AI Plus NVR Is a Powerful Surveillance Combination

An existing NVR network could be retrofit to add the AI capabilities, assuming the network and vision systems are adequate.

Network video recorders (NVRs) have been around for decades, and they serve many useful purposes. They are often used for video surveillance applications, and are connected to networked cameras. The number of cameras that can be connected is based on many variables, some having to do with the users’ needs and some having to do with the capabilities of the network hardware and software.

Typical features of an NVR include the ability support real-time video monitoring, recording, playback, and backup. Assuming it’s supported by the software, the NVR can retrieve video at certain events or time intervals. In more modern systems, video compression is used, which significantly reduces the amount of data that needs to be stored.

Enhancing Safety and Security

NVRs can help ensure both the safety and security of people and equipment/facilities. An attendant can be alerted if an emergency arises, like if a person is injured or a piece of equipment is not operating as it is intended. It can also call attention to a security situation, such as if an unauthorized entry occurs.

The NVRs can combine their vision systems with other sensors to further enhance the safety and security. For example, temperature sensors can detect when heat is rising and a piece of equipment should be shut down. Or a motion detector can be placed in an area that does not contain a camera. That could also potentially enable the camera so the higher power peripherals are not running 24/7.

In some environments, the NVR could help facilitate better traffic flow, whether that is people, cars, or products coming off an assembly line. Knowing which areas and at what times the traffic is highest can help with staffing issues as well.

Adding the AI Layer

The software behind the NVRs continues to evolve, but there is a point at which another technology is required to go to the next level. That technology is artificial intelligence (AI), and thankfully, it has arrived. The AI NVR is typically connected to the Cloud or an internal intranet. It uses video analytics to resolve problems far more quickly and, in some cases, even before they occur.

The Cloud connection ensures that administrators can access the video at any time or place and depending on the network configuration, take necessary actions without having to be on site. But it’s the AI capabilities that really raise the smart NVR to another level.

The AI-enhanced video analytics can analyze the video content in real-time, and extract metadata to be used and/or stored. If appropriate, alerts are transmitted and actionable intelligence is sent to administrators, security personnel, or other systems for further analysis or action. Depending on the configuration, the analytics can occur at the Edge, in the Cloud, or both.

The addition of AI allows for different scenes to analyzed, noting the changes that are occurring (or have occurred). Those scenes are compared to previous analytics to know what has changed, why it has changed, and what action (if any) should ensue.

The list of end applications for AI-based smart NVRs continues to grow. It currently includes anything from simple motion detection to people and/or object detection. The latter could be people in a hospital or other secured area, or objects such as license plates or manufactured goods.

Applications That Can Maximize the AI NVR

In a retail setting, the video analytics solution could provide proactive monitoring, meaning that by recognizing attributes of people, the most appropriates ads could be presented to them. In a transportation environment, the video analytics technology could be combined with traffic safety analysis information to provide higher accuracy traffic data and help maintain safer roads. And in a healthcare space, the smart NVR with AI technology would provide protection for both patients and staff.

An existing NVR network could be retrofit to add the AI capabilities, assuming the network and vision systems are adequate. That means the network must be able to handle the required bandwidth and the cameras must provide the desired resolution.

MiTAC’s AI-enhanced smart NVRs harness the power of the Intel 11th Gen Core processor to add AI analytics to video server applications.

MiTAC’s N series (NV1) of AI-enhanced smart NVRs can handle the described applications, with headroom to spare. The NV1 is housed in a multi-functional, compact three-in-one 1U case, combining the NVR with a PoE switch, and the AI server. The NV1 is powered by the latest Intel 11th Gen Core Processor (Tiger Lake-UP3) which is well suited for AI applications due to its embedded AI engine; it varies from 3.7 to 7.1 TOPS based on the processor, for up to 10 channels of AI video analytics. The processor can also handle up to 64-channel H.264/H.265 video decode and storage using hardware RAID hard-disk storage. The result is a platform that brings surveillance with AI analytics to the Edge for instant visual intelligence applications.