Glossary

 

COMPUTER VISION
Refers to one of the sensory input branches of artificial intelligence that endeavors to teach a computer to actively “see.”  Algorithms are written and employed in a variety of combinations that enable the software to reliably identify, classify and track objects of interest within motion video, then provide a stream of metadata immediately reporting the analysis activities.

DARPA
Defense Advanced Research Projects Agency, DoD organization that provides some R&D funding to ObjectVideo to solve specific problems with the application of computer vision science.

EVENT
Security-related activity or other activity of interest that takes place within a camera’s field of view. Or, it’s anything you care about.

INTELLIGENT VIDEO
Refers to video analytics software and functionality that is based on the science of computer vision. Intelligent video must be able to perform at least the following: Reliably separate foreground objects from the static background; create a complete metadata description based on that active analysis; allow the user to interact through application of rules to the metadata stream. (long def but just trying to help…)

ONR
Office of Naval Research, government organization that provides some R&D funding to ObjectVideo to solve specific problems with the application of computer vision science.

PIXEL
The smallest addressable element on a computer screen. The higher the pixel resolution, the more information displayed on the computer screen. Also, taken together with other pixels, the smallest addressable element for intelligent technology to analyse.

SOLUTION PROVIDER
A catch-all term used on this site referring collectively to anyone who maintains a direct selling relationship with the end user, i.e. installer; dealer; system integrator; reseller; platform provider; consultant; blah, blah.

UAV
One of those confounded government abbreviations – stands for Unmanned Aerial Vehicle. Not technically an acronym, as it’s pronounced ‘you-ay-vee’, not ‘you-av.’

VIDEO ANALYTICS
Sometimes confused with ‘intelligent video.’ While intelligent video is a type of video analytics, not all video analytics are actively intelligent. Video analytics, as a broad category, also includes more passive technologies like video motion detection, license plate and facial recognition and vision-based industrial inspection. (glad you asked?)

VIDEO MOTION DETECTION (VMD)
A type of video analytics that, in theory, registers an alert every time a pixel moves within motion video.  Purveyors of VMD apply a wide varity of filters to their version of this ancient technology in an attempt to keep false alerts to a tolerable level.  The technology has legitimate application in indoor environments where there should be no movement whatsoever within the camera’s view.  VMD is marketed somewhat erroneously by some mfrs as having “intelligent” capabilities; VMD is not actively intelligent in the same way computer vision-based technology is. (out of breath!)

Please share!