TELEDYNE FLIR Boson® Advanced Radiometry

Official Publication Date: 7/14/2021

Official Expiration Date: N/A

Document Number: 102-2013-100-08 Rev 101

1
EAR Controlled: EAR99
These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
TELEDYNE FLIR Boson® Advanced Radiometry

Table of Contents
Document Number: 102-2013-100-08 Rev 101 .......................................................................................................... 1
1 Introduction ................................................................................................................................................................. 3
1.1 Revision History ................................................................................................................................................. 3
1.2 Reference Documents ...................................................................................................................................... 3
1.3 SCOPE ..................................................................................................................................................................... 3
2 Temp Stable Output .................................................................................................................................................. 3
2.1 Mechanical Design ............................................................................................................................................. 4
2.2 Radiometric Implications ............................................................................................................................... 6
3 Flux to Temperature conversion ......................................................................................................................... 6
3.1 Background .......................................................................................................................................................... 6
3.2 T-Linear ................................................................................................................................................................. 7
3.3 Application of Environmental Factors ...................................................................................................... 8
3.4 RBFO Flux (T-stable) to Temperature Conversion ............................................................................ 10
3.5 Window Characterization ........................................................................................................................... 13
3.5.1 Requirements .......................................................................................................................................... 13
3.5.2 Procedure .................................................................................................................................................. 14
3.5.3 Alternate procedure.............................................................................................................................. 15
4 Spot Meter ................................................................................................................................................................. 16
4.1 Spot Meter Accuracy ...................................................................................................................................... 16
5 Calibrating ................................................................................................................................................................. 17
5.1 One Point Correction- Temperature Linear Offset Adjust .............................................................. 17
5.2 Two Point Correction- Temperature Linear Output Adjust ........................................................... 18
5.3 Adjusting the Radiometry Model- Temperature Stable Output ................................................... 20
5.3.1 Scene Dynamic Range adjust- minor radiometry model change......................................... 20
5.4 Thermal Model................................................................................................................................................. 21
6 Isotherms ................................................................................................................................................................... 21
7 Radiometry Implementation Notes, Best Practices .................................................................................. 25
8 Accuracy ..................................................................................................................................................................... 25
8.1.1 Test Conditions ....................................................................................................................................... 25

2
EAR Controlled: EAR99
These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
TELEDYNE FLIR Boson® Advanced Radiometry

1 Introduction
1.1 Revision History
Version Date Comments

100 6/30/2021 Initial Release

101 7/14/2021 Minor edits for accuracy, consistency
1.2 Reference Documents
Ref Number Document number Comments

1 102-2013-43 Boson datasheet also 102-2013-40

2 102-2013-44 Boson Software IDD

1.3 SCOPE
This document will demonstrate the use of features specifically available for radiometry enabled
Boson cameras, excluding isotherms, which is also available for standard configurations with
limited functionality.

2 Temp Stable Output

The Radiometry enabled Boson variant is factory calibrated to output a stable 16-bit output over
temperature. This is accessible via the T-stable tap point, which is the same as Radiometry
enabled, T-Linear disabled 16-bit output.

3
EAR Controlled: EAR99
These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
TELEDYNE FLIR Boson® Advanced Radiometry

For radiometrically capable configurations, Boson provides three radiometric output modes:
■ Radiometry Disabled (T-linear Enable/Disable has no effect on output): 16b output varies with both scene flux
and camera temeprature.
■ Radiometry Enabled, T-linear Disabled: 16b output value is intended to be proportional to scene-flux
only and independent of the camera temperature.
■ Radiometry Enabled, T-linear Enabled: 16b output value is intended to be proportional to scene-
temperature.

2.1 Mechanical Design

Adequate heatsinking must be provided to prevent the Boson core from overheating, particularly
when operated in temperatures approaching the upper temperature range of the device. The rear
camera cover must be maintained at a temperature at or below 80C at all times. If the camera’s
internal core temperature exceeds its maximum safe temperature, Boson signals an overtemp
condition, and by factory-default, enters a low-power non-imaging state 10 seconds later.

