A package used for OCR.
Project description
QQOcr
A package used for simple OCR.
-
Contact the author: jugking6688@gmail.com
-
Our bilibili: https://space.bilibili.com/3493127383943735
Examples are as follows:
For learn:
from qqocr import QQOcr
# You must provide a method to binarize the characters in the picture
# and import the external library in the function.
def binary(image):
import numpy as np
import cv2
low_range = np.array([0, 0, 0][::-1])
high_range = np.array([100, 100, 100][::-1])
return cv2.inRange(image, low_range, high_range)
qq = QQOcr()
# Dataset folder consists of many pictures and a 'label.txt'.
# For 'label.txt', the format of each line is "[filename]\t[text]".
# For example, it can be: "1.png 12345".
qq.load_dataset('./dataset')
qq.set_binary(binary)
qq.learn()
# The suffix must be '.qmodel'.
qq.save_model('./1.qmodel')
For predict:
from qqocr import QQOcr
import cv2
qq = QQOcr()
qq.load_model('./1.qmodel')
text = qq.predict(cv2.imread('test.png'))
print(text)
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
qqocr-0.3.2.tar.gz
(4.0 kB
view hashes)
Built Distribution
qqocr-0.3.2-py3-none-any.whl
(4.4 kB
view hashes)