Limelight saves precious time during the build-season. Get tracking in one session with an ethernet cable, four screws, and a small amount of code in any language. Limelight takes care of the rest. Several teams have reported setup times of less than one hour.
Introducing Limelight 3
Vision has never been easier.
Limelight is a plug-and-play smart camera purpose-built for the FIRST Robotics Competition. No experience is required - Limelight is easy enough for teams with no vision experience or expert mentors, and powerful enough for experienced teams who need a reliable, competition-ready vision solution.
The Perfect Intro to Deep Learning
Limelight supports Google Coral hardware acceleration for real-time object detection and image classification. Using real data, synthetic data, and a cloud GPU cluster, the Limelight team can quickly train neural networks for any application.
AprilTag Support
With its powerful quad-core cpu, Limelight tracks AprilTags in full 3D with ease.
Easy
Fast
Limelight tracks targets at 90 frames-per-second. Information is posted to network tables, so we support every programming language.
Versatile
Configure your vision pipelines with a simple web interface. Share your pipelines with other teams, and download updates as we improve the camera.
Limelight is also programmable - it supports full python scripting with openCV, numPy, SciPy, and more.
How Easy Is Limelight?
Q: How can I detect U-shaped targets?
A: Change the target "fullness" slider!
Q: How can I look for groups of targets rather than single targets? (2017, 2018 FRC Challenges)
A: Change the target "grouping" option!
Q: How can we aim our robot with an offset? Our shooter curves shots to the left.
A: Line up your shot by hand, and click the "calibrate" button!
Why?
Adding reliable, event-ready computer vision to an FRC robot is incredibly difficult. Limelight makes it incredibly easy, giving you time to focus on applying vision to improve robot performance.
Limelight combines over ten years of FRC vision experience to finally level the playing field.
Before Limelight
Choose a co-processor
Choose a camera
Setup Linux
Maintain your camera's exposure, white-balance, etc. after rebooting your robot
Mount a camera which probably lacks flat surfaces and mounting holes
3D-print a case for your co-processor
Mount LED rings around your camera
Wire your LEDs
Wire your camera and co-processor
Finally write your vision processing code
Write networking code (don't forget about blocked ports on FRC fields!)
Now that multiple members of your team have spent valuable time integrating vision, and the build-season's almost over, your programmers have mere hours to focus on actually using your vision information to guide your robot.
After Limelight
Mount your Limelight with the built-in mounting holes.
Run two wires to your PDP, and run an ethernet cable to your radio
Give your Limelight a team number, an IP address, and add a few lines of robot code.
Your robot now has targeting information streaming-in at 90 frames-per-second.
Limelight 2+ Stats
Tracking Speed: 90fps
Tracking Resolution: 320 x 240 pixels
Field-of-View: 59.6 x 49.7 degrees
Dimensions: 3.819 x 2.194 x 0.984 in (97.005 x 53.730 x 25.00 mm)
Weight: .25 lbs
Tracking Interfaces: Network Tables
Total latency (Photons -> Robot) : 21-25 milliseconds
Pipeline Latency: 3-7 milliseconds
NetworkTables -> Robot latency: .3 milliseconds
(NT limits bypassed to instantly submit targeting data.)
Luminous Flux: 400 lumens
60% more light than the standard dual-ring setup
Illuminance is increased by gloss-finish LED cones
Constant-brightness LEDs down to 7 volts
LL3 Stats coming soon.