Skip to content

captus perception

Automatic target recognition for drones

SEE IT IN ACTION

The perception layer for any drone 

Watch Captus Perception detect and identify objects across a scene, then let the operator designate a target for persistent tracking.

Lock on

Designate a target by selection, detection model, description, or image.

Hold on

Maintain tracking through motion, changing conditions, and occlusion.

Reacquire

Recover the target when it reappears after it’s lost from view.

use case

Put advanced perception on the drone you already fly

Captus Perception is onboard software that gives Group 1-3 drones intelligent perception and target tracking. Operators can designate a target through direct selection, a text description, an image, or an existing detection model. Captus locks on, remembers what that target looks like as it turns, moves, and changes, and picks it back up when it disappears. It runs on the compute already on your platform, with no cloud link.

WHAT IT ENABLES

Human control stays in the loop

Captus Perception works inside the operator’s existing workflow. The operator confirms the target before anything downstream acts on it. Captus keeps it in view, so the operator can focus on the mission instead of needing to keep eyes on the target.

Mission capability without the network

Tracking runs onboard. Once the operator designates a target, Captus holds it through jamming, GPS denial, and lost comms, with no datalink required.

Never lose track of a target

Captus remembers what the target looks like, not just where it was last seen. When the target passes behind cover or the drone looks away mid-mission, Captus picks the same target back up instead of making the operator start over.

the problem

ATR breaks when the target changes

Most automatic target recognition falls into a few familiar patterns.

  • Detection-only systems identify objects frame by frame but don’t remember which one the operator cares about. A new target type means going back to the lab.
  • Conventional trackers hold a pixel lock until the target turns, speeds up, or disappears, then lose it.
  • Platform-specific systems bundle perception with one airframe, sensor, or mission stack, so it can’t move to the next platform.

In the field, the target, the camera, the platform, and the network all change at once. That’s where most ATR breaks.

the solution

Find it. Hold it. Finish the mission.

FIND

Lock on to the target

FIX

Hold custody

FINISH

Act on it

  • Cue targets however you find it: automated detection, manual selection, a plain-language description, or a photo.
  • Start from one look: no new target library needed to track something you’ve just spotted.
  • Day or night: the same tracking across visible and thermal cameras.
  • Hold through change: Captus keeps up as the target turns, moves, and changes in light and angle.
  • Pick it back up: reacquires the target after it disappears behind cover, instead of making the operator start over.
  • Track multiple targets: separate tracks and IDs for several targets in the same scene.
  • Follow through the final approach: target memory updates as the target grows in frame and its appearance changes.
  • Hand targets across platforms: shared target memory gives another platform the information it needs to reacquire the same target.
  • Enable autonomy: move from operator-confirmed cueing to machine-directed tracking.

MISSION APPLICATION

Defending against drones? Meet Captus C-UAS.

Captus C-UAS uses the same lock-on and tracking core, with detection built for small drones. Find them sooner, hold track through clutter and occlusion, and feed your C-UAS stack.

Explore Captus C-UAS

field proven

Proven across platforms, environments, and missions

Army Project Linchpin (DEVCOM GVSC)

C-UAS detection, ATR, and multi-sensor fusion validated across EO, IR, RF, and SAR at the tactical edge.

Read the case study

Army Operation Jailbreak

Interoperability across autonomous systems and mission partners.

Learn more

Navy Project AMMO

Model update cycles reduced from six months to days in a disconnected maritime environment.

Read the case study

CDAO Swarm Forge Crucible 3

Iterative testing of AI-enabled capabilities with elite warfighting units and technology innovators.

SOCOM JATF

Selected as the perception solution for collaborative autonomy.

NATO SOFCOM C-UAS Bold Machina 26

Integrated detection and tracking with EO/IR sensors, C2, edge compute, and tactical networks from multiple partners in 48 hours.

Ready to get started?

Schedule a meeting with an AI expert today.