4
EAR Controlled: EAR99
These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
TELEDYNE FLIR Boson® Advanced Radiometry

To the extent possible, Boson should be insulated from rapid thermal transients. Extreme thermal
shock will reduce the effectiveness of calibration and degrade the quality of the image. For best
results, the camera should be isolated from the thermal effects of window heaters, variable fans, and
other intermittent thermal loads. It is particularly important to minimize temperature gradients across
the camera, especially in the axes perpendicular to the optical line of sight. Avoid mounting conditions
which expose the camera to asymmetric heating from heaters, high-powered devices, and other
thermal loads. If convective cooling is required, airflow should be constant and as symmetric as
possible about the optical line of sight.

5
EAR Controlled: EAR99
These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
FLI R B oson Adv an ced Rad iom et ry
A PPL I CA TI ON NOTE

2.2 Radiometric Implications

The thermal characteristics of the camera and its environment influence the out of field
irradiance from the lens assembly, shutter assembly, and other internal components of the
camera. Out of field irradiance is spurious signal not representative of the scene being imaged.
Radiometric configurations of Boson undergo a factory-calibration process which includes
compensation for out-of-field irradiance. However, operating the camera in a method
substantially different than the configuration during calibration will degrade the quality of the
compensation. All of the thermal considerations described in Section 9.4 are important not just
for good image quality, but also for good radiometric accuracy. Consider the following for best
radiometric results:

1) Provide adequate heatsinking to keep camera temperatures well under the operational
maximum, preferably from the rear surface of the camera only.
2) Minimize conduction from any other camera surface other than the rear.
3) Minimize airflow across the camera assembly. If possible, utilize a non-insulative
enclosure.
4) Avoid embedding the camera in insulative materials (e.g. foam). Doing so affects the
self-heating and relative temperature of internal camera components.
5) If using an enclosure with a window, specify its parameters such that attenuation effects
and unwanted radiation reflected/emitted from it can be compensated using the
camera’s environmental factors correction feature in Figure 1.

3 Flux to Temperature conversion

Flux to temperature conversion can be done off-camera using the T-stable output in the
case of some applications, but generally it will be done on-camera, and the result of this
conversion is available via the T-linear output.
3.1 Background
When radiometric mode is enabled, the 14-bit video from Boson is stabilized and normalized
such that a scene with a given temperature will nominally correspond to a particular digital
value in the video stream, independent of Boson’s own temperature. The signal from the
camera, denoted as S, is called flux-linear because it is linear to the radiometric flux within

EAR Controlled: EAR99

These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
FLI R B oson Adv an ced Rad iom et ry
A PPL I CA TI ON NOTE
Boson’s spectral band. The flux-linear signal is related to scene temperature by the Planck
curve:
2
2hc 2 1
S = W (TK ) =   5
exp( ch ) −1
( ) • 
1
kTK ,

Equation 1

where S is the output signal, 1 and 2 define the spectral band, h is Planck’s constant, k is
Boltzmann’s constant, c is the velocity of light, R( ) is the camera responsivity (as a function of
wavelength), and TK is absolute temperature in units of Kelvin. Since Equation 1 is impractical
to calculate in software, the temperature-to-signal conversion is typically approximated by the
so-called RBFO equation:

R
S = W (TK ) = +O
exp( B / TK ) − F ,

Equation 2

where S is the output signal; R, B, F, and O are parameters generated during calibration; and TK
is absolute temperature. The conversion from flux to temperature is performed using the
inverse of Equation 2:

 R 
TK = B / ln  + F 
 SR − O .

Equation 3

In general, compensation for emissivity and other external parameters is performed before
applying flux-to-temperature conversion
3.2 T-Linear
The T-linear output mode of Boson takes the signal from the T-stable tap point and then
applies RBFO correction along with the environmental factor correction to produce an
output that is equal to the Temperature * 100.

EAR Controlled: EAR99

These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
FLI R B oson Adv an ced Rad iom et ry
A PPL I CA TI ON NOTE
3.3 Application of Environmental Factors
The external scene factors described in the Background section are configurable parameters in
the CCI. The parameters are used to calculate the scene temperature measurements provided
in the spotmeter feature and TLinear data output.

The parameters should be set according to the specifications of the final system and
application.

