In [93]:
from PIL import Image
import matplotlib.pyplot as plt
import pytesseract
import cv2
In [94]:
def cv2ToPil(img):
return Image.fromarray(cv2.cvtColor(img,cv2.COLOR_BGR2RGB))
In [123]:
img = cv2.imread('validateCode.jpeg')
grayImg = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
thresh = cv2.threshold(grayImg, 0, 255, cv2.THRESH_BINARY | cv2.THRESH_OTSU)[1]
retImg = cv2.medianBlur(thresh, 3)
In [124]:
grayImg = cv2ToPil(grayImg)
thresh = cv2ToPil(thresh)
retImg = cv2ToPil(retImg[2:-2, 2:-2])
In [125]:
plt.imshow(img)
Out[125]:
<matplotlib.image.AxesImage at 0x118e20a90>
In [126]:
plt.imshow(grayImg)
Out[126]:
<matplotlib.image.AxesImage at 0x1193c8cc0>
In [127]:
plt.imshow(thresh)
Out[127]:
<matplotlib.image.AxesImage at 0x119475978>
In [128]:
plt.imshow(retImg)
Out[128]:
<matplotlib.image.AxesImage at 0x1196707f0>
In [130]:
pytesseract.image_to_string(retImg)
Out[130]:
'3858'
定义函数，添加参数增加识别率¶
- 转灰度图
- 分离图片的前景与背景
- 过滤噪音
- 去除边框
In [142]:
def recogVerifCode(imgPath):
img = cv2.imread(imgPath)
grayImg = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
thresh = cv2.threshold(grayImg, 0, 255, cv2.THRESH_BINARY | cv2.THRESH_OTSU)[1]
retImg = cv2.medianBlur(thresh, 3)
verifcodeImg = retImg[2:-2, 2:-2]
return pytesseract.image_to_string(verifcodeImg, config='--psm 6 --oem 3 -c tessedit_char_whitelist=0123456789')
In [143]:
recogVerifCode('validateCode.jpeg')
Out[143]:
'3858'
In [144]:
recogVerifCode('verif_code.jpeg')
Out[144]:
'3213'
In [145]:
recogVerifCode('verif_code1.jpeg')
Out[145]:
'0465'