ocr - python识别验证码

Question

我只能做成这样：

用pytesseract来提取，还是不够准确。有谁知道怎样去掉那些黑点吗？

去掉黑点已解决。
但为什么这时用image_to_string函数会报错的？平时都没错

Traceback (most recent call last):
  File "test1.py", line 10, in <module>
    text = pytesseract.image_to_string(Image.open('temp2.jpg'))
  File "/usr/local/lib/python3.5/site-packages/pytesseract/pytesseract.py", line 163, in image_to_string
    errors = get_errors(error_string)
  File "/usr/local/lib/python3.5/site-packages/pytesseract/pytesseract.py", line 111, in get_errors
    error_lines = tuple(line for line in lines if line.find('Error') >= 0)
  File "/usr/local/lib/python3.5/site-packages/pytesseract/pytesseract.py", line 111, in <genexpr>
    error_lines = tuple(line for line in lines if line.find('Error') >= 0)
TypeError: a bytes-like object is required, not 'str'

from PIL import Image,ImageEnhance,ImageFilter
import pytesseract

im = Image.open("temp.jpg")
im = im.filter(ImageFilter.MedianFilter())
enhancer = ImageEnhance.Contrast(im)
im = enhancer.enhance(2)
im = im.convert('1')
im.save('temp2.jpg')
text = pytesseract.image_to_string(Image.open('temp2.jpg'))
print(text)

Answer 1

第二张图要去掉黑点，就是一个二值化降噪的过程。可以用PIL（Python Image Library）试试

代码：

from PIL import Image,ImageEnhance,ImageFilter

im = Image.open("1.jpeg")
im = im.filter(ImageFilter.MedianFilter())
enhancer = ImageEnhance.Contrast(im)
im = enhancer.enhance(2)
im = im.convert('1')
im.show()

效果：

处理前

处理后

Answer 2

http://www.jianshu.com/p/9f74...