdftxt = pdftxt + "#" + "".join(line.rstrip() for line in myfile)
    findNum(pdftxt,oldfilename)

    folder ="PATH"

    for the_file in os.listdir(folder):
        file_path = os.path.join(folder, the_file)
        try:
            if os.path.isfile(file_path):
                os.unlink(file_path)
        except Exception, e:
            print e

def pdf2ocr(filename):
    pdffile = filename
    os.system('pypdfocr -l heb ' + pdffile)

def ocr2txt(filename):  
    pdffile = filename


    output1 = pdffile.replace(".pdf","_ocr.txt")
    output1 = "PATH" + os.path.basename(output1)

    input1 = pdffile.replace(".pdf","_ocr.pdf")

    os.system("pdf2txt" -o  + output1 + " " + input1) 

    with open(output1) as myfile:
        pdftxt="".join(line.rstrip() for line in myfile)
    findNum(pdftxt,filename)


def findNum(pdftxt,pdffile):
    l = re.findall(r'\b\d+\b', pdftxt)


    output = open('PATH' + os.path.basename(pdffile) + '.txt', 'w')
    for i in l:
        output.write(",")
        output.write(i)
    output.close()    

def is_ascii(s):
    return all(ord(c) < 128 for c in s)

i = 0     
files = glob.glob(path + '\\*.pdf') 
print path  
print files 
for file in files:
    if file.endswith(".pdf"):
        if is_ascii(file):
            print file
            pdf2ocr(file)    
            ocr2txt(file)
        else:
            newname = "PATH" + str(i) + ".pdf"
            shutil.copyfile(file, newname)
            print newname
            secFile(newname,file)
        i = i + 1

files = glob.glob(path + '\\' + '*_ocr.pdf')         

for file in files:
    print file
    shutil.copyfile(file, "PATH" + os.path.basename(file))
    os.remove(file)

Answer 1

看看我的代码，它对我有用。

import os
import io
from PIL import Image
import pytesseract
from wand.image import Image as wi
import gc



pdf=wi(filename=pdf_path,resolution=300)
pdfImg=pdf.convert('jpeg')

imgBlobs=[]
extracted_text=[]

def Get_text_from_image(pdf_path):
    pdf=wi(filename=pdf_path,resolution=300)
    pdfImg=pdf.convert('jpeg')
    imgBlobs=[]
    extracted_text=[]
    for img in pdfImg.sequence:
        page=wi(image=img)
        imgBlobs.append(page.make_blob('jpeg'))

    for imgBlob in imgBlobs:
        im=Image.open(io.BytesIO(imgBlob))
        text=pytesseract.image_to_string(im,lang='eng')
        extracted_text.append(text)

    return (extracted_text)

我通过编辑 /etc/ImageMagick-6/policy.xml 为我修复了它，并将 pdf 行的权限更改为“读|写”：

打开终端，改变路径

cd /etc/ImageMagick-6
nano policy.xml
<policy domain="coder" rights="read" pattern="PDF" /> 
change to
<policy domain="coder" rights="read|write" pattern="PDF" />
exit

当我将 pdf 图像提取为文本时，我遇到了一些问题，请通过以下链接

https://stackoverflow.com/questions/52699608/wand-policy-error- 
error-constitute-c-readimage-412

https://stackoverflow.com/questions/52861946/imagemagick-not- 
authorized-to-convert-pdf-to-an-image

Increasing the memory limit  please go through the below link
enter code here
https://github.com/phw/peek/issues/112
https://github.com/ImageMagick/ImageMagick/issues/396

Answer 2

转换 pdf，使用 pytesseract 进行 OCR，并将 pdf 中的每一页导出为文本文件。

安装这些....

conda install -c conda-forge pytesseract

conda install -c conda-forge tesseract

pip 安装 pdf2image

import pytesseract
from pdf2image import convert_from_path
import glob

pdfs = glob.glob(r"yourPath\*.pdf")

for pdf_path in pdfs:
    pages = convert_from_path(pdf_path, 500)

    for pageNum,imgBlob in enumerate(pages):
        text = pytesseract.image_to_string(imgBlob,lang='eng')

        with open(f'{pdf_path[:-4]}_page{pageNum}.txt', 'w') as the_file:
            the_file.write(text)

Answer 3

看看这个库：https://pypi.python.org/pypi/pypdfocr 但 PDF 文件中也可以包含图像。您可能能够分析页面内容流。一些扫描仪将单个扫描页面分解为图像，因此您不会使用 ghostscript 获得文本。

Answer 4

PyPDF2 是一个作为 PDF 工具包构建的 Python 库。它能够：

Extracting document information (title, author, …)
Splitting documents page by page
Merging documents page by page
Cropping pages
Merging multiple pages into a single page
Encrypting and decrypting PDF files
and more!

要安装 PyPDF2，请从命令行运行以下命令：

pip install PyPDF2

代码：

import PyPDF2 

pdfFileObj = open('myPdf.pdf', 'rb') 


pdfReader = PyPDF2.PdfFileReader(pdfFileObj) 

print(pdfReader.numPages) 

pageObj = pdfReader.getPage(0) 

print(pageObj.extractText()) 

pdfFileObj.close() 

Answer 5

此解决方案适用于 Linux 操作系统（NoelOCR）

1. 安装 NoelOCR

    pip3 install NoelOCR
   

2. 使用它

    import NoelOCR as nm
    text = nm.processPDF('input.pdf')
    print(text)
   

之后，您应该从扫描的 PDF 中获取纯文本。