IR Seeker Guidance System: FPS Challenges Explained
Explore the complexities of IR seeker guidance systems, focusing on why increasing frame rates on uncooled LWIR sensors isn't the solution. Learn about microbolometer physics, sensitivity issues, and how to optimize tracking with a focus on delay and stabilization.
9/29/20262 min read
1. A microbolometer is not a CMOS camera.
An uncooled LWIR pixel heats up and changes resistance. It needs time. If exposure time is too short, the reading is not clean. The image may look faster, but it is less true.
2. Faster frames reduce sensitivity.
Shorter exposure lowers SNR and raises NETD. For CUAS, this matters. A distant target is often not a bright fireball. It may be only a small contrast against sky, ground, clouds, or hot surfaces. If you chase fps and lose NETD, you may stop seeing the target before you can track it better.
3.The real mistake: mixing sensor rate and guidance rate.
IR seeker does not need every control decision from a fresh frame.
The sensor is a truth anchor. It can run at honest 30-60 Hz.
The guidance loop can run faster if a model estimates target state between frames.
What we did at SpearX:
-keep sensor at 30-60 Hz;
-use a motion model and state filter;
-use platform motion data;
-add limited optical flow in the ROI;
-correct the model when a new frame arrives.
A new frame is not the only truth. It confirms or corrects the prediction.
4. Delay matters more than pretty fps
At 300-350 m/s, a 60-90 ms loop means the target moves 18-31 m. If you aim at the visible pixel, you aim at the past.
We optimized predictable delay:
-move some preprocessing close to the sensor;
-run tracking and state estimation in parallel;
-do not block the loop on one frame;
-let the filter output state between frames;
-send predicted target state to the actuator.
5.Stabilization is part of the seeker
Low fps means longer exposure. Small platform jitter can blur the target. For LWIR with 0.5 mrad IFOV, blur can make the target a streak. Then tracking becomes unstable.
Key lesson:
On uncooled LWIR, you cannot beat physics by forcing higher fps. Build an architecture that 1.respects physics:
2.honest sensor rate;
3.faster guidance rate through prediction;
4.delay budget as a system parameter;
5.LOS stabilization inside the seeker;
6.tracking not dependent on one frame.
In next post: how to keep track when the target drops an IR flare, and why the brightest pixel is not always the target.
Part 1 of 2: Why 120 fps on uncooled LWIR is not the answer?
Jet Shahed-class targets can fly at 130-170 m/s. In a head-on intercept, closing speed can reach 300-350 m/s.
The first idea is:
raise camera fps to 90-120 Hz. On uncooled LWIR, this is not only costly. It fights microbolometer physics.
Related : GEOCOM Co. LLC www.geocomco.eu DeepTechRnD www.deeptechrnd.eu SpearXAgro www.spearxagro.eu
Contacts
SpearX ￪ Strike with Precision
+359 877 620 210
D-U-N-S® 525540791
© 2025 - 2026. All rights reserved. SpearX Projects. Geocom Co. LLC
SpearX is an EU-Ukrainian company located in a strategic areas for R&D in UAV