Power-On Telemetry Line

Parameter SDK Commands
Default Location
RAD Flux Linear Parameters (0.98, 295K, radiometrySetTempWindow() TBD
(ε, TBkg, win, Twin, atm, Tatm, 100%, 295K,
radiometrySetTransmissionWindow()
Rwin, Trefl) 8192, 295K, 0,
295K) radiometrySetReflectivityWindow()
radiometrySetTempWindowReflection()
radiometrySetTransmissionAtmosphere
()
radiometrySetTempAtmosphere()
radiometrySetEmissivityTarget()
radiometrySetTempBackground()

For best radiometric accuracy results it is recommended that instead of utilizing the
transmission values provided by the supplier, that the window transmission and window
temperature are characterized on a small sample set in the final system to be applied to all
systems using the same configuration. The window transmission should be characterized first,
then the window temperature at a single ambient temperature point at a minimum. Note that
the window temperature measurement error is high when the window transmission is very
high, but in this case the impact of this parameter is lower. The following section discusses the
procedure for characterizing the window parameters.

In use, the window temperature will vary as a function of the environment temperature. If it is
not practical to measure the window temperature in use, it can be estimated by using nearby
known temperatures such as that of the focal plane array (TFPA) plus an offset determined
experimentally. This constant offset may be measured at room temperature ambient point as
described in the following section, where TFPA(measured) is determined in Step 7 and
TWin(measured) is determined in Step 12:

EAR Controlled: EAR99

These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
FLI R B oson Adv an ced Rad iom et ry
A PPL I CA TI ON NOTE
Offset = TWin (measured) − TFPA (measured)

TWin (runtime) = TFPA (runtime) + Offse𝑡

Note that updating the window temperature in runtime based on ambient temperature
requires constant update of the TWin parameter in software via the CCI.

In the Boson App, they are available in the Radiometry Tab. The parameters are described as
follows:

▪ Window Temperature: The temperature

of the IR transmissive window in the
optical path of the camera
▪ Window Transmission: The transmission
of the IR transmissive window
▪ Window Reflection: The reflection of the
IR transmissive window.
▪ Window Reflection Temperature: The
reflection of the internal temperature on
the camera side of the window, off the IR
transmissive window.
▪ Atmospheric Transmission: Transmission
of the atmosphere in the optical path
between the camera and the target
▪ Atmospheric Temperature: Temperature
of the Air between the camera and the
target
▪ Emissivity Target: Emissivity of the target
being measured
▪ Background Temperature: Temperature of
Figure 1: Default Environmental Factors as shown in
the scene background being imaged. the Boson App
Generally, this will be near ambient.

EAR Controlled: EAR99

These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
 FLI R B oson Adv an ced Rad iom et ry
A PPL I CA TI ON NOTE
3.4 RBFO Flux (T-stable) to Temperature Conversion

Object of
interest

Window, Twin
Scene,
Tscene Radiometric
Atmosphe IR-camera
image
re, Tatm

𝜀𝑊(𝑇𝑠𝑐𝑒𝑛𝑒 ) + (1 − 𝜀)𝑊(𝑇𝑏𝑘𝑔 ) 𝜏𝑎𝑡𝑚⋅ (𝜏𝑎𝑡𝑚⋅ [𝜀𝑊(𝑇𝑠𝑐𝑒𝑛𝑒 ) + (1 − 𝜀)𝑊(𝑇𝑏𝑘𝑔 )] + (1 −

𝜏𝑎𝑡𝑚 )𝑊(𝑇𝑎𝑡𝑚 ))+rwinW(Trefl)+(1-τwin-rwin)W(Tatm)

𝜏𝑎𝑡𝑚⋅ [𝜀𝑊(𝑇𝑠𝑐𝑒𝑛𝑒 ) + (1 − 𝜀)𝑊(𝑇𝑏𝑘𝑔 )] + (1 − 𝜏𝑎𝑡𝑚 )𝑊(𝑇𝑎𝑡𝑚 )

Figure 2: Illustration of the radiation model

The model illustrated in Figure 2 shows the camera signal being a summation of:

1. signal from the scene

2. signal reflected by the background
3. signal emitted by the atmosphere
4. signal reflected by the window
5. signal emitted by the window

𝑊(𝑇) = 𝜏𝑎𝑡𝑚⋅ (𝜏𝑎𝑡𝑚⋅ [𝜀𝑊(𝑇𝑠𝑐𝑒𝑛𝑒 ) + (1 − 𝜀)𝑊(𝑇𝑏𝑘𝑔 )] + (1 − 𝜏𝑎𝑡𝑚 )𝑊(𝑇𝑎𝑡𝑚 ))+rwinW(Trefl)+(1-τwin-rwin)W(Tatm)

Equation 4

where the various parameters of equation 4 are described in Table 1.

