用python搭建"百万赢家"自动搜索答案
用python搭建百万答题、自动百度搜索答案。
使用平台
windows7
python3.6
MIX2手机
代码原理
手机屏幕内容同步到pc端
对问题截图
对截图文字分析
用浏览器自动搜索文本
使用教程
1、使用Airdroid 将手机屏幕显示在电脑屏幕上。
也可使用360手机助手实现。不涉及任何代码。实现效果如图:
2、在提问出现时,运行python程序,将问题部分截图。
这里要用到两个函数:
get_point() #采集要截图的坐标,以及图片的高度宽度
window_capture() #截图
def get_point(): '''采集坐标,并返回w,h,x,y。 作为window_capture() 函数使用''' try: print('正在采集坐标1,请将鼠标移动到该点') # print(3) # time.sleep(1) print(2) time.sleep(1) print(1) time.sleep(1) x1,y1 = pag.position() #返回鼠标的坐标 print('采集成功,坐标为:',(x1,y1)) print('') # time.sleep(2) print('正在采集坐标2,请将鼠标移动到该点') print(3) time.sleep(1) print(2) time.sleep(1) print(1) time.sleep(1) x2, y2 = pag.position() # 返回鼠标的坐标 print('采集成功,坐标为:',(x2,y2)) #os.system('cls')#清除屏幕 w = abs(x1 - x2) h = abs(y1 - y2) x = min(x1, x2) y = min(y1, y2) return (w,h,x,y) except KeyboardInterrupt: print('获取失败')
def window_capture(result,filename): '''获取截图''' #宽度w #高度h #左上角截图的坐标x,y w,h,x,y=result hwnd = 0 hwndDC = win32gui.GetWindowDC(hwnd) mfcDC = win32ui.CreateDCFromHandle(hwndDC) saveDC = mfcDC.CreateCompatibleDC() saveBitMap = win32ui.CreateBitmap() MoniterDev = win32api.EnumDisplayMonitors(None,None) #w = MoniterDev[0][2][2] # #h = MoniterDev[0][2][3] # w = 516 # h = 514 saveBitMap.CreateCompatibleBitmap(mfcDC,w,h) saveDC.SelectObject(saveBitMap) saveDC.BitBlt((0,0),(w,h),mfcDC,(x,y),win32con.SRCCOPY) saveBitMap.SaveBitmapFile(saveDC,filename)
运行后截图如下
3.对图片文字分析提取
参考链接: * 图片转文本 * 配置方式
代码部分:
def orc_pic(): #识别中文 text=pytesseract.image_to_string(Image.open('jietu.jpg'),lang='chi_sim') #识别英文 # text=pytesseract.image_to_string(Image.open('jietu.jpg')) text = ''.join(text.split()) return text
4.对文本进行搜索
#浏览器搜索 url = 'http://www.baidu.com/s?wd=%s' % text webbrowser.open(url)
所有代码如下:
#coding:'utf-8' import win32gui, win32ui, win32con, win32api from PIL import Image import pytesseract import webbrowser #先下载pyautogui库,pip install pyautogui import os,time import pyautogui as pag #获取sdk http://ai.baidu.com/。 #获取aip pip install git+https://github.com/Baidu-AIP/python-sdk.git@master from aip import AipOcr import json status=0 """ 你的 APPID AK SK """ APP_ID = '****' API_KEY = '***' SECRET_KEY = '***' client = AipOcr(APP_ID, API_KEY, SECRET_KEY) """ 读取图片 """ def get_question(path): '''百度识别图片文字''' with open(path, 'rb') as fp: image=fp.read() res = client.basicGeneral(image) words = res['words_result'] lines = [item['words'] for item in words] question = ''.join(lines) if question[1] == '.': question = question[2:] elif question[2] == '.': question = question[3:] return question.replace('?', ' ') #采集坐标 def get_point(): '''采集坐标,并返回w,h,x,y。 作为window_capture() 函数使用''' try: print('正在采集坐标1,请将鼠标移动到该点') # print(3) # time.sleep(1) print(2) time.sleep(1) print(1) time.sleep(1) x1,y1 = pag.position() #返回鼠标的坐标 print('采集成功,坐标为:',(x1,y1)) print('') # time.sleep(2) print('正在采集坐标2,请将鼠标移动到该点') print(3) time.sleep(1) print(2) time.sleep(1) print(1) time.sleep(1) x2, y2 = pag.position() # 返回鼠标的坐标 print('采集成功,坐标为:',(x2,y2)) #os.system('cls')#清除屏幕 w = abs(x1 - x2) h = abs(y1 - y2) x = min(x1, x2) y = min(y1, y2) return (w,h,x,y) except KeyboardInterrupt: print('获取失败') #获取截图 def window_capture(result,filename): '''获取截图''' #宽度w #高度h #左上角截图的坐标x,y w,h,x,y=result hwnd = 0 hwndDC = win32gui.GetWindowDC(hwnd) mfcDC = win32ui.CreateDCFromHandle(hwndDC) saveDC = mfcDC.CreateCompatibleDC() saveBitMap = win32ui.CreateBitmap() MoniterDev = win32api.EnumDisplayMonitors(None,None) #w = MoniterDev[0][2][2] # #h = MoniterDev[0][2][3] # w = 516 # h = 514 saveBitMap.CreateCompatibleBitmap(mfcDC,w,h) saveDC.SelectObject(saveBitMap) saveDC.BitBlt((0,0),(w,h),mfcDC,(x,y),win32con.SRCCOPY) saveBitMap.SaveBitmapFile(saveDC,filename) def get_point_txt(status): #如果status=y,则重新获取坐标 '''如果存在point.txt,则询问是否重新采集,删除point.txt;如果不存在txt,则直接采集。''' if not os.path.isfile('point.txt') : result = get_point() with open('point.txt', 'w') as f: f.write(str(result)) return result else: if status=='y': result = get_point() with open('point.txt', 'w') as f: f.write(str(result)) return result else: with open('point.txt', 'r') as f: result = f.readline() result = eval(result) return result def orc_pic(): #识别中文 text=pytesseract.image_to_string(Image.open('jietu.jpg'),lang='chi_sim') #识别英文 # text=pytesseract.image_to_string(Image.open('jietu.jpg')) text = ''.join(text.split()) return text #百度识别 def orc_baidu(): text=get_question('jietu.jpg') return text status='y' start = time.time() result=get_point_txt(status) for i in range(10): window_capture(result,'jietu.jpg') # text=orc_baidu() text=orc_pic() print(text) #浏览器搜索 url = 'http://www.baidu.com/s?wd=%s' % text webbrowser.open(url) # url2='https://www.google.com/search?q=%s' % text # webbrowser.open(url2) end = time.time() time=end-start print('此次耗时%.1f秒' % time)
来源:PY学习网:原文地址:https://www.py.cn/article.html