Send us a Message


The following scenarios offer illustrative examples of Cognika's powerful video & imagery solutions in Intelligence, Surveillance and Reconnaissance (ISR) activities. Our solutions are also available for geo-spatial imagery interpretation/analysis, and medical imagery interpretation.

Scenario 1: Coast Guard Search & Rescue

The coast-guard has received a distress call from a fishing vessel off the coast of Alaska. Rough seas and poor visibility limit the identification of the life-raft reportedly occupied by the survivors of the vessel. The coast-guard runs Pegasus software running on an on-board laptop is monitoring the video feed from the helicopter SAR craft. It monitors the heavy seas and uses its sophisticated background substraction algorithms to mark likely Objects of Interest - possible liferafts, debris from vessels and humans.

  • "Real-time" search of OOI (life-raft) locates possible hits on-board in a live video stream
  • Algorithm self-adjusts for varying lighting and sea conditions
  • On-board assets alert coast-guard personnel in real-time


Scenario 2: Law Enforcement

A bank robbery has taken place in an urban area, Police are on the lookout for a suspect Red Sedan leaving the scene. Using feeds from a number of surveillance cameras, traffic cameras, helicopter cameras, dashboard cameras etc., Cognika's software create a list of possible egress routes ordered by likelihood. The process is initiated by using a sample of a similar vehicle as a query, and applying Cognika's temporal and spatial filtering algorithms to establish the likely egress routes. The data is augemnted by publicly available videos such as on Youtube taken by passers-by with cell-phone cameras. With a likely profile of the suspect vehicle, law enforcement also set an alert on their camera assets for the Object of Interest to be alerted by text, e-mail etc., anytime "similar" vehicles appears across multiple camera assets within the area of interest.

  • "Visual" search of OOI locates possible hits across available camera assets
  • Temporal Positional Localization Algorithms estimates likely routes of egress
  • Camera assets alert law-enforcement in real-time for visually "similar" vehicles


Scenario 3: Border Security

Customs and law-enforcement agencies are observing a large compound in a border area suspected of drug smuggling and human trafficking. The agencies have installed cameras to monitor activities within the compound. They also have aerial assets observing the compound to be alerted on any anomalous activity. This is a particularly challenging situation since the compound is in a heavily trafficked area with surrounding human & vehicular activity. The officers set a Region of Interest (ROI) to monitor just the compound, Cognika's machine-learning algorithms learn the "patterns-of-life" across the multiple video feeds (while ignoring the ambient "noise").The unblinking "stare" capabilities alerts the agencies to anomalous activities e.g. unusual number of humans present in compound, increase in vehicular activity etc. Furthermore, the agents are shown potential "hits" about the vehicles of interest from across their camera assets for them to build a profile.
Instead of assigning resources constantly to constantly watch the feeds, agents are alerted as-needed-when-needed, freeing them to focus more on analysis & interpretation.

  • Identifies "patterns-of-life" and sequences in the data
  • Filters out "noise" and focuses on critical data points
  • Real-time monitoring and alert systems


Scenario 4: Perimeter Security

A US government installation in a middle-eastern country is under threat from terrorist attacks - especially from Vehicle-borne Improvised Explosive Devices (VBIED). Intelligence has indicated the compound is being scoped by terrorist operatives looking for weak-spots to exploit. Multiple cameras are installed across the perimeter to provide a 360° surveillance of the surrounding activity. Constant human monitoring is especially challenging due to multiple human shifts, stress on the analysts, and the challenges in working with unpredictable and constantly adapting enemy.
Cognika's unblinking "stare" capabilities looks for vehicles or humans that appear repeatedly moving in an unusual fashion (e.g a vehicle stopping unusually and moving away quickly, or humans loitering with minimal activity). The system is self-adaptive to learn the varying "patterns-of-life", to alert human supervisors to such anomalies.

  • Learns patterns and sequences in the data
  • Adaptive Anomalous Activity Detection
  • Instant forensic look-back for OOI to view history


Scenario 5: WAMI Tracks

A route taken by NATO troops in Afghanistan, is the target of frequent IED strikes. The battle-field commanders suspect this is the work of a secretive IED network that operates with little/no signal "foot-prints" negating SIGINT efforts to break it. A UAV and Wide-area monitoring asset are purposed to observe the road and nearby villages. Humvee-mounted cameras as they drive-through villages, also offer additional data-points for monitoring. A WAMI (wide-area motion imagery) asset observes a suspicious "move-stop-move" event with a motorcycle and a small pick-up truck on the road. Before an interdiction team could reach the spot, the suspects have vanished, instantly, The analysts perform a "visual" monitoring of all compounds in the geographic area and get a hit from a UAV asset, showing a group of people gathering near a suspect compound. The network is apprehended following this lead.
Cognika's tools thus provide a complement of real-time and forensic analysis to be employed in conjunction for maximal, effective FMV exploitation.

  • WAMI data monitoring to detect activity
  • Forensic visual search of multiple camera assets
  • Works with multiple-cameras, formats and video types for augmented interpretation.