Required directories under the source code:-

new_images archives

Required libraries import

import os import shutil from datetime import datetime from watchdog.observers import Observer from watchdog.events import FileSystemEventHandler import cv2 from tkinter import Tk, Button, Label, Frame, StringVar, LEFT from PIL import Image, ImageTk from ultralytics import YOLO

Operation

Run the code with python object_detect.py Drag files into the new_images directory The system will detect the principle obect in the view, then crop the image to a 500*500 square. A discripton and confidence score is added.

Next enhancement is to read the new_images directory and move the cropped images to the directory that is waiting for latest.jpg for meter reading.