Notes on how to wait for processing when using pyautogui for image recognition. Initially I used a primitive method of sleeping for a fixed period, but when the time for the target image to appear is unknown, it’s necessary to sleep with extra margin, resulting in a lot of wasted time overall.
time.sleep(15)
pos_check=pyautogui.locateOnScreen('path/to/target/image', confidence=0.6)
pyautogui.click(pos_check)
I revised it to loop and sleep until the target image appears. When the image is not found, an ImageNotFoundException is raised, which is caught to sleep and continue.
import pyscreeze
from pyscreeze import ImageNotFoundException
～omitted～
def click_from_imagename(imagename,confidence,message):
for tryCount in range(30):
# Processing when the target image exists on screen
try:
if pyautogui.locateOnScreen(imagename, confidence=confidence):
print('Execute: {0}'.format(message))
pos_check=pyautogui.locateOnScreen(imagename, confidence=confidence)
pyautogui.click(pos_check)
time.sleep(1)
print('Done: {0}'.format(message))
break
# Processing when the target image does not exist on screen
except pyautogui.ImageNotFoundException:
print('Waiting: {0}'.format(message))
time.sleep(1)
continue
～omitted～
When calling:
click_from_imagename("path/to/target/image", 0.6, "execution name")