EAR Controlled: EAR99

These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
FLI R B oson Adv an ced Rad iom et ry
A PPL I CA TI ON NOTE
Table 1: Description of the parameters of Equation 4.

Notation Description

S Value of the 16-bit digital video, in counts

 Emissivity of the scene.
win Transmission coefficient of the window
Twin Window temperature, in Kelvin
rwin Window reflection
Trefl Temperature reflected in the window, in Kelvin
atm Transmission coefficient of the atmosphere between the scene and the camera
Tatm Atmospheric temperature, in Kelvin
Tbkg Background temperature reflected by the scene, in Kelvin
Tscene Scene temperature, in Kelvin
W(T) Radiated flux (in counts) as function of the temperature of the radiating object

Prior to conversion of the camera signal to scene temperature, the unwanted signal (i.e., from
the background, atmosphere, window reflection, and window emission) must first be removed
and then the resulting difference scaled to compensate for attenuation factors. Re-arranging
Equation 4 to isolate the term W(Tscene) yields the following:
𝑆 (1 − 𝜀) (1 − 𝜏𝑎𝑡𝑚 ) 𝑟𝑤𝑖𝑛
𝑊(𝑇𝑠𝑐𝑒𝑛𝑒 ) = − ⋅ 𝑊(𝑇𝐵𝑘𝑔 ) − ⋅ 𝑊(𝑇𝑎𝑡𝑚 ) − 𝑊(𝑇𝑟𝑒𝑓𝑙 )
(𝜏𝑤𝑖𝑛 ⋅ 𝜏𝑎𝑡𝑚 ⋅ 𝜀) 𝜀 (𝜏𝑎𝑡𝑚 ⋅ 𝜀) (𝜏𝑤𝑖𝑛 ⋅ 𝜏𝑎𝑡𝑚 ⋅ 𝜀)

(1 − 𝜏𝑤𝑖𝑛 − 𝑟𝑤𝑖𝑛 )
− ⋅ 𝑊(𝑇𝑤𝑖𝑛 )
(𝜏𝑤𝑖𝑛 ⋅ 𝜏𝑎𝑡𝑚 ⋅ 𝜀)

Equation 5

All of the W(T) terms on the right-hand side of Equation 5 can be calculated via Equation 2.
After doing so, the right-hand side of Equation 5 is reduced to a single scalar value, which can
then be converted back to temperature via Equation 3. An example is shown below.

EAR Controlled: EAR99

These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
FLI R B oson Adv an ced Rad iom et ry
A PPL I CA TI ON NOTE
Example:

RBFO = (1673427.6586, 1548.009792, 1, 18491.9693)

• S = 24047 counts. (Note: without compensation for scene factors, this output value
from the camera would be converted to a value of 271K using the RBFO parameters
substituted into Equation 3.)
•  = 92%
• win = 
• Twin = 
• rwin = 
• Trefl = 
• atm = 
• Tatm = 
• TBkg = 

Calculations:
𝑅 1673428
𝑊(𝑇𝐵𝑘𝑔 ) = +𝑂 = + 18491 = 24903
𝑒𝑥𝑝( 𝐵/𝑇𝐵𝑘𝑔 ) − 𝐹 𝑒𝑥𝑝( 1548/278) − 1

𝑅 1673428
𝑊(𝑇𝑎𝑡𝑚 ) = +𝑂 = + 18491 = 24903
𝑒𝑥𝑝( 𝐵/𝑇𝑎𝑡𝑚 ) − 𝐹 𝑒𝑥𝑝( 1548/278) − 1

𝑅 1673428
𝑊(𝑇𝑟𝑒𝑓𝑙 ) = +𝑂 = + 18491 = 30867
𝑒𝑥𝑝( 𝐵/𝑇𝑟𝑒𝑓𝑙 ) − 𝐹 𝑒𝑥𝑝( 1548/315) − 1

1 0.08 0.15 0.03

𝑊(𝑇𝑠𝑐𝑒𝑛𝑒 ) = ⋅ 24047 − ⋅ 24903 − ⋅ 24903 − 30867
(0.97 ⋅ 0.85 ⋅ 0.92) 0.92 (0.85 ⋅ 0.92) (0.97 ⋅ 0.85 ⋅ 0.92)
−0

𝑊(𝑇𝑠𝑐𝑒𝑛𝑒 ) = 31701 − 2166 − 4777 − 1221 − 0 = 23537

EAR Controlled: EAR99

These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
FLI R B oson Adv an ced Rad iom et ry
A PPL I CA TI ON NOTE
𝐵
𝑇𝑠𝑐𝑒𝑛𝑒 = = 266.6𝐾
𝑅
𝑙𝑛 ( + 𝐹)
23537 − 𝑂

In this particular example, the difference between the temperature estimate after
compensating for scene parameters and the estimate prior to compensation is (271.0K –
266.6K) = 4.4K. Depending upon values for the various parameters, much larger estimation
errors are possible if compensation is not performed.

