This website and in particular this paper, describe how one can 'see' the 'sightfield' of a camera. To quote,

[A] linear array of computer-controlled LED lights is held in front of a surveillance camera. Even though the camera is concealed inside a dark dome with a shroud to hide any information about which way the camera is aimed, its sightfield may be visualized by abakography. An abakographic wand is moved through the space to trace out a “darkpainting” or “sightpainting” (inverse “lightpainting”) that allows one or more abakographic cameras (e.g. one or more people wearing Meta-View Spaceglasses), to observe the surveillance camera’s sightfield. This may be done by way of 3D (three-dimensional) AR (Augmediated Reality). The color change of the abakographic wand (from black to blue in this example) indicates, over time (animated from left-to-right), the surveillance camera’s field-of-view and extent of coverage.

PixStix details

Accompanying this figure is the following description.

A PixStix Abakographer is swept through space like a wand. Typically there are hundreds of light sources along its length, but for simplicity, only 8 are shown here, numbered L1 through L8. Here the lightfield test color, is green. When the lightfield falls within the sightfield of a surveillance camera, the color changes to the abakographic color which is red in this example. A second camera, the abakographic camera, captures the abakograph of the wand, to build up a representation of the sight-field over time. The PixStix Abakographer contains a processor to present a special pattern, while sensing changes in response to nearby surveillance cameras.

Can someone explain how this works in simple terms (and without jargon or quoting verbatim from the paper)? How can a series of LEDs be used to detect whether a camera is 'seeing' them?

  • $\begingroup$ The LEDs are not detecting the field they are being controlled by a computer to display the field. Presumably that computer is taking images from the camera as input, and is able to apply pattern recognition to locate the wand. Then the LEDs are illuminated accordingly. The clue is in the sentence " This may be done by way of 3D (three-dimensional) AR (Augmediated Reality)." $\endgroup$ – JMLCarter Sep 5 '17 at 18:33
  • $\begingroup$ @JMLCarter: "The LEDs are not detecting the field [but instead] they are being controlled [...]" -- That's my understanding, too. But I'm not sure what to make of the last sentence of the OP description: "The PixStix Abakographer contains a processor to present a special pattern, while sensing changes in response to nearby surveillance cameras." ... $\endgroup$ – user12262 Sep 5 '17 at 18:57
  • $\begingroup$ Thanks for clearing things up. I was under the impression they did not have access to the camera's video/picture output. But if they do, then mystery solved. $\endgroup$ – Jet Blue Sep 6 '17 at 0:20

From the linked paper:

There are three ways in which a sightfield may be visualized [...]

  1. a "friendly" camera that we control;
  2. a hostile camera that is controlled by another entity, to which we may apply signals intelligence to receive its signals;
  3. a camera we can't receive signals from, possibly because it does not yet exist


Scenario (3) arises when we are unable to do so [...]. In this case we merely affix a small wireless camera into the space in alignment with the existing (unreadable) or not yet existing (i.e. proposed) camera/camera-location.


By "painting" with a test color or test pattern, a response manifests itself and is sensed, for which the abakographic wand displays an abakographic pattern (different than the test pattern).

In other words, the system is dependent on getting access (wittingly or unwittingly) to the video recorded by the surveillance camera, or a stand-in camera put in its place. When the surveillance camera detects the wand, the wand lights a corresponding pixel for the light-painting picture. It doesn't have any novel way of detecting whether photons are landing on a camera sensor or not; it simply observes the camera's output. So the science is legitimate, and the system is realistic — they just try not to mention the major caveat for PR purposes.

  • $\begingroup$ Ah I see, thanks. If you have access to the video the camera is recording, what is the point of this? Can you not just look at the video to determine the camera's "sightfield"... $\endgroup$ – Jet Blue Sep 6 '17 at 0:16
  • $\begingroup$ @JetBlue the main advantage (other than looking cool and generating buzz, which may be the primary uses of this technology) is that you get to see it from a different perspective. The perspective from the camera itself isn't great for spatial reasoning, especially for the space between the camera and the nearest opaque object. Seeing a blob of light emanating from the camera and filling the room is more impactful and/or useful. $\endgroup$ – hobbs Sep 6 '17 at 0:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.