Programming

Don't know about PGE's API, but for the OCR stuff, you may get better results with additional preprocessing before you pass images into tesseract. e.g. crop it to just the region of interest, and try various image processing techniques to make the text pop out better if needed. You can also run tesseract specifically telling it you're looking at a single line of data, which can give better results. (e.g. --psm 7 for the command line tool) OCR is indeed finicky...

I suspect that PG&E's smart meters might: 1) support an infrared pulse through an LED on the top of the meter, and 2) use a fairly-open protocol for uploading their meter data to the utility, which can be picked up using a Software Defined Radio (SDR).

Open Energy Monitor has a write-up about using the pulse output, where each pulse means a quantity of energy was delivered (eg 1 Watt-hour). So counting 1000 of such pulses would be 1 kWh, and that would be a way to track your energy consumption for any timescale.

What it won't do is provide instantaneous power (ie kW drawn at this very moment) because the energy must accumulate to the threshold before sending a pulse. For example, a 9 Watt LED bulb that is powered on would only cause a new pulse every 6.7 minutes. But for larger loads, the indication would be very quick; a 5000 W dryer would emit a new pulse after no more than 0.72 seconds.

The other option is decoding the wireless protocol, which people have done using FOSS software. An RTL-SDR receiver is not very expensive, is very popular, and can also be used for other purposes besides monitoring the electric meter. Insofar as USA law is concerned, unencrypted transmissions are fair game to receive and decode. This method also has a wealth of other useful info in the data stream, such as instantaneous wattage in addition to the counter registers.

import cv2
import easyocr
import numpy as np
from PIL import Image
from collections import Counter

# Initialize the EasyOCR reader
reader = easyocr.Reader(['en'])

def preprocess_image(image):
    # Convert to PIL image for EasyOCR processing
    return Image.fromarray(cv2.cvtColor(image, cv2.COLOR_BGR2RGB))

def recognize_text(image):
    processed_image = preprocess_image(image)
    results = reader.readtext(np.array(processed_image), allowlist='0123456789')
    # Concatenate all recognized text results
    recognized_text = ''.join(result[1] for result in results)
    return recognized_text

def format_number(text, length=6):
    # Remove non-numeric characters and pad with zeros if necessary
    formatted = ''.join(filter(str.isdigit, text))
    return formatted.zfill(length)[-length:]

def most_common_number(numbers):
    # Find the most common number from the list of numbers
    counter = Counter(numbers)
    most_common = counter.most_common(1)
    return most_common[0][0] if most_common else ''

def main():
    cap = cv2.VideoCapture(2)

    if not cap.isOpened():
        print("Error: Could not open webcam.")
        return

    print("Press 'q' to quit.")
    frame_count = 0
    text_history = []

    while True:
        ret, frame = cap.read()
        if not ret:
            print("Error: Failed to capture image.")
            break

        # Recognize text from the current frame
        recognized_text = recognize_text(frame)
        formatted_number = format_number(recognized_text)

        # Update the history with the latest recognized number
        text_history.append(formatted_number)

        # Keep only the last 10 frames
        if len(text_history) > 20:
            text_history.pop(0)

        # Determine the most common number from the history
        most_common = most_common_number(text_history)
        print(f"Most common number from last 10 frames: {most_common}")

        # if cv2.waitKey(1) & 0xFF == ord('q'):
        #     break

    cap.release()

if __name__ == "__main__":
    main()