The repo contains the codebase for building and deployment of computer vision solution for digitalizing electric meters.
- Software
- Hardware
Run the data collection script data_collection_img.py. The script captures images every snapshot_interval.
The captured data is further processed to obtain each digit in each sample image using the following notebook.
The cropped images are uploaded the Edge Impulse and a computer vision model was built to recongize digits.
To use the inference script, provide the following details in the secret.py:
user = "" # MQTT username
password = "" # MQTT password
SSID = "" # Network SSID
KEY = "" # Network key
server = "" # MQTT server
Run the deployment script main.py. To perform inference.
Screen Scraping in Action
Demo video
- Build lighting system for obtaining reading at night
- Deep sleep and energy saving mode
- Alerting system for notification