INTELLIGENT IMAGE ANALYSIS

AxxonSoft offers video analytics solutions that deliver multiple business and security benefits through neural networks powered by AI (artificial intelligence) and deep learning algorithms.

Multi-domain monitoring is useful for geographically dispersed systems, e.g. in retail chains with many locations. Configuration and operation of the remote supervision system is possible from one workstation. Additional advantages:

• Connecting a single workstation to multiple systems in different domains
• All server-related settings and devices are consolidated into one convenient view
• Operators can simultaneously access cameras from multiple independent surveillance systems.

Web clientin AxxonNext

Web client features:

• Secure connection over HTTPS
• Support for camera layouts, identical to the standard client
• Live video preview and convenient search of recordings
• Controlling PTZ cameras
• Digital zoom
• Audio from camera microphones
• Export still images and video
• Camera status view.

AxxonNet Cloud Service

AxxonNet is a free cloud service that connects to AxxonNext surveillance servers via the Internet. SSL encryption ensures secure data transmission.

You can create users and permissions with AxxonNext. The service also allows you to display a live view and access to the archive. Video from events / freeze frames are saved in the cloud and can be viewed by clicking on the appropriate link. The user can receive notifications about alarms and selected events according to his preferences, e.g. via the AxxonNet mobile application.

Mobile client

The application allows you to: view and search video recordings, control PTZ cameras, work with fisheye cameras, receive push notifications, run macros and use maps. Free apps are available for iOS, Android and Apple TV.

GreenStream

GreenStream automatically selects the lowest possible camera stream that matches the current video resolution on the client screen. For example, on a 1920×1080 screen with a 4×3 camera layout, each camera frame is only 480×270. GreenStream eliminates the need to send full resolution streams from all cameras when we are not able to really use it.

Metadata from IP devices

Metadata is a compressed description of the moving objects in the scene. It is used for real-time video analytics or Moment Quest smart search. Metadata is generated in cameras with built-in image analysis. Thanks to them, there is no need to decompress the video on the server side, which contributes to a significant offload of the video server's processor. This, in turn, reduces hardware costs or increases the number of cameras per server.

GPU offload

Work without GPU:

• Reduced CPU power requirements, especially when decoding H.265 video sources
• Reducing decoding time
• Servers can simultaneously handle more HD video streams and higher.

Working with the GPU:

• You can connect more cameras to one server, reducing hardware expenses and performance issues
• Thanks to this technology, the client station can display more video channels or use lower performance components.

UDP and multicasting

Axxon Next offers a range of tools to reduce network load and increase the efficiency of security systems. Live video can be sent from the server to client stations via UDP. Multicasting is also supported to minimize network usage and optimize resource consumption.

Handling of external events - quick and easy integration with third-party systems

Connection to external devices and systems: SKD, SSP, SSWiN and more. Axxon Next can: handle external events and store them in the database, and associate the event with the image. It can also search for events by string or display them in real time in the events panel.

Macros – event response wizard

Axxon Next supports flexible configuration of complex system response to any set of events. Use if x then y logic to create a macro that will automatically perform an unlimited number of actions on the system. Macros allow you to program reactions to specific events at the system and device level.

Redundancy– your VMS is always up to date

Fallover mode quickly switches to a standby server when communication with the primary server is lost. The standby server automatically takes over all the functions of a failed or malfunctioning server. Instantly resume live video and archive recording. It is possible to shut down any server without data loss, e.g. for maintenance purposes.

Updating servers and client stations

Silent update of servers. You can update all servers in a cluster silently. To do this, select the required software distribution of the zip file or provide a www link.

Automatic update of client stations, when the Axxon Next client connects to a server that has a newer version of the software, the operator will be informed about the possibility of updating. After confirmation, the update process is performed automatically.

LDAP authentication

This feature greatly simplifies user management for system administrators in large companies. Operators can log into the surveillance system by entering their Windows domain credentials. The system administrator connects the LDAP directory to Axxon Next and selects the users to add. It can also bind VMS access rights to corporate directory groups. After deleting a user profile on the LDAP server, the system can also automatically delete it from Axxon Next.

Measurement and detection of elevated temperature

The configurable temperature values are color coded both in real time and on video, eg green for normal temperature and red for elevated temperature. It is possible to configure automatic scenarios in case of detecting elevated temperature.

Thermal cameras measure body temperature with an accuracy of +/-0.5 (+/-0.3 °C with Blackbody calibration). It takes 1 second, so there is no risk of congestion at the measurement site. The temperature is measured remotely  at a distance of more than 2 m from a person. In addition, integrated video analysis ensures that only human faces are captured for temperature measurement. In this way, false alarms related to, for example, hot drinks are avoided.

Axxon Next VMS is not medical software and is only used as an entry checkpoint. The obtained results should then be confirmed by authorized medical devices.

Detection of social distancing behaviour

Our behavioral analytics, based on artificial intelligence and using patterns of behavior, can detect violations of predefined social distancing restrictions. If a violation is detected, notifications can be sent to the appropriate people. All cases are recorded and quickly available in video recordings by searching by alarm events.

