结合Tesseract引擎解决文本混淆反爬-Python Spider

In [1]: from PIL import Image

In [2]: import pytesseract

In [3]: text = pytesseract.image_to_string(Image.open('C:/Users/ASUS/totals.png'))

In [4]: text #测试tesseract引擎是否能用
Out[4]: '1'

In [5]: import json

In [6]: import requests

In [7]: render = 'http://www.porters.vip:8050/execute'

In [8]: url = 'http://www.porters.vip/confusion/movie.html'

In [9]: script = """
   ...: function main(splash)
   ...:   assert(splash:go('%s'))
   ...:   assert(splash:wait(0.5))
   ...:   total_png = splash:select('.movie-index-content.box .stonefont'):png()
   ...:   return{
   ...:   total = total_png
   ...:   }
   ...: end
   ...: """ % url

In [10]: header = {'content-type':'application/json'}

In [11]: data = json.dumps({"lua_source":script})

In [12]: resp = requests.post(render,data = data,headers = header) #截图

In [13]: resp
Out[13]: <Response [200]>

In [14]: image = resp.json()

In [15]: import base64

In [16]: import os

In [17]: for key,value in image.items():#有可能有多张图片
    ...:     #splash返回的图片数据以base64进行编码，我们需要解码
    ...:     image_body = base64.b64decode(value)
    ...:     filename = "%s.png"%key
    ...:     path = os.path.join(os.path.dirname(os.path.abspath('C:/Users/ASUS')),filename)
    ...:     print(path)
    ...:     with open(filename,'wb') as f:
    ...:         f.write(image_body) #存入指定位置
    ...:
    ...:
D:\total.png

In [21]: print(pytesseract.image_to_string('C:\\Users\\ASUS\\total.png')) #调用引擎进行识别
56.83

Tesseract安装地址：安装地址

安装教程：傻瓜式下一步+将Tesseract-OCR文件夹添加至环境变量+pip install pytesseract