Python实现答题辅助工具：OCR识别与自动搜索

from screenshot import pull_screenshot
import time, urllib.request
 
try:
 import Image
except ImportError:
 from PIL import Image, ImageDraw
 
import pytesseract
 
# 屏幕顶端到问题的距离/屏幕高度，随分辨率变化（默认1920*1080）
top_off_c = 0.15
# 问题高度
que_h = 300
# 答案高度
ans_h = 170
 
# 左右偏移量
l_r_off = 40
 
 
def check(question, ans):
 q_url = "http://www.baidu.com/s?word=" + urllib.parse.quote(question)
 resp = urllib.request.urlopen(q_url)
 
 print(resp.read().decode("utf-8"))
 
 
def draw():
 img = Image.open('autojump.png')
 w, h = img.size
 draw = ImageDraw.Draw(img)
 draw.line((40, h * 0.15, w - 40, h * 0.15), fill="red")
 draw.line((40, h * 0.15 + 300, w - 40, h * 0.15 + 300), fill="red")
 
 draw.line((40, h * 0.15 + 470, w * 0.7, h * 0.15 + 470), fill="red")
 draw.line((40, h * 0.15 + 640, w * 0.7, h * 0.15 + 640), fill="red")
 draw.line((40, h * 0.15 + 810, w * 0.7, h * 0.15 + 810), fill="red")
 
 img.show()
 
 
def main():
 while True:
  print(">>>>>>")
  pull_screenshot()
  img = Image.open('autojump.png')
  img = img.convert('L')
  w, h = img.size
  img_q = img.crop((l_r_off, h * top_off_c, w - l_r_off, h * top_off_c + que_h))
  img_a = img.crop((l_r_off, h * top_off_c + que_h, w * 0.7, h * top_off_c + que_h + ans_h))
  img_b = img.crop((l_r_off, h * top_off_c + que_h + ans_h, w * 0.7, h * top_off_c + que_h + ans_h * 2))
  img_c = img.crop((l_r_off, h * top_off_c + que_h + ans_h * 2, w * 0.7, h * top_off_c + que_h + ans_h * 3))
  question = pytesseract.image_to_string(img_q, lang='chi_sim')
  ans_a = pytesseract.image_to_string(img_a, lang='chi_sim')
  ans_b = pytesseract.image_to_string(img_b, lang='chi_sim')
  ans_c = pytesseract.image_to_string(img_c, lang='chi_sim')
  question = question.replace(" ", "").replace(".", "")
  ans = ["1", "1", "1"]
  ans[0] = ans_a.replace(" ", "").replace(".", "")
  ans[1] = ans_b.replace(" ", "").replace(".", "")
  ans[2] = ans_c.replace(" ", "").replace(".", "")
 
  print(question.replace(" ", "").replace(".", ""))
  print(ans)
 
  check(question, ans)
  # draw()
 
  time.sleep(1)
 
 
if __name__ == '__main__':
 main()

sudo pip3 install pytesseract
sudo apt-get install tesseract-ocr