Alarm about missing mask

The tool recognizes both the presence and absence of a mask. After detecting irregularities, the system sends an alarm to the operators. All violations are recorded and the video of the event can be searched immediately.

AxxonNet is video management software that enables communications service providers and system integrators to build cloud-based video surveillance (VSaaS) solutions that leverage the power of AxxonSoft's neural network analysis and intelligent search for recorded video.

Web client characteristics:

• Microservices architecture; horizontal scalability; one-click deployment.
• Management via Kubernetes.
• Application monitoring via Prometheus; logging using the ELK stack.
• Sending events to loT platforms; export video recordings of alarm events to streaming services, e.g. Amazon KVS.
• Fastly customizable GUI with OEM GUI development kit.

VMS servers in the cloudis a dynamic allocation of resources for each new client or task, scaling up to 10k. servers and the failover service.

AxxonData's reporting systemsenable an unlimited scalable reporting system and create new reports with a simple configurator.

POSSIBILITIES OF BUILDING AXXONNET CLOUD SYSTEM

On-site VMS servers that can be configured and monitored via the cloud (web client, mobile apps) or locally (desktop client):

• The video archive is stored locally; recordings are not affected by internet connection failures.
• You can store selected video fragments in the cloud for quick access from different devices.
• The cloud makes it easy to access, configure and monitor any local system.

Smart Video Surveillance – the ability to centrally operate an unlimited number of cameras and servers. Designed by professionals, the user-friendly interface meets all the conditions for comfortable and efficient use of the system. First of all, it offers easy access to the information collected by the system and the ability to make immediate decisions in the event of a security breach.

Audio processing thanks to the network audio module - a high degree of customization. Ability to separate or synchronize audio and video recording.

Image Processing - the highest level of motion detection, face detection and classifying objects and their behavior. Intelligent filtering by the system of image fragments with detected suspicious activity.

Cooperation with IP systems - trouble-free operation of most network cameras available on the market, continuous updating of the list of devices and supported functions.

Integration with any hardware and software – easy integration of Axxon Soft into an existing system of devices.

Access control – an advanced access control module ensures the highest security of the facility. Available options include e.g. working time calculation module, pass control module or photo ID module.

Fire alarm module - advanced integrated protection in case of potential fire or burglary.

Automation of event handling - the Axxon system recognizes events, analyzes them and takes action according to scenarios. It helps to fully improve the operation of the security system by automating most processes.

Mobile application Axxon Intellect enables remote viewing of camera images and access to archive materials. In addition, the user can easily manage camera settings, notifications, interactive maps and many other functions.

Axxon Intellect Enterprise abandons the classic client-server architecture and three-tier architecture in favor of a distributed model. The platform uses three components: server, client and gateway/router.

Servers form a network segment with a P2P (peer-to-peer) structure - all data and connections triggered by events are configured as S2S (server-to-server). Clients connect to one or more servers, either directly or through gateways.

Gateways serve as routing devices and optimizers that define segments in a distributed network. The Axxon Intellect Enterprise peer-to-peer network uses multiple types of connectivity (LAN, WAN, cellular) between network users and their cumulative bandwidth, unlike conventional centralized resources where a relatively small number of servers is essential to the service or application.

Distributed technology means Axxon Intellect Enterprise is highly reliable and failsafe - if there is a problem with a module or network connection, Axxon Enterprise automatically switches to another live server. The Axxon Intellect Enterprise server component currently runs on NT, Linux and AIX.

The most important features offered by the POS Intellect module:

• video surveillance using analog and IP cameras (support for over 9900 different device models)
• intelligent customer behavior analysis (people counting, queue length detection, heat maps)
• control of the cashier's work - synchronization of data from the cash register with the video recorded by the camera located at the cash register; highlighting a line for an event where a certain content will appear
• Archive analysis, including video search by bill item (receipt text search) and slow-motion playback
• possibility to work remotely - creating statistics of cash transactions in each store of the retail chain (web reports), real-time remote video surveillance and access to archival video material
• configuration of the system's automatic response, creating scenarios for specific events.

POS Intellect is efficient video surveillance, profiled cash registers and full control of the customer's basket.

The Axxon Intellect POS module offers the tools necessary to fight and prevent fraud, theft and other abuses faced by small and large stores on a daily basis. Intelligent video surveillance and tight integration with other systems will ensure full control over all transactions and activities - from checkout points to store space.

POS Intellect offers video management software for every POS in the store. Automatically records all events, controls the content of accounts and saves all collected information in the database on an ongoing basis. The system synchronizes the video image recorded from the cash desk. It controls both the list of products on each account and the payment process, regardless of whether it is a cash or card payment. The module verifies the correctness of the data entered by the cashier and supervises the act of giving change to the customer. This minimizes the likelihood of fraud and ensures tighter control over the actions of employees and customer behavior.

