A non-radiometric thermal camera tells you something is hotter than its surroundings; a radiometric (thermometry) camera tells you it is 87 °C — and that difference decides whether your inspection produces evidence or just pictures.
- Per-pixelCalibrated temperature data
- ±3 °C / ±3%Published LX-9A accuracy
- 0Firmware paths to radiometry
- At orderWhen you must decide
Key takeaways
- Radiometric adds per-pixel calibrated temperature, spot and region measurement, and configurable high/low temperature alarms.
- Detection missions — SAR, security patrol, wildlife, navigation — need contrast, not measurement. Skipping radiometry saves cost and simplifies operation.
- Measurement is mandatory wherever a number backs a maintenance or safety decision: utility reports, solar IEC classification, industrial trending, fire overhaul.
- Radiometry depends on factory calibration of the specific core. Firmware cannot add it later — choose at purchase.
On this page
What Radiometric Adds
The core difference is what each pixel stores. A non-radiometric camera records relative brightness: this pixel is warmer than that one, mapped to a display range that shifts as the scene changes. A radiometric camera records a calibrated temperature value per pixel, which survives into the file and can be interrogated afterwards.
That enables three things. Spot and region measurement — click anywhere in a recorded frame and read a temperature, including in post-processing on imagery captured weeks ago. Configurable high and low temperature alarms, which turn the payload into an automated attention system rather than something an operator has to watch continuously. And thermometry cores typically offer a high-gain range around −20 to 150 °C for ambient work and a low-gain range around 0 to 550 °C for equipment and fire applications.
The alarm capability is underrated. Painting only pixels above a set threshold converts a thermal feed into a screening tool that flags its own findings — see isotherms and temperature alarms for how that works in practice, and radiometric temperature measurement on UAV payloads for the underlying calibration chain.
When Contrast-Only Is Enough
Search and rescue, security patrol, wildlife observation, navigation aid: you need detection, not measurement. Nobody asks what temperature the missing hiker is. The question is whether there is a warm object in the treeline, and a non-radiometric camera answers it exactly as well as a radiometric one.
Skipping radiometry saves cost and simplifies operation. It also removes a class of operator error — a non-radiometric camera cannot report a wrong temperature, whereas a radiometric one in the hands of someone who has not set emissivity will do so confidently. For a patrol programme that will never issue a measurement, that is a genuine advantage rather than a consolation.
Missions in this category: SAR, night security patrol, wildlife monitoring, livestock checks and deer recovery. All of them are contrast problems.
When Measurement Is Mandatory
Anywhere a number backs a maintenance or safety decision. If your deliverable causes someone to spend money, shut down equipment or defer a repair, the finding needs a measurement behind it and the measurement needs traceable conditions.
| Mission | Radiometric needed? | Why |
|---|---|---|
| Search and rescue | No | Detection only — nobody measures the subject |
| Security and perimeter patrol | No | Presence, not temperature |
| Wildlife and livestock observation | No | Counting and locating, with health as a relative cue |
| Utility inspection reporting | Yes | Severity grading requires delta-T against comparable phases |
| Solar IEC-style classification | Yes | Standards define thresholds in degrees |
| Industrial trend monitoring | Yes | Comparison across time needs absolute values |
| Fire overhaul verification | Yes | A residual hotspot is defined by temperature, not brightness |
| Building envelope audit | Yes | Conditions and delta must be logged for the report to stand |
The pattern is consistent: detection missions do not need it, documentation missions do. Solar farm inspection and substation work sit firmly in the second group, and what a report should contain shows why.
UAVThermal Options
Thermal payloads including the MV-2P, LX-9B, OP-90A and OP-125A are offered in thermometry versions on request — specify radiometric when you quote. The published entries in the specification PDFs show how each model expresses it.
| Payload | Published temperature measurement entry |
|---|---|
| LX-9A | ±3 °C or ±3% of the reading |
| LX-9B | Optional — thermometry type |
| OP-125A | Optional — thermometry type |
| OP-90P | Optional — thermometry type |
| MV-2P | Supported via subsequent firmware updates |
Note that “optional — thermometry type” means a distinct factory configuration, not a menu setting. The OP-90A is likewise available in a thermometry version on request. When quoting, name radiometric explicitly rather than assuming the standard configuration includes it; the two units look identical and behave very differently.
