Uniview Smart Intrusion Prevention – Human and Vehicle Detection
Introduction
Uniview is a video surveillance manufacturer which invests ~15%[1] of annual revenue in research and development. Over 50% of their employees are from the R&D department[2], resulting in higher quality and more advanced products every year.
Towards the end of 2020, Uniview refreshed their video surveillance lineup with a new series of NDAA compliant cameras. This camera refresh improved the image quality, added LightHunter technology, and also improved the intelligence of the cameras with Uniview Smart Intrusion Prevention.
What is Smart Intrusion Prevention?
Uniview Smart Intrusion Prevention is a series of AI products and solutions that adopts an independent AI chip and deep learning algorithm. It focuses on human and vehicle detection with high accuracy, enabling users to focus on defined targets. Smart Intrusion Prevention is built into certain Uniview cameras with the goal of providing more accurate alerts and notifications.
With intelligent mark enabled in the camera, you will have a colored box surrounding humans and vehicles. A green box will outline pedestrians, yellow box for non-motor vehicles, and orange box for motorized vehicles. When one of these objects triggers an event, you will see the box change to a red color.
Uniview’s Deep Learning Algorithm is able to distinguish between humans, motor vehicles, and non-motor vehicles.
Which Cameras Support Smart Intrusion Prevention?
Smart Intrusion Prevention is available on the Prime 4MP series cameras but is limited to human body detection only. It’s also available on 5MP/8MP camera series ending with -I0, providing advanced human/vehicle detection. If you are not sure that your camera supports this feature, you can email our tech support at: [email protected]
How to Enable Smart Intrusion Prevention on Uniview Cameras
Human and Vehicle detection can be enabled from the NVR or directly through the camera’s web interface.
- NVR Setup: Under the “Smart” tab, select a Uniview AI camera and enable “Perimeter Protection”. You can choose from Cross Line Detection, Intrusion Detection, Enter Area, or Leave Area.
- Camera Setup: Log into the camera from the web interface, and navigate to Setup à Intelligence à Smart. As with the NVR, you can choose from Cross Line Detection, Intrusion Detection, Enter Area, or Leave Area.
Configuration of Smart Intrusion Prevention
While configuring Smart Intrusion Prevention, you can filter between Motor Vehicle, Non-Motor Vehicle (bicycle/motorcycle), and Pedestrian. This can help you narrow down the triggering of the analytic based on the type of object present in the scene. You can also set the maximum and minimum size of the object in pixels, which can help reduce false positives in the scene. You should only configure this option if you are not getting the desired results on the default setting.
Conclusion
Uniview’s new AI chipset gives the camera an ability to differentiate between humans, vehicles, and non-important objects in the scene. It uses a deep learning algorithm to accomplish this feat. By enabling Smart Intrusion Prevention, you can focus the notifications or playback to only certain types or objects. This also reduces 99% of false positives from trees, clouds, rain, and other non-important types of motion viewed by the camera.
Because Uniview pours a lot of resources into R&D, we are confident that their deep learning algorithm will continue to improve and develop. Be sure to keep your Uniview NVR and cameras updated to the latest firmware to ensure best performance from Smart Intrusion Prevention. If you have any questions, be sure to visit our contact page to get in touch!