Note that if the RBFO curve is calibrated with the system’s protective window in place, as
recommended, then it is not required to correct for window transmission. That is, win is
considered to be 100% since the calibrated RBFO curve already compensates for window
transmission. Furthermore, if the window is located in close proximity to the Boson such that
the dominant source of reflection is the Boson camera-lens assembly itself, then it is not
essential to compensate for window reflection. If the scene temperature is high and window
transmission is <100%, the scene flux is replaced by window or Boson temperature so the
correction is needed even if the window is close to the Boson. Lastly, emission from the window
itself can generally be considered negligible since window emissivity is usually very low (i.e.,
100% – rwin – τwin is typically low) and window temperature not very different from camera
temperature. For example, the Si windows for FLIR One have 80% transmission and the 20% is
mostly absorption; if the camera isn’t calibrated with the window in place it has to be corrected
for. Assuming window effects can be ignored, Equation 5 reduces to the following:

𝑆 (1 − 𝜀) (1 − 𝜏𝑎𝑡𝑚 )
𝑊(𝑇𝑠𝑐𝑒𝑛𝑒 ) = − ⋅ 𝑊(𝑇𝐵𝑘𝑔 ) − ⋅ 𝑊(𝑇𝑎𝑡𝑚 )
(𝜏𝑎𝑡𝑚 ⋅ 𝜀) 𝜀 (𝜏𝑎𝑡𝑚 ⋅ 𝜀)

Equation 6

3.5 Window Characterization

3.5.1 Requirements
• Two black body sources with known temperature and emissivity, e.g. 20°C and 60°C
• Software interface to communicate with Boson
• Video interface to capture images
• Final enclosure with the window

EAR Controlled: EAR99

These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
FLI R B oson Adv an ced Rad iom et ry
A PPL I CA TI ON NOTE
3.5.2 Procedure
1) Power on the Boson within the final system (including the window).
2) Align the camera with the black body source such that the surface of the black body
subtends at a minimum the 10x10 center pixels. Ensure the black body is far enough
away to be in focus. If the black body source is relatively hot, this can cause undesired
camera heating affects, so a shield may be used until the data collection starts.
3) Ensure the black body source is stable in temperature. Ensure the camera temperature
is stable. For example, to verify the camera temperature is stable, read back the FPA
temperature via the bosonlookupFPATempDegKx10() SDK command multiple times over
the course of 10 seconds.
4) Change the Flat Field Correction (FFC) mode to “Manual” to ensure no FFCs occur during
the calibration via the bosonSetFFCMode() SDK command.
5) Perform an FFC via the bosonRunFFC() SDK command and wait at least one minute prior
to data collection.
6) Set the TLinear state to disabled via the TLinearSetControl() C-SDK command, or
alternatively just use the t-stable output in the GUI.
7) Capture a frame or record the mean of the center pixels. The mean output of the center
region of interest, ideally smaller than the black body (e.g. 6x6), is variable S1Win. Record
the FPA temperature (TFPA) in degrees Celsius at the time of this measurement (either in
the telemetry data or via query).
8) Move the camera to view the second, hotter black body source.
9) Capture a frame or record the mean of the center pixels. The mean output of the center
ROI is variable S2Win.
Note: Alternatively, both black body sources can be viewed simultaneously and the
captured frame analyzed with the appropriate region of interest for each of the S1Win and
S2Win measurements.
10) Remove the window and repeat the previous steps where the new measurements
without the window are variables S1NoWin and S2NoWin.
11) Calculate the average percent transmission reduction due to the window with the
following equation:

𝑆2𝑊𝑖𝑛 − 𝑆1𝑊𝑖𝑛
Window Transmission = 𝜏𝑊𝑖𝑛 = ɛ ∙
𝑆2𝑁𝑜𝑊𝑖𝑛 − 𝑆1𝑁𝑜𝑊𝑖𝑛

12) Calculate the window temperature with the following equations. The flux due to the
window temperature W(TWin) must first be calculated before converting to temperature
TWin with the RBFO parameters that were calibrated at the factory on a per-camera

