You may have heard about people counting sensors or people counter depth cameras. In this post, we will discuss and evaluate pros and cons of people counting devices vs camera-based people counters like IP camera or CCTV camera.
1. Physical Installation: Camera-Based People Counter Solutions win
If you choose to use camera-based people counter, you start in minutes using your existing IP cameras. You can configure them simply and start getting results and analytics from your video sources.
But if you choose to use people counting devices like door sensors or depth cameras, you will have an installation process which is strictly defined by the hardware vendor. Also please consider that these hardwares mostly guarantees a confidence level only valid for certain physical conditions. So, unfortunately, you may be out-of-coverage easily.
2. Effective Area: People counting device solutions win
Quick refreshes here: a camera-based video analytics solution like VidMetrica can analyze only the camera stream data it receives. So, if you need hundreds of m2 of the area to perform video analytics, you will need enough cameras to cover all of these areas.
Sensor or depth camera-based solutions often run only on entrances and exit doors of an area. In this case, it would be enough to place just one piece of hardware for each entrance even for hundreds of m2 of area. Consider your requirements and may be counting from doors is just enough for your use-case. But you need to give up on very rich in-field analytics like social distancing and homogenization metrics which we’ll discuss in the next section.
3. In-field Metrics: Camera-Based People Counter solutions win
As you can guess, VidMetrica produces a rich set of in-field analytics which a people counting device can’t. Social distancing violations, social distancing safety, people homogenization, people per m2, people distribution metrics are just initial ones that can be obtained out-of-the-box. VidMetrica can even produce 2D representation of the area on dashboards in real-time. Please check out our IGA Istanbul Airport case study page to see details.
4. Stateful vs stateless data, Fault-tolerance: Camera-Based People Counter solutions win
Camera-based people counters count video sources in every processing. Thus, you will have refreshed data after every processing. This allows the system to adopt video processing frequency. It can be possible to run analyses just once in a minute on cameras to optimize computing power. Even if the video data is processed just once, it will immediately have up-to-date analytics from real world.
For people-counting devices, the device will produce each in and out records. But in any case of hardware failure, power outage, or network disconnection, hardware will miss some in or out measurements. If 3 of the outs are missed while an outage, it’s not possible to get this missing data to the device. Then, this error will likely be moved until a reset applied intentionally.
Final Words
Regardless of your choice, we’re here to help you to explore best people counting analytics solution works for you. You can contact us and get detailed guidance about your use-case. We would love to help on not only people counting solutions but also your all other video based analytics needs.
