Though available for many years, Auto-Tracking PTZs are generating renewed interest with Axis' recent release of their 233D camera. The claim is that "auto-tracking enables a single PTZ dome camera to cover a large area and to automatically detect, follow and record video of moving persons or vehicles."
Auto-Tracking has never become an widely adopted feature despite its introduction and marketing by various manufacturers. The question then becomes: should you use it and where should you use it?
Below view a video of Axis demonstrating auto-tracking PTZ functionality or download the clip. After, we will consider its value.
The demo provides a very honest portrayal of the key issues involved in setting up and using auto-tracking PTZs. Essentially, the tracking is designed to track the object with the most movement or greatest size. There are some major drawbacks including:
- The feature can track irrelevant objects like flags or other uninteresting moving objects.
- The feature can switch tracking from a relevant to an irrelevant object simply because the second object was bigger.
- These problems increase as the size of the area and the amount of activity increases. For instance, this would likely work extremely poorly in a football stadium parking lot during game day as hundreds of cars and people move around.
As such, it seems this feature is likely most useful in small areas with light traffic. This is where I have used this feature with other camera manufacturer's in the past. Often, we wind up turning it off because the camera made mistakes and the security manager found setting a wide field of view preset or a continuous pan to be the least confusing setup for the PTZ.
Today, I would prefer to use a 2MP - 5MP camera for the type of scenario pictured in this demo. Rather than risk the problems of incorrectly tracking, the megapixel camera guarantees evidence across the entire area. Operators can then use digital zooms either in real time or for investigations to get exactly the image they need.
Even on a cost basis, a megapixel camera plus the addiitonal storage is no more expensive than a PTZ. For instance, the 233D costs about $3,000 USD while a megapixel camera costs around $1,000 with an extra $2,000 or so for more storage.
I see this as part of the general trend why megapixel is gaining rapid acceptance and positive satisfaction while video analytics are still dealing with accuracy and optimization issues. While megapixel does not offer as much as video analytics, what it does it does well. The reduced risk and increased simplicity spurs uptake. I see this as the better choice to auto-tracking PTZs. Thoughts?
