Digifort’s 7.4 upgrade to its VMS

About Digifort

The Digifort open-platform VMS is proven, unified, trusted, and secure. It is used in many applications across the UK, including power stations, ports, smart motorways, industrial sites, public space, universities, multi-site retail and shopping malls. Digifort operates a one-time-buy, camera licensing model with no annual maintenance fees and with free software upgrades for users.

The Digifort VMS is future-proof, with simple expansion options available for cameras, devices and systems. Modular options for next-generation neural analytics, edge analytics, LPR, Facial Recognition and Point-of-Sale can be added at any time, to enhance site operation. Digifort is SDK integrated with over 400 camera manufacturer brands and over 10,000 cameras. It supports server failover in mirrored, ratio and parallel recording, along with cloud streaming and backup. Digifort also allows control of NVRs and DVRs, coordinating a mixed estate of brands into one common platform for live viewing and playback.

Digifort’s graphic user interface (GUI) runs on Windows-based Operating Systems, such as Server 2019 and Windows 10 with familiar display screens, functions, menus, help windows and manuals.

Analytics based on AI

Digifort has enhanced its Deep Learning (DL) AI video analytics engine to include metadata searching and processing. The analytics is state-of-the art, optimised to run on NVidia Graphics Processing Units GPUs using CUDA processing technology. Analytics can be used to identify object types and colour content in real-time– such as an intruder in a blue jacket shown here with a colour profile.

In version 7.4, metadata allows retrospective searches of the recorded video and can look for different objects and events. So, even if the system was configured to identify intruders with blue jackets, operators can apply new search criteria on the recorded video instead, such as identifying vans or bikes. The system only looks for metadata content, which relates to objects detected in the video, retrieving the associated recorded video for review. This means retrospectively processing 30 days of video takes just a few minutes.

License Plate Recognition (LPR)

LPR comes as a module within Digifort. In the 7.4 upgrade, there are three different recognition engines for plate capture. These are optimized for different vehicle criteria, such as stationary vehicles at a barrier or in a car park, through to fast-moving vehicles on highways. In a car park type application, Digifort will tag a plate as ‘entry’ or ‘exit’ to give a vehicle’s entry and exit dates and times, length of stay, count, car park occupancy and average occupancy. Gates and barriers can be opened automatically for authorized vehicles, using a centrally managed database. Overstays can be identified for ticketing and short stays or multiple car park entries by the same vehicle can be designated as a suspicious activity for review. In fast-moving vehicle applications, Digifort enables real-time traffic management on highways allowing for multiple traffic lanes, high traffic density, and most vehicle types.

A Region of interest (ROI) can be applied to specific areas within a camera’s field of view to look for plates, ignoring all other areas. This makes plate reading faster and improves server efficiency as processing demands decrease. There can be multiple ROIs per camera, supporting multi-lane, plate capture.

The Digifort LPR Bridge, also an additional module, links to external databases to give far more comprehensive vehicle and owner information. Depending on the database, this can include criteria like stolen cars, wrong plate, incorrect vehicle colour and much more.

Events (and alarms)

Events and alarm status in Digifort version 7.4 are listed within the surveillance client, with customizable features such as colour coding and grouping alarm types; alarm status; days to display alarms before archive; physical alarm mute; and alarm response and processing actions. Alarms can be reviewed, regardless of status, with actions such as pop ups and messages shown.

Operators are informed of new, incoming events and alarms, selecting whether these should take priority over live, operator actions, or not. These new events are presented with a snapshot of what triggered the alarm. Alarms can be ‘pushed’ to a specific screen for review in the surveillance clients, video walls, mobile devices (ID) and control rooms, for a real-time response and full audit trail. These include associated information such as video snapshots, selected camera views, maps, analytics, LPR, web pages and much more.

As an example, a stolen car plate could be used to trigger an alarm, sending the plate image and live video to a workstation. At the same time, the plate and multiple, relevant camera views could be sent to the video wall on monitor 18 and to operator four in the control room, with just a plate and vehicle image sent to mobile phones of relevant staff. The flexibility is endless!

Surveillance Client and Videowalls

Each Digifort Surveillance Client can control up to eight monitors with multiple camera views and split screens. They can be set up as an operator workstation or joined with others to create a virtual matrix videowall of unlimited numbers of monitors.

Custom screen styles can be configured in the Surveillance Client to build multi-screen camera views from a grid. Blue or red highlighting can be used to show feature selection and toolbars can clear when the client is not in use, to give maximum viewing area. Server online status can also be monitored in the surveillance client, raising an alarm if it goes offline.

Object Linking allows navigation from camera to camera by clicking arrows in the camera view, intuitively tracking people and objects around a site in live, playback, and export mode. Buttons and zones can be added to camera views and maps, allowing the operator to trigger alarms, relays, open doors etc. This is particularly useful for larger sites or where operators are not familiar with camera locations.

Administration Client and system configuration

The Administration Client resides on the system servers, or a PC with admin rights, and is used for setup and configuration. Camera configurations can now be duplicated, edited and applied to new cameras for fast setup. The videowall targeting, discussed under events, is configured using a simple event response tool and object linking is setup for each camera.

Maps, including interactive Google Maps, can now include icons and graphics which appear in the Surveillance Client and Videowall. Cyber security has also been improved with password management and One-Time-Password (OTP) support. The system audit logging allows more user actions to be logged and easily filtered for quick review.

To encourage even more integrations with third party systems, essential in larger projects, the scope of Digifort’s http communications commands has been extended beyond http GET and PUT, enabling greater exchange of information with third-party systems without the need for complex software.

Digifort’s extensive failover capability now includes failover and failback to an edge device, such as a failover server or camera with edge storage. This also enables self-healing, where streamed video interrupted by a network failure is repaired once the network goes back online.

License structure and server density

Digifort has four license levels, Explorer, Standard, Professional and Enterprise, ranging in capability. Version 7.4 doubles the channels per server, taking Explorer to 32 channels, Standard to 64 channels and Professional to 128 channels, with Enterprise remaining unlimited. This change supports greater camera density per server.