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在其上进行ocr，但出现错误：

“在通常的地方找不到ghostscript”

搜索后，我找到了在Windows平台中将Ghostscript链接到pypdfocr的解决方案，我尝试下载GhostScript并将其放入环境变量中，但仍然存在相同的错误。

如何使用python在扫描的pdf文件中查找文本？

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”：

打开终端并更改路径

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，请查看以下链接

/sf/ask/3688972591/ 
error-constitute-c-readimage-412

/sf/ask/3700336251/ 
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

看看这个库：https : //pypi.python.org/pypi/pypdfocr 但 PDF 文件中也可以包含图像。您或许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，因此您不会获得带有 ghostscript 的文本。

Answer 3

转换pdf，使用pytesseract做OCR，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)