Table 2.

VVSP concept contribution vs. traditional concept approach.

Task	Traditional Approach	VVSP Concept	Contribution
Video signal monitoring	Analogue PAL or NTSC digital (HD) SDI or CoaxPress.	Digitalized H264.	Easier interfacing to command control systems. No need for additional hardware/software.
Camera interfacing	PAL/NTSC—easy (HD)SDI/CoaxPress—easy, limited range CameraLink or LVDS—difficult, converter to HD-SDI needed introduced latency and possible format change.	Easy connection to any interface on camera with just choosing proper interface board. This is important for thermal and SWIR cameras, which predominantly utilize CameraLink or LVDS interface for maximal performance.	Interfacing to any camera interface is direct without any resolution change or additional latency.
Pan-tilt control	Via Pelco-D, ONVIF or similar protocol.	Via ONVIF protocol. For a critical algorithm such as target tracking, pan-tilt drive is direct from the VVSP without any additional latency.	VVSP has an advantage in applications executed directly on HW- and AI-based control.
Pan-tilt positioner slip ring communication	Limited to coaxial or Ethernet interfaces; otherwise, converters are needed. Additional interface is needed for pan-tilt, LRF and lenses control.	Only uses one Ethernet interface for both video streaming, camera, lenses, LRF and pan tilt control.	Reduces requirements for pan-tilt positioner slip ring.
Fanless operation	Some processor modules may require additional cooling.	VVSP is completely passively cooled.	Efficient system cooling which is tested even in desert conditions.
Resolution	Limited to PAL or NTSC. High-resolution/dynamic-range imaging not supported. Digital interfaces have no limitations.	Full HD resolution and high dynamic range that is critical for, e.g., thermal imaging, which is fully supported.	Much better image quality.
Latency	Minimal latency before external compressing hardware is used.	About 300 ms. Please note that critical algorithms (e.g., tracking) are executed on vVSP, without any latency.	If an MSEOS is integrated into the C2 system, there are no differences.
Target tracking	Executes on external hardware that controls pan-tilt. For good performance, the controller should be installed close to the MSEOS.	Executes on VVSP. Utilizes full resolution and frame rate of all sensors. No additional latency. Very compact solution.	Considerable advantage for VVSP due to direct sensor access.
Image stabilization	Limited to image-based algorithms. Depends on the particular scene.	Can utilize different methods (e.g., IMU stabilization [10]).	Much more flexibility than the traditional approach.
Image fusion from multiple cameras (e.g., thermal and visible)	External hardware. Time synchronization and image registration might be difficult to calibrate.	A dedicated VVSP module can run image fusion algorithm acting as a “virtual video channel”.	More flexibility and easier calibration.
AI target recognition	Need to digitize the image prior to the application of AI modules. The concept limits resolution, which reflects on AI-based solution performance.	Full resolution of the image can be utilized for AI-based target detection application. In cases of a lack of computational power, additional VVSP modules can be added.	Much more flexibility, especially if target detection is linked to target tracking with pan-tilt movement in order to keep a tracked target in the center of the scene.
Image enhancement	No possibilities.	Easy implementation of various image enhancement algorithms.	Especially important for thermal imaging.
Laser range finder (LRF) application	External command for measurements. Targeting reticle is very hard to implement in continuous zoom cameras.	Simple implementation of targeting reticle even in continuous zoom systems [6].	Easier integration. Additionally, it enables the implementation of complex applications (e.g., LRF-aided target-in-lock indication for tracking systems).
Overall moving payload weight	Minimal or no extra payload.	Each VVSP module adds about 0.45 kg in weight.	This is a small drawback since the weight of cameras and lenses in long-range MSEOS is much bigger.
Private network for Ethernet devices	Additional manageable GbE switch is required.	A manageable GbE switch is integrated in the VVSP module.	VLANs configuration enabled. Important devices connected to the network can be hidden from the outside world.
Command–control (C2) system interfacing	Requires protocol conversion box for video digitizing and range extension.	Everything needed for C2 system integration is obtained via a single Ethernet port.	Easier interfacing to the C2 system.
System scalability (adding a new camera to the system)	System is limited to a certain number of cameras or requires additional hardware and/or system architecture redesign.	Only an additional VVSP is required. This module is connected to the existing system network over GbE.	This is made possible by the initial concept of distributed system architecture.