Python如何读取pdf中的图片

<p style="line-height: 1.75em;"><span style="font-family: 微软雅黑, "Microsoft YaHei";">import fitz

import time

import re

import os<br></span></p>

<p style="line-height: 1.75em;"><span style="font-family: 微软雅黑, "Microsoft YaHei";">def pdf2pic(path, pic_path):

    '''

    # 从pdf中提取图片

    :param path: pdf的路径

    :param pic_path: 图片保存的路径

    :return:

    '''

    t0 = time.clock()

    # 使用正则表达式来查找图片

    checkXO = r"/Type(?= */XObject)"

    checkIM = r"/Subtype(?= */Image)"  <br></span></p>

<p style="line-height: 1.75em;"><span style="font-family: 微软雅黑, "Microsoft YaHei";">    # 打开pdf

    doc = fitz.open(path)

    # 图片计数

    imgcount = 0

    lenXREF = doc._getXrefLength()

    # 打印PDF的信息

    print("文件名:{}, 页数: {}, 对象: {}".format(path, len(doc), lenXREF - 1))<br></span></p>

<p style="line-height: 1.75em;"><span style="font-family: 微软雅黑, "Microsoft YaHei";">    # 遍历每一个对象

    for i in range(1, lenXREF):

        # 定义对象字符串

        text = doc.getObjectString(i)

        isXObject = re.search(checkXO, text)

        # 使用正则表达式查看是否是图片

        isImage = re.search(checkIM, text)

        # 如果不是对象也不是图片，则continue

        if not isXObject or not isImage:

            continue

        imgcount += 1

        # 根据索引生成图像

        pix = fitz.Pixmap(doc, i)

        # 根据pdf的路径生成图片的名称

        new_name = path.replace('\\', '_') + "_img{}.png".format(imgcount)

        new_name = new_name.replace(':', '')<br></span></p>

<p style="line-height: 1.75em;"><span style="font-family: 微软雅黑, "Microsoft YaHei";">        # 如果pix.n<5,可以直接存为PNG

        if pix.n < 5:

            pix.writePNG(os.path.join(pic_path, new_name))

        # 否则先转换CMYK

        else:

            pix0 = fitz.Pixmap(fitz.csRGB, pix)

            pix0.writePNG(os.path.join(pic_path, new_name))

            pix0 = None

        # 释放资源

        pix = None

        t1 = time.clock()

        print("运行时间:{}s".format(t1 - t0))

        print("提取了{}张图片".format(imgcount))<br></span></p>

<p style="line-height: 1.75em;"><span style="font-family: 微软雅黑, "Microsoft YaHei";">if __name__=='__main__':

    # pdf路径

    path = r'E:\dogcat\提取图片\计算机视觉算法工程师.pdf'

    pic_path = r'E:\dogcat\提取图片\测试'

    # 创建保存图片的文件夹

    if os.path.exists(pic_path):

        print("文件夹已存在，请重新创建新文件夹！")

        raise SystemExit

    else:

        os.mkdir(pic_path)

    m = pdf2pic(path, pic_path)<br></span></p>