README.md

object-detection

Detect specific objects in images, video feed and games using OpenCV (Open Source Computer Vision Library) & Python 3. Here's a list of the current scripts. Some are incomplete or for specific platforms.

1. Images can be JPEG, JPG, PNG or other commonly used image file types.

Add our images file path (Tested on Windows & Linux)

my_image = cv2.imread('./images/harry-meghan.jpg')

Run the image script

$ py image_object_detection.py

2. Video feed can come trough USB or camera modules.

Our default video camera for capture (Tested on Winndows, Mac & Linux)

video_capture = cv2.VideoCapture(0)

Pi Camera module (Tested on Raspberry Pi 3 B+)

# PiCam settings
camera = PiCamera()
camera.resolution = (width, height)
camera.framerate = 30

# Video capture
video_capture = PiRGBArray(camera, size=(width, height))

Run the picam script

$ python video_object_detection_with_picam.py

3. Game capture is taken from a specific position on the primary screen.

Our game screen to capture (Tested on Winndows)

game_screen = np.array(ImageGrab.grab(bbox=(0, 0, 800, 600)))

Run the game script

$ py game_object_detection.py