Now that we're confident that we can at least get the screen data, the next thing we want to do is run some OpenCV operations on the data that are typical for vision tasks.

For example, we'll want to conver the image to grayscale to simplify things (one value vs RGB's 3 values per pixels) and edge detection to eventually be used for finding the lines that will be our lanes.

The code up to this point:

import numpy as np
from PIL import ImageGrab
import cv2
import time

# just so this doesn't go on forever:

def screen_record(): 
    last_time = time.time()
    while(True):
        # 800x600 windowed mode
        printscreen =  np.array(ImageGrab.grab(bbox=(0,40,800,640)))
        print('loop took {} seconds'.format(time.time()-last_time))
        last_time = time.time()
        cv2.imshow('window',cv2.cvtColor(printscreen, cv2.COLOR_BGR2RGB))
        if cv2.waitKey(25) & 0xFF == ord('q'):
            cv2.destroyAllWindows()
            break
            
#screen_record()

Now let's add some grayscale and edges:

import numpy as np
from PIL import ImageGrab
import cv2
import time


def process_img(image):
    original_image = image
    # convert to gray
    processed_img = cv2.cvtColor(image, cv2.COLOR_BGR2GRAY)
    # edge detection
    processed_img =  cv2.Canny(processed_img, threshold1 = 200, threshold2=300)
    return processed_img

def main():
    last_time = time.time()
    while True:
        screen =  np.array(ImageGrab.grab(bbox=(0,40,800,640)))
        #print('Frame took {} seconds'.format(time.time()-last_time))
        last_time = time.time()
        new_screen = process_img(screen)
        cv2.imshow('window', new_screen)
        #cv2.imshow('window',cv2.cvtColor(screen, cv2.COLOR_BGR2RGB))
        if cv2.waitKey(25) & 0xFF == ord('q'):
            cv2.destroyAllWindows()
            break

#main()
from IPython.display import Image
Image(filename='edge-detection.png') 

Looking great!

In my excitement of getting the screen recording working, I haven't yet tested PyAutoGUI, so we can do that next.