Fine-Tuning Avigilon Analytics for Accurate Pet Detection
Avigilon's powerful video analytics are designed to intelligently classify objects, distinguishing between people, vehicles, and other movement. However, achieving reliable pet detection can sometimes be a challenge. You might find your system is generating false alarms for your dog roaming the garden, or worse, completely missing smaller animals.
This professional guide will help you understand how Avigilon analytics work and how you can configure your system to accurately detect pets while minimising unwanted alerts.
### How Avigilon Classifies Objects
Avigilon cameras with self-learning analytics don't just detect motion; they analyse the characteristics of moving objects, including their size, speed, shape, and behaviour. The system is pre-trained on a massive dataset to recognise humans and vehicles. While "animal" is a classification, it can be more nuanced than the others. The accuracy depends heavily on the quality of the video feed and the configuration of the analytic rules.
Key Steps to Improve Pet Detection Accuracy
Optimising pet detection is a process of refinement. Work through these steps to dial in your system's performance.
1. Camera Placement is Critical
The foundation of good analytics is a good view. No amount of software tuning can fix a poorly positioned camera.
- Angle and Height: The camera should be mounted high enough to have a clear, angled view of the detection area. A top-down view can make it difficult for the analytics to determine an object's shape. Follow Avigilon's recommended mounting heights for your specific camera model.
- Clear Line of Sight: Ensure there are no obstructions like tree branches, bushes, or garden furniture that could partially obscure an animal as it moves through the scene.
- Lighting: Good, even lighting is essential. Strong backlighting (e.g., pointing at the rising or setting sun) can create silhouettes that are hard to analyse. Ensure the area is adequately lit, especially at night, for IR cameras to be effective.
2. Configure Analytic Rules in ACC
The Avigilon Control Center (ACC) software is where you will do most of the fine-tuning.
- Define a Precise Region of Interest (ROI): Don't just draw a box around your entire garden. Create a more specific ROI that covers the ground where pets are likely to be, avoiding areas with irrelevant motion like public footpaths or rustling trees.
- Adjust Object Size Parameters: In the rule setup, you can define the minimum and maximum size of an object to be detected. This is a powerful tool. Measure your pet and set the size parameters to create a "detection window" that is more likely to include your pet and exclude smaller things (like birds) or larger things (like people).
- Tweak Sensitivity and Threshold: The sensitivity settings control how much an object must change from the background to be detected. Start with the default and adjust it incrementally. A setting that is too high will lead to false alarms from shadows and rain.
3. Use "Classified Object" as the Trigger
When setting up your rule, ensure the event trigger is set to "Classified Object" and that "Animal" is a selected class. Avoid using the generic "Motion Detection" trigger, as this will alert on any pixel change and will not use the advanced analytics engine.
4. Keep Your System Updated
Avigilon continuously improves its analytics algorithms. Performance can be significantly enhanced by running the latest versions of both the camera firmware and the ACC server software.
- Check for Firmware: Regularly check the Avigilon website for firmware updates for your specific camera models.
- Update ACC Software: Ensure your ACC server is on the latest release to take advantage of the most recent analytic capabilities and bug fixes.
5. Be Patient and Refine
Video analytics, especially for nuanced subjects like animals of various sizes, require a period of observation and adjustment. After making changes, monitor the system's performance for a day or two. Review any missed detections or false positives and use that information to make further small adjustments to your rules. By taking a methodical approach to camera placement and software configuration, you can significantly improve the reliability of your Avigilon system's pet detection.