Font Size: A A A

CCTV

AI cameras

New from Hanwha Techwin Europe are two AI-powered dual-channel multi-sensor cameras, the PNM-C12083RVD and PNM-C7083RVD. The AI cameras are for monitoring two areas with two different fields of view, such as for monitoring staircases or elevators, looking around corners, and securing two adjacent areas such as check-in desks.

The two-in-one cameras feature deep-learning based object detection and classification while representing a lower total cost of ownership (TCO) over installing two equivalent cameras, the product company says. Besides lower hardware costs, they use one cable and in most cases require only a single VMS (video management softwrae) licence. The PNM-C12083RVD comes with two 6MP cameras at 15 frames per second (fps), while the PNM-C7083RVD has two 2MP at 30fps. Both come with an infra-red viewable length of 25m, are NDAA compliant for the United States, and feature cybersecurity with TPM 2.0 and are FIPS 140-2 certified.

AI-based object detection combats false alarms, while object classification can aid operational efficiency during forensic search. The cameras can identify human faces as well as identify particular vehicles and recognise number plates, while ignoring irrelevant motion such as wind-blown trees, shadows, stray objects, or animals moving within the field of view.

Review events

For post-event analysis, the dual-channel cameras use BestShot technology to support operators in seeing the best shot of an event occurring. Camera-generated metadata (person, face, vehicle, and licence plate), is embedded in the video and delivered back to a control room, so a compatible VMS can carry out a forensic search.

AI also works on the image quality of video. WiseNR II noise reduction uses AI to identify object movement and reduce blur in noisy, low-light uses. AI-based Preferred Shutter technology automatically adjusts shutter speed based on classified objects in motion and the lighting, to reduce motion blur and deliver the clearest images possible.

WiseStream III dynamically controls encoding so that when a classified object is detected, the compression rate on the area is minimised to improve image quality, while the rest of the scene is compressed at a higher rate. Combined with H.265 compression the bandwidth efficiency can be improved by up to 75 per cent it is claimed compared to H.264 compression.

The cameras feature a metal-shielded RJ-45 to protect the cameras from transient voltage, a hard-coated dome bubble providing better resistance to scratching and a GORE vent to prevent condensation and maintain performance. As for installation, CVBS video output and a Wi-Fi dongle (optional) are for viewing and adjusting cameras from a smartphone.

Full product information can be found at https://bit.ly/3HEuaDg.


Tags

Related News