Hikvision is launching a new level of analogue surveillance by introducing Deep Learning. A technology that delivers AI via the Acusense Turbo DVR, marketed as a world first, and also provides vivid images 24 hours a day, via Colorvu Turbo cameras.
Acusense technology is based on Hikvision’s Deep Learning algorithm, which first started to appear in IP products last year. Deep Learning algorithms are much deeper programming compared to conventional intelligent algorithms, which according to Hikvision only operate on the surface level. These perform feature learning, designed to provide astonishingly accurate and consistent video content analytics (VCA) performance.
Acusense is able to extract human body and vehicle information from a massive amount of objects. Not only can it detect human and vehicle intrusion, but it can also filter out false alarms that would have been triggered by non-threatening moving objects.
The technology also brings a powerful and quick target search function that is capable of analysing recordings and picking out all human or vehicle images which link to the related video clip. This means that users are able to quickly search for a target, for example, as the system is able to extract human and vehicle information from the feed. These exciting features make the DVR more powerful and reliable than ever before.
While the Acusense provides the brain of the system – its eyes are provided by the Colorvu camera, with an advanced F1.0 large aperture lens, giving more light, brighter images and balanced exposure. With the addition of high sensitivity in ultra low light environments, the camera can provide vivid, full-colour video 24/7 in almost 90% of scenarios. What’s more, the cameras also have warm supplementary lighting and an anti-glare diffusion lens to further guarantee the vivid picture in very dark areas.
These technologies will prove useful for installers who will be able to bring many of the smart features to an existing system by simply replacing an older DVR with the Acusense model. Users will find it easy to use, saving resources in both manpower and finances.
“We have been proving the value of Deep Learning in IP solutions for a while now – the next step is to bring this innovative technology to analogue users,” says Peter Guan, Channel and Marketing Director at Hikvision Europe. “This is the first time the surveillance industry has seen this quality of vivid colour in video images throughout the whole day. We think this will bring a lot of change to the analogue market.”
