August 31, 2006 Panoptic’s proposed C-Thru 3D Video Surveillance System could be loosely described as a formalised, scalable implementation of Superman’s X-ray vision. The system enables one or more surveillance agents, using a single high resolution, auto-stereoscopic display, to remotely monitor the security situation of an arbitrarily large number of locations at-a-glance. Agents can see, hear and transport their focused viewpoint through walls, floors and ceilings, zooming into a specific location to monitor it at a level so acute that it seems beyond the levels of even science fiction. Designed to enable both wholistic site-wide and granular-level security, the system is ideal for monitoring airports, shipping ports, transit sites and other ports-of-entry, hotels, casinos, shopping malls, campuses, military bases, large buildings and building complexes, offering total situational awareness at a glance.

Though it seems almost preposterous, the remarkable capability of the Panoptic system is the result of the combination and integration of multiple technologies and insights. Most of these individual component technologies have been developed to varying degrees by other companies and Panoptic acknowledges that some groups of developers have combined varying subsets of these technologies, but describes its proposed solution as unique and unprecedented in its integration/combination with its proprietary Intelligent HyperCompression (IHC) being the key to the system's success.


More than 1,200 New Atlas Plus subscribers directly support our journalism, and get access to our premium ad-free site and email newsletter. Join them for just US$19 a year.


It describes IHC as “an object oriented wavelet-based compression system for security and surveillance applications” and a “programmable compression hardware and software solution.”. IHC enables object selection and extraction and features high-resolution viewing while conserving bandwidth, meaning high quality images can be selected and viewed with higher data rates and higher frame rates than the significantly reduced background.

The system is claimed to enable a single agent to monitor thousands of places at the same time without overwhelming them with the massive amount of information being processed and displayed. It enables this by using artificial intelligence to determine what’s important to look at. The agent's mental load is radically optimized towards monitoring only what is most critical to security: presence and movement of (groups of) people and vehicles inside a campus/site and its buildings as well as any security-sensitive events (such as fires, smoke, badge access, assemblies, gun-shots, people running etc).

These security-critical events and alerts are adapted into one system then correlated against each other based upon defined security policy and rules. This automated, root cause analysis enables disparate real-time security data to be combined into “an integrated panoptic security command centre (over)view.”

All information streams and elements which may detract from the overarching goal of transit-site security surveillance are dynamically subtracted from the agents view and not displayed unless triggered by the system as relevant or requested by the agent. In other words, agents only monitor what they need to relevant to surveillance and transit-site security.

Events which need analysis by the human operator are flagged by what Panoptic calls the “Artificially Intelligent suspicious activity detector.” This detection system uses pattern recognition and possibly also neural nets to detect potential security or safety indicative events such as: the assembly of groups larger than X people; loitering; running; stressed shouting; suspect clothing (balaclavas, squad-type clothing, bullet-proof vests) or suspect items hidden under clothing; nervousness of individuals and groups; fires; smoke and gasses; explosions etc.

The three dimensional C-Thru surveillance system is effectively a complex form of Augmented Reality which fuses real-world video imagery with volumetric models in a real-time 3D display to help observers comprehend multiple streams of temporal data and imagery.

The “C-Thru” imaging is enabled by the combination of imagery supplied by many low-cost 360º cameras and dynamic background subtraction, which makes the walls of buildings and vehicles transparent, offering at-a-glance situational awareness of security-relevant people, objects and vehicles in motion.

Key elements/components of the Panoptic C-Thru surveillance system include: * Concept of real time (stereoscopic) display of video-sprites of people and vehicles inside and among user navigable "transparent" 3D wire frames of building(s); building complexes or battlefield areas. * Usage of panoramic video imaging systems (instead of classic cameras) * Rapid production of 3D model of site/campus through reverse "volumetric painting" using DV camcorder-based capture system * Real-time dynamic fusion of panoramic video streams into 3D model of site/campus/battlefield. * Real-time extraction of video imagery of moving and semi-static objects (people, vehicles, animals etc.) with live transposition into 3D scenes. * Live video-based 3D tracking of moving objects within designated security-sensitive areas. * Autostereoscopic display for added accuracy of depth perception by surveillance agent. * Intuitive 3D user-interface for empowered analysis and management of security threats or breaches, or battlefield situations (including 3C activities) * Optimized detection & remediation of surveillance coverage blind spots * Real-time dynamic fusion of video data from mobile cameras (on vehicles (e.g. security patrols or battlefield reconnaissance vehicles); helmet mounted; on micro-UAVs or small helium blimps;) * Integrated display of results from triangulation based gunshot locator into the 3D model. * Alternative panoptic modes for single screen video only panoptic surveillance leveraging latent capacity of human brain for massively parallel processing of defocused viewing combined with peripheral field-of-view imagery. * Integrated on going or on demand identification through facial recognition against customer definable database of security sensitive persons. * Integration and correlation of multiple events from physical security monitoring systems.

View gallery - 10 images