POS Intellect gives you access and control over all your retail sales. It allows for instant display and printing of any video frame, simultaneous monitoring of multiple POS terminals (Point of Sale) on the same screen and customization of sales reports. Just specify the required parameters and the system will automatically sort and search for data, respectively - by date/time, terminal ID, group assignment, cashier's name, price, etc. In addition, the system calculates statistical sales data, making it a powerful business analysis tool.< /p>

Auto Intellect – License plate recognition and traffic monitoring system

Monitoring, controlling and managing traffic or fleets of vehicles is often critical to the operation of many companies. Therefore, they may need effective license plate recognition technology to track vehicles, manage parking spaces and warehouse traffic in real time.

Axxon Auto Intellect  it easily connects to your existing hardware and software – regardless of type. It's the choice for simple installation, easy configuration, and increased availability.

Basic Auto Intellect tasks include:

• vehicle license plate recognition, vehicle flow monitoring and specific vehicle search
• automatic speed limit enforcement by capturing photos/videos
• automatic red light enforcement by capturing photos/videos
• collecting traffic data and transferring it to an automated traffic control system
• recognition of wagon registration number and freight container number (intelligent railway monitoring).

Number Plate Recognition

The system, equipped with the best Adaptive Recognition technologies, recognizes and registers car registration plates, even moving objects, and compares them with the database. It also records all vehicles that pass through a defined zone, a photo of the car and license plate, and notes the date, registration time and direction of each vehicle.

• More than 95% accuracy
• Integration with access control systems
• Blacklist functionality
• Speedometer integration.

Vehicle traffic monitoring

Traffic jams are a recurring problem for both public services and commercial enterprises.

Auto Intellect makes it easier to manage a high-traffic area. Intelligent transport and motion sensors provide real-time data you need to prevent problems before they occur and efficiently organize your supervised space.

• Collection of traffic statistics and their analysis
• Up to 6 lanes per camera
• Automatic detection of traffic violations - speed limit, no stopping, no overtaking, driving in the opposite direction, etc.
• Pre-configured settings for vehicle types - motorcycles, cars, trucks (up to 12 m long), buses.

The benefits of Axxon Face Intellect:

• Non-intrusive, non-contact identification process
• Detection of human face regardless of facial hair, hairstyle, glasses, aging and other conditions, e.g. background
• Compatibility with other databases
• Real-time notifications of recognized people
• Easy integration with existing systems
• Automatic database synchronization
• Integration with access control systems
• Recognition algorithms powered by Cognitec SDK v 5.0.

Face recognition module is used to automatically identify people based on recorded video images. It recognizes faces recorded by video cameras and then compares their parameters with digital templates stored in a special database. This module can be used e.g. in conjunction with an access control system. The facial recognition module guarantees high accuracy.

First, the face detection tool registers the presence of a face within the frame. Then the module adopts one of two operating modes: identification or verification mode.

In identification mode, the parameters of the faces recorded by the camera are compared to the templates in the database. This helps, for example, to determine whether a person is blacklisted or VIP.

In the face verification mode of the access card holder or a person holding another identity document, when passing through the control, their parameters are compared with the parameters of the rightful owner whose photo is stored in the database. This allows you to verify that the person seeking access is really who they say they are.

Enhanced access control

In this case, facial recognition is performed in verification mode. The camera at the entrance registers the faces of incoming people and is connected to the face detection tool and the Intellect facial recognition module. This scenario is triggered when the access card is presented to the card reader. Face Detection captures a video image of the face of the appearing person. The module analyzes the parameters of the image and compares them with the data of the real cardholder in the database.

The system can be programmed to automatically grant or deny access after proper identification if there is a low probability of the person's authenticity. This process can also be controlled manually by the operator. The image of the registered face, personal information and the degree of similarity are displayed on the operator's screen. In this scenario, it is the operator who decides whether access can be granted.

Automated face check

In this case, the facial recognition module is in identification mode. A camera with an active face detection tool is installed at the entrance. When a person comes into view, the image of their face is captured, processed and compared with images in a database, e.g. in a database of VIPs or people on the so-called blacklist.

When a face is recognized, the display shows the captured image, a reference image from the database, personal information, and comments about the category the person belongs to. If the image matches the reference image from the database, no comment is needed anymore as the system automatically classifies the person into a specific group. Depending on the outcome of the diagnosis, appropriate action may be taken. For example, if a person is considered a VIP, the operator must notify the facility manager. If a person is blacklisted, guards must prevent them from entering the facility.

The face recognition module can be used for various operations related to the automatic identification of people. The Intellect platform allows you to create scenarios that can cover any of the security subsystems. The type of scenario, the level of operator involvement, as well as the type of interaction between the facial recognition module and other system components are determined by the security policy of a given facility.

The face search module allows you to create a database of all faces registered by the cameras. Allows you to search this database for similar faces. The module allows you to speed up the search of the archive of people you are interested in and to obtain statistics on their capture by various cameras.

Applications and uses

Face Intellect is the perfect solution for applications in public places: airports, stadiums, transverse border control zones, prisons and military or restricted access areas. It will also work in the detective industry.

The Face Intellect module authenticates the cardholder by automatically comparing his face with the image stored in the database.