As video security is continuously evolving, Arteco presents OMNIA VCA 1.5: the video analysis suite that brings the Extreme VMS license to the most advanced level of intelligence. It automatically pulls meaningful and relevant information from video streams to detect people, objects and their surroundings, while continuously analyzing them, their movements and behavior.
Based on pattern analysis and machine intelligence, OMNIA VCA algorithms can improve the efficiency of security systems, reduce costs of management and decrease risk margins with the prospect of further developments on the horizon
The Omnia VCA 1.5 suite is a stand-alone application that must be installed on a dedicated workstation with Windows based OS, or on a Virtual Machine running in a Windows environment.
HOW IT WORKS
OMNIA VCA, in the OMNIA interface framework, processes video, identifies objects in the video footage (people, vehicles, and other items), and indexes them so that footage can be easily and quickly browsed and analyzed for different purposes:
1) to conduct post-event investigation and forensics searches
2) to enhance security situational awareness in real-time.
DEEP LEARNING
OMINA VCA technology is rooted in the Deep Learning machines learning model in which neural networks are being trained to extract patterns from massive amounts of data. As a result, the video analytics software is able to process video streams and contemporarily detect and classify every object that appears on screen.
Omnia VCA is equipped with two Deep Learning engines, one that performs the object detection (DL Object tracker) while the other performs exclusively the people detection function (DL People tracker).
The DL Object Tracker allows the automatic classification of the following objects in the scene:
•People
•Cyclist
•Car
•Bus
•Van
•Motorcycle
•Bicycle
•Bag
•Truck
The DL People Tracker allows the automatic classification of people present in the scene, therefore it is particularly suitable in the following cases:
•Presence detection
•People Counting
ANALYTICS
TAMPER
Tampering detection, lack of video signal
OBJECT CLASSIFIER
Reliable 3D calibration and object tracking engine with customizable object classes
RULES AND ALGORITHMS
Presence, Enter, Exit, Appear, Disappear, Stop, Dwell, Direction, Count, Abandoned object, Color filter, Speed filter, Tailgating, AND/OR/NOT logical rules
OBJECT DEEP LEARNING
Automatic detection and classification of 8 object classes.
PEOPLE TRACKER DEEP LEARNING
Automatic detection and classification of people and faces
MINIMUM SYSTEM AND CAMERA REQUIREMENTS
OS
Windows 10 Pro or Windows Server 2019
CPU
Intel i3 generation 8 or latest, or Intel Xeon E series
VIRTUAL MACHINE SUPPORT
VmWare, Docker
RAM
8GB or higher
GPU
Intel, NVIDIA Cuda series
NETWORK
Ethernet, 1 Gigabit
HDD SPACE
1TB HDD
ARTECO VMS VERSION
10.0 or higher
ENCODER VIDEO STREAM
H.264
RESOLUTION
320*240
FRAMERATE
10 fps
RECOMMENDED SYSTEM AND CAMERA REQUIREMNTS
OS
Windows 10 Pro or Windows Server 2019
CPU
Intel i7 or i9 generation 8 or latest, or Intel Xeon W series or higher
VIRTUAL MACHINE SUPPORT
VmWare, Docker
RAM
16 GB or higher
GPU
NVIDIA Cuda series
NETWORK
Ethernet, 1 Gigabit
HDD SPACE
1TB HDD
ARTECO VMS VERSION
10.0 or higher
ENCODER VIDEO STREAM
H.264
RESOLUTION
480P [Max 720p]
FRAMERATE
15 fps