更快的图像识别？

如何解决更快的图像识别？

Castimage = pag.locateCenterOnScreen('Cast.png',region=(426,579,643,676))
PullImage = pag.locateCenterOnScreen('pull.png',676)) 
stopper = pag.locateCenterOnScreen('stopper.png')
ensure = pag.locateCenterOnScreen('ensure.png',region=(154,88,624,132))

我希望这部分代码更快，因为在屏幕上搜索照片会消耗太多时间（我什至指定了要搜索此照片的区域，但没有帮助）。也许我应该使用其他图书馆？如何更快地在屏幕上定位图像？

这是我的完整代码：

import pyautogui as pag
import time


def cast():
    flag = True
    while flag:
        Castimage = pag.locateCenterOnScreen('Cast.png',676))
        print('searching for cast image')
        if Castimage is not None:
            time.sleep(0.2)
            click = pag.click(Castimage)
            print('clicking cast image')
            flag = False


def pull():
    flag = True
    while flag:
        PullImage = pag.locateCenterOnScreen('pull.png',676))
        print('searching for pull image')
        if PullImage is not None:
            time.sleep(0.2)
            click1 = pag.click(PullImage)
            print('clicking pull image')
            flag = False


def strength():
    flag = True
    stopper = pag.locateCenterOnScreen('stopper.png')
    ensure = pag.locateCenterOnScreen('ensure.png',132)) # ensures that this is right window(with game)
    while flag:
        print('Searching for strengthBAR')
        strengthgreen = pag.locateOnScreen('strength.png',region=(355,304,774,528))
        if strengthgreen is None and ensure is not None:
            clickPull = pag.mouseDown(534,634) # FIXED VALUE
            print('PULLING')
        if stopper is not None:
            flag = False
            print('Stopping strength function')


while True:
    cast()
    pull()
    strength()