EAR Controlled: EAR99

These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
FLI R B oson Adv an ced Rad iom et ry
A PPL I CA TI ON NOTE
basis; the RBFO parameters can be read from the camera software via the SDK
command radiometryGetRBFOHighGainFactory() and radiometryGetRBFOLowGainFactory().
The background temperature TBkg must be assumed and the corresponding flux W(TBkg)
calculated using the RBFO function. The expected blackbody signal W(TBB) must also be
calculated with the known black body temperature and the RBFO function.

S1𝑊𝑖𝑛 − τ𝑊𝑖𝑛 {ɛW(T𝐵𝐵 ) + (1 − ɛ)W(T𝐵𝑘𝑔 )}

W(TWin ) =
(1 − τ𝑊𝑖𝑛 )
 R 
TWin = B / ln  + F 
 W (TWin ) − O 
where
• ɛ = emissivity of the black body
• S1Win = measured signal against black body 1
• τWin = characterized window transmission
• W(TBB) = expected signal against the black body
• W(TBkg) = flux due to background temperature
• R, B, F, and O = camera specific constants readable via CCI

13) Write the characterized external scene parameters τWin and TWin to camera software via
the Environmental Compensation parameters.
14) Verify accuracy against both black body sources to confirm the newly characterized
parameters.

3.5.3 Alternate procedure

The simpler way to do this measurement would be to use the built-in functionality of the window
transmission environmental parameter. Imaging a blackbody from a set distance, measure the
temperature using the t-linear output while the camera is in thermal equilibrium. Ideally wait 30
minutes to reach this state, and put the camera in manual FFC mode to avoid any small jumps in
the radiometric output. Then, mount or hold the window infront of the camera FOV and while
changing the environmental parameter radiometrySetTransmissionWindow() to lower values
near where you expect transmission to be, attempt to match the windowless temperature
that was previously recorded. The number that matches the output of the camera pre and
post calibration should be the approximate transmission of the window. The blackbody
being measured would ideally be at a higher or lower temperature (+/-15C) than the
camera’s Tfpa.

EAR Controlled: EAR99

These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
FLI R B oson Adv an ced Rad iom et ry
A PPL I CA TI ON NOTE
4 Spot Meter
Prior to Release 3.0, Boson provided an Image Statistics feature, reporting average pixel value,
minimum pixel value, maximum pixel value, and standard deviation within a specified region of
interest (ROI). Beginning with Release 3.0, Boson additionally provides a separate Spot Meter
feature. Key differences between the Image Statistics feature and the Spot Meter feature are
as follows:

• Both features report mean, standard deviation, minimum value, and maximum value
upon command and via the telemetry line. The Spot Meter additionally reports the
location (x,y coordinates within the sensor) of the pixel with minimum value and the
pixel with maximum value.
• The Image Statistics feature collects its statistics after the Non-Uniformity Correction
block in the signal pipeline. The Spot Meter feature collects its statistics at the end of
the 16b pipeline. Consequently, for configurations which provide radiometric capability,
the Spot Meter is for the temperature-stable signal (if Radiometry is enabled and T-
linear conversion is disabled) or the temperature-linear signal (if Radiometry and T-
linear conversion are both enabled). In other words, units are reported in temperature
units (Kelvin x 100) if the 16b tap is configured for temperature-linear output.
• The Spot Meter mean value can be displayed on-screen as a numeric value or
thermometer-style gauge.
• The regions of interest (ROI) for both the Image Statistics and Spot-Meter feature are
independently configurable. For example, the Image Statistics feature can be
configured to collect statistics from the full array while the Spot Meter feature is
configured to collect statistics from a 16x16 pixel area in the center of the array.
• The Image Statistics mean value is the basis for gain-state switching

4.1 Spot Meter Accuracy

Spot Meter Accuracy represents a measure of the expected radiometric performance by
reporting an integer value based on a comparison between the as-calibrated environment
and the current camera conditions. A low value indicates more optimal camera conditions
and therefore will occur when camera conditions are such that the camera is able to
generate a more accurate temperature measurement.

