When a customer asks us “which module should I use,” the first question we always ask back is: what do you need to see, at what range, and in what package? The answer to those three questions almost completely determines the correct detector format — before you even touch spectral band selection or cooling requirements.
This article walks through the practical framework our engineering team uses to guide format selection for uncooled LWIR and cooled MWIR imaging modules.
Why Format Matters More Than Resolution Alone
Resolution figures — 640×512, 1280×1024 — tell you the pixel count, but they don’t tell you whether a target will be resolved at the required range. What matters is the combination of:
- Pixel pitch (μm) — physical size of each detector element
- Array format — total number of pixels
- Focal length of the lens paired with the detector
- Target size and range — what you need to detect, recognize, or identify
Two modules with the same 640×512 array but different pixel pitches (12 μm vs 17 μm) will have very different detector sizes, requiring different optics to achieve the same field of view — and different sensitivity characteristics.
Common Format Options and Their Use Cases
| Format | Pixel Pitch | Detector Size | Typical Application |
|---|---|---|---|
| 320×256 | 17 μm | 5.4×4.4 mm | Short-range handheld, low-cost UAV |
| 640×512 | 12 μm | 7.7×6.1 mm | Standard ISR UAV, vehicle, security |
| 640×512 | 17 μm | 10.9×8.7 mm | Legacy-compatible retrofits |
| 1280×1024 | 12 μm | 15.4×12.3 mm | Long-range surveillance, high-detail |
| 1280×960 | 10 μm | 12.8×9.6 mm | Compact high-res, next-gen payloads |
IRmodules offers formats from 640×512 (SPECTRA L06) through 1280×1024 (SPECTRA L12) in uncooled LWIR, and 640×512 (SPECTRA M06) through 1024×768 (SPECTRA H10) in cooled MWIR.
The Johnson Criteria: Translating Pixels to Real-World Performance
The Johnson Criteria define the minimum number of pixels required across a target to accomplish different recognition tasks:
- Detection (target present/absent): 1.5–2 pixels across the minimum target dimension
- Recognition (vehicle vs. person): 6–8 pixels across
- Identification (specific vehicle type): 12–16 pixels across
Worked Example
You need to detect a standing person (0.5 m wide) at 800 m range using a 640×512 module with a 25 mm lens.
- Instantaneous Field of View (IFOV) = pixel pitch / focal length = 12 μm / 25 mm = 0.48 mrad
- Pixel footprint at 800 m = 0.48 mrad × 800 m = 0.38 m/pixel
- Pixels across 0.5 m target = 0.5 / 0.38 = 1.3 pixels
That’s marginal for detection and insufficient for recognition. Options: move to a longer focal length (50 mm), use a higher-format module (1280×1024), or accept a closer engagement range.
SWaP Trade-offs
In constrained platforms — UAV gimbals, handheld devices, missile seekers — every millimeter and milliwatt counts. Larger format arrays draw more power and generate more heat that the detector assembly must handle.
For most UAV payloads in the 200–500 g class, a 640×512 module at 12 μm pitch strikes the best balance between detection capability and SWaP budget. Systems with relaxed weight constraints or long-range requirements benefit from moving to 1280×1024.
Checklist for Format Selection
- Define the minimum target size you need to detect (in meters)
- Define the maximum detection range (in meters)
- Calculate required IFOV: target size / (detection pixels × range)
- Select pixel pitch based on available focal lengths and module envelope
- Verify array format meets field-of-view requirements for your scenario
- Check total detector power budget against platform thermal design
If you’ve worked through these steps and still need help, IRmodules’ application engineering team can run a range performance model against your specific requirements. We support this for both standard module configurations and custom optical integrations.
