Utilizing AI Features in QNAP QVR Pro
Surveillance camera systems may not be very familiar to the general public. However, nowadays, surveillance cameras are installed everywhere—schools, supermarkets, stations, buildings, parks, factories, farms, rivers, and more.
Even though we simply call them surveillance cameras, each has slightly different purposes. For example:
- Cameras installed at stations are not only for crime prevention but also for counting pedestrian traffic.
- In supermarkets, cameras are used not only to prevent shoplifting but also to monitor for any irregularities at the cash register.
- On farms, cameras help prevent vegetable theft and also check if wild animals are causing trouble.
- River cameras monitor suspicious individuals and, during heavy rain, watch water levels to prevent flooding.
And so on.
With QNAP’s QVR Pro, you can do more than just record video; you can use AI features to detect events according to your specific needs. By using AI, you can pick up and review only the events that match your purpose more precisely, which I would like to introduce here.
What Should Trigger Event Recording?
A relatively old trigger method is the so-called “motion trigger,” which records an event when there is movement in the image. With this method, for example, even if you want to record an event only when a person is present in an area where people are usually absent, the turning on/off of lights might be detected as “movement” and recorded as an event, or “trees swaying in the wind” might be recorded as an event, causing the actual “person present” events to get buried.
In such cases, when you want to pinpoint events, AI-based event detection becomes useful.
For example, when using motion detection as an event trigger, even the turning on/off of lights is detected, as shown here:
With this, even though you only want to check when a person passes by, the on/off of lights is also recorded as an event, making it more time-consuming to review later.
Here is an example when using AI for human detection:
The turning on/off of lights is not considered an event, and only when a person passes by is it recorded as an event.
With recordings like this, it becomes much easier to review surveillance footage later.
Setting Up AI-Based Detection Features
In QVR Pro and QVP series, you can configure this in the “Event Rules” section of the settings screen.
When adding an event, you can select types such as “Human Detection,”
Vehicle Detection,
and more, allowing for highly accurate detection using AI.
Note: For AI-based detection, the camera must support this feature.
Summary
This time, we introduced event detection in surveillance camera systems using the QVR Pro/QVP series.
While motion detection alone “records all changes within the screen as events,” by using human detection features, “you can focus on recording only what you actually want to see or keep, such as when a person passes by.”
When building a surveillance camera system, it is a good idea to consider whether AI-based detection is necessary or possible.
Demo Information
You can see demos of AI-based event detection and recording with QVR Pro and QVP, as introduced here, at the QNAP booth at Security Show 2025 @ Tokyo in March 2025.
QNAP staff will be present at the booth on the day, so you can ask questions or consult with them.
We look forward to your visit!