Dahua DHI-IVSS7008-1T 128-Channel Intelligent Video Surveillance Server

The Dahua Intelligent Video Surveillance Server Series (IVSS) combines Deep Learning algorithms, a powerful Graphics Processing Unit (GPU), and an advanced Network Video Recorder (NVR) into an all-in-one Artificial Intelligence server. Dahua built the IVSS series from the ground-up as a dedicated AI surveillance server that eliminates the complications of browser updates, incompatible camera plug-ins, incompatible flash memory, and outdated Web technology.

The integrated GPU module and advanced deep learning algorithms perform powerful video structure analysis with metadata, achieve precision human facial analysis, and support four (4) channels of real-time face recognition based on a database that stores up to 100,000 faces. IVSS technology captures, records and categorizes various facial features, including age, gender, expression, and whether the target face wears glasses or sports facial hair; compares the data to faces in the database, and displays real-time facial recognition data.

IVSS offers advanced features to customize the powerful Artificial Intelligence for different applications. For example, facial feature filtering displays only those faces that meet target criteria and Stranger Mode detects a face not stored in the database and triggers an alarm, takes a snapshot, or records the face for further scrutiny. The IVSS is ideal for applications that require entrance/exit management, where knowing who is coming and going is a valuable asset.

Functions

Real-time Face Recognition
The IVSS7008 performs real-time facial recognition on up to four (4) channels simultaneously. The server captures and analyzes facial features to determine gender, age, expression, glasses, moustache, and mask, and then can record the faces and store the associated structured data. The server also filters incoming video to display faces that match target features.

Stranger Mode
IVSS supports two facial detection modes: Regular and Stranger. In Stranger mode, when IVSS detects a face not stored in the device’s database, it can trigger an alarm, a buzzer, take a snapshot, or initiate recording. IVSS also incorporates a Similarity Threshold, where the server triggers an alarm for a face that does not match a minimum recognition level. The Stranger Mode and the Similarity Threshold are designed for critical infrastructure sites where access control is of vital importance.

Efficient Search
IVSS searches images by facial features to improve search efficiency. The server support uploading and storing 100,000 facial images for comparison with recorded faces. This search capability allows operators to quickly and easily search multiple channel recordings to determine when and where a person of interest appeared in a recording.

Face Database Management
In addition to the 100,000 facial images, the IVSS database also stores names, genders, birthdays, nationality, address, and ID information associated with each facial image. IVSS offers powerful and configurable database management features that can be applied to each face recognition channel independently.

High Efficiency Video Coding (H.265)
The H.265 (ITU-T VCEG) video compression standard offers double the data compression ratio at the same level of video quality, or substantially improved video quality at the same bit rate, as compared to older video compression technologies. H.265 offers such impressive compression by expanding the pattern comparison and difference-coding, improving motion vector prediction and motion region merging, and incorporating an additional filtering step called sample-adaptive offset filtering.