The upgraded IVSS 7000-I features a scalable GPU design to meet the ever-growing project requirements. The unique Dahua deep-learning algorithm with GPU group enables IVSS7000-I to perform face recognition, perimeter protection and metadata with high accuracy. Offering advanced AI features, it even enables danger prevention via faster video content inquiry, rapid discovery of targets as well as real-time and accurate alarm in a variety of applications, providing AI-based all-in-one surveillance solution for customers.

Dahua Technology’s AI-powered face recognition technology enjoys a high reputation in the industry, as testified by multiple first prizes in international competitions.

Equipped with Dahua face recognition technology, IVSS7000-I series records faces by structured data and performs real-time crosscheck with reliable accuracy to figure out the face by features of the target. Supporting up to 256-channel face recognition by device and up to 256-channel face recognition by camera, IVSS7000-I can analyse up to 50 face databases with 300,000 face pictures in total. In addition, the upgraded IVSS7000-I also supports search by face image which means users can upload face pictures to IVSS7000-I and compare them with recorded ones by similarity.

These abilities enable the NVR to be utilised in diverse applications such as, VIP recognition, blacklist alarm, stranger identification, attendance management, access control, etc. It divides faces detected into different categories as defined by the users, and acts accordingly, like notifying the manager that a VIP customer is coming or sending an alert when a persona non grata is identified. Users can choose regular/stranger mode to trigger the required alarm action according to different applications.

Focusing on people and vehicles, perimeter protection automatically filters out false alarms caused by animals, rustling leaves, bright lights, camera shaking and other irrelevant elements. It also enables the system to act as a secondary recognition for the targets, thus significantly improving alarm accuracy with saved costs.

In terms of perimeter protection, IVSS7000-I supports up to 256-channels with each channel allowing up to 10 IVS rules.

Metadata is feature attribute information of a target object which can be used for data retrieval. Currently, there are mainly four types of metadata that are human face, human body, motor vehicle and non-motor vehicle. IVSS7000-I supports up to 64-channel video metadata, making it possible for operators to quickly and easily search through multiple channels and over a long duration to find out when and where a person/vehicle of interest appeared, which significantly enhances accuracy and efficiency.

The IVSS7000-I employs an optimised GUI which is easier for users to learn and use. With the same GUI working for local output and PC client, the training cost for system operators is reduced. Comparable to VMS client, the new GUI design is more user-friendly for project-level customers.

In addition to the above features, IVSS7000-I also allows independent PC client for remote access so that users will not be bothered by things like web plug-ins and browser updates. It's also worth mentioning that with a new security centre to keep hackers away, IVSS7000-I series provides a safer network environment for users.

Being multifunctional, IVSS-I can meet variable needs of users, thus it can be applicable to a wide range of scenes, including entrances and exits of banks, business parks, schools, prisons and transportation centres, such as airports, railway stations that are sensitive to strangers and criminals; or the shopping malls that want to protect customers from petty thieves; or football stadiums that choose to ban those known hooligans for good.