Pixel Velocity Overview

Published Jun 21, 2012 00:00 AM
PUBLIC - This article does not require an IPVM subscription. Feel free to share.

High end video analytic systems that are still in business are few and far between. One company that we hear about from time to time is Pixel Velocity. Since they do not get, and evidently do not pursue much press, we decided to learn more about how they compare to other video analytic and VMS offering. This note shares our findings.


Pixel Velocity's main offering, called Pixel Video Fusion [link no longer available], is an end-to-end surveillance and analytic system, targeted at wide-area monitoring. Typically, this includes transportation, military, critical infrastructure, and high-end corporate security, not everyday applications. The system consists of three key components:

  • Proprietary cameras: Pixel Velocity offers a single model of IP camera, in box form factor, with 5MP resolution. These cameras must be used with Pixel Video Fusion's analytics. Analytics processing is performed on uncompressed video at the camera, with alarms sent to the server for processing. Third-party analog and IP cameras may be integrated to Video Fusion, but this is typically only for live viewing, when other types of camera are required, such as PTZs or thermal cameras.
  • Video Commander VMS: A proprietary VMS, called Video Commander, is used for recording and monitoring of the Pixel Video Fusion system, and most suited for wide area monitoring applications. Video Commander uses image stitching to create overall views of large areas, and in conjunction with Pixel Velocity's cameras, allows for a region of interest to be created within this overall stitched view. This region of interest may be dynamically moved and changed, allowing for a higher-resolution virtual PTZ control thoughout the larger FOV. This differs from many virtual PTZ controls in that it is not simply a digital zoom function, but an increase in image resolution within this region. Partners may use pre-loaded COTS servers from Pixel Velocity, or install this software on their own (or customer-provided) hardware. 
  • Analytics: Pixel Velocity key analytic offering is the Perimeter Protection System, a wide-area activity tracking system, similar to SightLogix's SightSensor system, detailed below. Additionally offered are application specific analytics used to detect wrong-way travel through airport exit lanes, as well as falling object detection, used to detect slip and fall incidents.

Pricing and Channel

A four-camera Pixel Velocity kit, including cameras, appropriate lenses, server with 4TB of storage, software, and UPS has an MSRP of $28,990 USD, about $7,250 per channel. Adding the geo-locating Perimeter Protection System to this kit increases price to $48,990, or over $12,250 per channel.

Pixel Velocity is sold only through select integrators, with no sales direct to end users. According to Pixel Velocity, since they prefer to be involved in each project, from design stages through installation, this channel is kept strictly controlled, to keep support issues to a minimum, and increase confidence in the product.

Perimeter Protection System

Pixel Velocity's main analytic, the Perimeter Protection System is used for tracking objects, people and vehicles, across a wide area. Similar to SightLogix, this is performed by correlating object position within the image with specific coordinates on a satellite image. Within this field of view, created from a stitched multi-camera image, areas of interest may be configured, so that only a specific driveway or fenceline creates alarms, instead of the entire area. These areas may be configured to alarm only on people or vehicles, time of day active, and alert actions. These functions are demonstrated in the following video:

Third Party Integration

Pixel Velocity offers limited integration to third-party VMS and PSIM systems, displaying video and alarms in these respective products. These integrations are performed on a project basis, and include only Genetec and CNL, to date. Users wishing to integrate Pixel Velocity to their VMS or PSIM should be aware of this, as additional integration expenses may range in the tens of thousands of dollars, if even possible.

Competitive Options

Pixel Velocity's most likely comparison is to SightLogix, as few, if any, geo-location analytics are commercially available, and none with the same presence. Though not a direct comparison due to differences in resolution and Pixel Velocity's end-to-end approach, SightLogix is likely the less expensive of these two options, taking the following into account:

  • Four-camera Pixel Velocity Perimeter Protection System: $48,990 USD MSRP
  • 4 x SightLogix Visible Sightsensor: $19,980 ($4,995 each)

Accounting for potential differences in dealer discount, based on this pricing, users could likely purchase and configure a server with a third-party VMS and SightSensors for less than a similar Pixel Velocity system. The two differ in some key ways, however:

  • Unlike Pixel Velocity's end-to-end approach, SightLogix's SightSensors are intended to be integrated with third-party VMS systems. SightLogix provides software (SightMonitor) only for configuration and monitoring, not recording. This allows for more flexibility in platform, but requires the integrator to potentially perform more integration work, and the level of integration varies depending on the VMS platform. Integration between Pixel Velocity's proprietary equipment, on the other hand, may be simpler and more predictable, since all components are made to work together.
  • Pixel Velocity offers only one camera, a visible 5MP model, which should may detection at greater ranges and wider FOVs than SightLogix, since SightLogix offers VGA resolution only. Pixel Velocity is therefore likely more useful in instances where visual inspection is required, instead of simply object detection. This is made more useful by Pixel Velocity's image stitching and region of interest functions, as well, which most VMSs lack. However, SightSensors are also available in thermal models, which may be useful in areas where visible detection is more difficult, such as ports or facilities with treelines.