Welcome to HomeAIVision! I'm here to make your home safer and smarter by bringing advanced AI vision capabilities right into your Home Assistant setup. Utilizing Azure Cognitive Services, I can analyze your home camera feed in real-time to detect human presence, ensuring that you only receive notifications that matter.
Before installing HomeAIVision, ensure you have:
- Home Assistant Installation: Running locally, on a device like a Raspberry Pi, or a VM. More details can be found here.
- Camera Setup: Accessible via a URL, testable in a web browser. Ensure your camera stream is accessible from the network Home Assistant is on.
- Azure Account: Active account with an API key and endpoint URL from Azure Cognitive Services. Get your API key here.
-
Download the Integration:
- Visit the GitHub repository.
- Click on the 'Code' button and select 'Download ZIP'. Extract the file.
-
Install the Custom Component:
- Ensure a
custom_components
directory exists in your Home Assistant configuration directory (/config
). - Copy the
homeaivision
folder from the extracted ZIP into thecustom_components
directory.
- Ensure a
-
Restart Home Assistant:
- This step is necessary to recognize the new integration. Restart via the server management page or restart the Home Assistant service.
-
Add the Integration:
- Navigate to Configuration > Integrations in Home Assistant.
- Click on Add Integration and search for HomeAIVision.
- Fill in your Azure API key, endpoint URL, and camera URL as prompted. These can be obtained from your Azure Cognitive Services account.
Parameter | Description | Default |
---|---|---|
azure_api_key |
Azure Cognitive Services API key. | |
azure_endpoint |
Endpoint URL for Azure Cognitive Services. | |
cam_url |
URL to access the camera feed. | |
time_between_requests |
Seconds between image requests. | 30 |
send_notifications |
Enable or disable notifications. | False |
organize_by_day |
Organize saved images by day. | True |
detected_objects |
Select which objects to detect (e.g., person, car). | person |
confidence_threshold |
Minimum confidence threshold for detections. | 0.6 |
Parameter | Description | Default |
---|---|---|
max_images |
Max images per folder or in total. | 30 |
days_to_keep |
Days to keep images if organized by day. | 7 |
notification_language |
Notification language. | English |
- Human Detection: Utilizes Azure Cognitive Services to identify human presence in the camera feed.
- Notifications: Alerts are sent to your devices upon detection to ensure you're always informed.
- Image Saving: Captures and saves images where human activity is detected, organizing them based on your settings.
- Verify Azure API and endpoint accuracy: Ensure that your Azure credentials are correctly entered.
- Check Camera Accessibility: The camera URL must be accessible from your Home Assistant installation.
- Review Logs: Check the Home Assistant logs for errors related to this integration, which can provide clues for troubleshooting.
If you like HomeAIVision and find it useful, consider giving it a ⭐ on GitHub! Your support encourages further development and helps others discover the project.
Encounter issues or have suggestions? Please report them on our Issues page. Your feedback is invaluable and helps improve HomeAIVision!
This project is licensed under the MIT License - see the LICENSE file for details.