Say “thermometry version” in the purchase order. This is the single most common and most expensive specification omission in thermal payload procurement. The correction after delivery is a new payload, not a firmware key.
How Accurate Is It, Honestly
The commonly quoted figure across the industry is ±2 °C or ±2% of the reading, whichever is greater, under controlled emissivity and distance. The published figure for the LX-9A is ±3 °C or ±3% of the reading. Either way, the qualifier matters more than the number: those tolerances assume you have set emissivity correctly for the material and accounted for distance.
Field accuracy depends on correct emissivity setting and standoff, and both degrade in aerial work. More atmosphere sits between sensor and target than in handheld use, the viewing angle is whatever the flight geometry gave you, and emissivity varies across a frame that may contain painted steel, bare metal and concrete simultaneously. Emissivity in drone thermography covers why shiny metal in particular will lie to your camera.
The practical consequence is not to distrust the numbers but to use them the way a thermographer does: comparatively. Grading one phase against the other two under identical conditions is robust; quoting an absolute temperature to two decimal places for a warranty claim is not. That is also why critical readings get verified at close range before high-cost decisions.
Why You Cannot Decide Later
Radiometry depends on factory calibration of the specific core — a characterisation of how that individual detector responds across its temperature range, stored on the unit. It is not a software feature being withheld; there is no calibration data to unlock.
That makes this an unusually consequential purchase decision. Most payload specifications can be worked around in the field or upgraded at the next refresh. This one cannot: a non-radiometric unit will never produce a defensible temperature, and discovering that when a client asks for one means buying the payload again.
The decision rule is simple. If there is any realistic prospect of the programme issuing reports that quote temperatures — utility, solar, building, industrial — buy the thermometry version even if today’s missions are detection-only. The premium is smaller than the replacement. If the programme is unambiguously detection work, save the money and the operator training. The five-year cost of a payload puts the trade in context.
Related reading
- Radiometric Temperature Measurement on UAV Payloads
- Shutters, Blackbodies and Two-Point Correction
- Emissivity in Drone Thermography
- Isotherms and Temperature Alarms: Automating the Find
- What a Thermal Inspection Report Should Actually Contain
- Drone Thermal vs Handheld Thermal Camera
- Annual Verification: When a Thermal Payload Needs Recalibration
- The Five-Year Cost of a Thermal Drone Payload
- MV-2P — 640×512 thermal at 130 g
- LX-9B — multi-sensor pod, thermometry optional
- OP-125A — 30x zoom, thermometry optional
- Solar Farm Thermal Inspection with UAV Payloads
FAQ
Can firmware upgrade a camera to radiometric later?
Generally no. Radiometry depends on factory calibration of the specific core — a stored characterisation of how that individual detector responds across its range. It is not a software feature being withheld, so there is nothing for a firmware update to unlock. Choose it at purchase; the correction afterwards is a new payload rather than a licence key.
How accurate are UAV radiometric readings?
The commonly quoted industry figure is ±2 °C or ±2% of the reading under controlled emissivity and distance; the published figure for the LX-9A is ±3 °C or ±3%. In both cases the qualifier does more work than the number. Field accuracy depends on setting emissivity correctly for the material and on standoff, and aerial work makes both harder than handheld use — more atmosphere, an uncontrolled viewing angle, and mixed materials in one frame.
Should a detection-only programme still buy thermometry?
Only if there is a realistic prospect of the work changing. Since the capability cannot be added later, the question is not “do I need it today” but “will I need it before this payload is replaced”. A public-safety or SAR programme that will never issue a temperature can skip it and gain a simpler operation. An inspection programme that currently only does screening should buy it, because the first client asking for a measured report otherwise costs a whole payload.
Questions about the technology? Talk to our engineers — we reply within 2 business days.