EAR Controlled: EAR99

These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
FLI R B oson Adv an ced Rad iom et ry
A PPL I CA TI ON NOTE
Table 2: Spot Meter Accuracy look up table

Spot Meter Uncertainty Description

Accuracy Factor
Best
Scenario 1 Best-case conditions stated in Test Conditions 8.1.1
Good
Scenario 2 Steady-state conditions at ambient temperatures within the valid
operating temperature range but outside those of Scenario 1.
Stable
Scenario 3 Thermally dynamic conditions within the valid operating temperature
range with some allowance for a slow rate of temperature change.

Unstable
Scenario 4 Any condition outside of Scenario 3.
Disabled
Scenario 5 Radiometric data not relevant. Either radiometry or spot meter is
disabled.

5 Calibrating
The Boson radiometry model as shipped assumes the camera is used outside an enclosure,
and thermally well mounted as described in the Boson ICD using the tripod mount.
Variations in the camera mounting, use of an enclosure, or situations where there are
significant deviations from nominal imaging conditions, including use of a window in the
optical path, will require adjusting the camera radiometry model settings. In this section,
methods for fine-tuning the radiometric output are described to address errors associated
with variations in scene temperature, and not necessarily the thermal state of the camera
itself. If error significantly increases over the operating temperature of the camera, this is
most likely due to the mechanical mounting and the thermal environment of the camera,
and may not be addressed by these two methods.

5.1 One Point Correction- Temperature Linear Offset Adjust

The most coarse and simple adjustment to the T-linear output is to apply an offset
adjustment based on temperature error of the camera. If the camera runs 2 degrees high
consistently over the operational temperature range of the camera, a simple -2 degree
offset can be applied to address this error.

EAR Controlled: EAR99

These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
FLI R B oson Adv an ced Rad iom et ry
A PPL I CA TI ON NOTE
The commands adjust the Tlinear output only for the specified Gain mode, and they require
refreshing the T-linear LUT using the TLinearRefreshLUT command. HG refers to high gain
and low gain refers to low gain. Please review the datasheet if these terms are unclear.

Command Maximum correction

radiometrySetTemperatureOffset_HG() +/- 15 C
radiometrySetTemperatureOffset_LG() +/- 15 C

5.2 Two Point Correction- Temperature Linear Output Adjust

In this section, the Temperature Linear output is adjusted using at least two external
reference sources (black bodies). This leaves the radiometry model generating the
temperature table output unchanged and only modifies the conversion of temperature
stable output to temperature linear output.

This process creates modified Tlinear RBFO R and O terms and must be done for each gain
(high and low) separately. Using at least two black body sources sufficient to span the
region of interest, a single parameter (AdjustValue) is varied until the Tlinear output
matches the known scene temperatures. Typically, this value should only need to be
adjusted by less than +/- 20%. The adjustment pivots the camera Tlinear output around
the calibration camera ambient temperature. Example of how the Tlinear Output and
Tlinear Error changes are shown in Figure 3.

The RBFO R term is modified,

𝑅𝐹𝑎𝑐𝑡𝑜𝑟𝑦𝑁𝑒𝑤 = 𝑅𝐹𝑎𝑐𝑡𝑜𝑟𝑦𝑂𝑙𝑑 ∙ 𝐴𝑑𝑗𝑢𝑠𝑡𝑉𝑎𝑙𝑢𝑒

The RBFO O term is modified according to,

𝑅𝐶𝐴𝐿 ∙ (1 − 𝐴𝑑𝑗𝑢𝑠𝑡𝑉𝑎𝑙𝑢𝑒)
𝑂𝐹𝑎𝑐𝑡𝑜𝑟𝑦𝑁𝑒𝑤 = 𝑂𝐶𝐴𝐿 +
𝑒𝑥𝑝( 𝐵𝐹𝑎𝑐𝑡𝑜𝑟𝑦 /(𝑇𝑎𝑚𝑏𝑖𝑒𝑛𝑡 + 273.15) − 𝐹𝐹𝑎𝑐𝑡𝑜𝑟𝑦

Where the CAL subscript refers to the RBFO terms generated when the camera is near
enough to a block body during factory calibration to completely subtend the camera FOV
and 𝑇𝑎𝑚𝑏𝑖𝑒𝑛𝑡 is the ambient temperature during RBFO re-calibration. The CAL subscripted
terms can be calculated knowing the camera conversion factor .
𝑅𝐹𝑎𝑐𝑡𝑜𝑟𝑦
𝑅𝐶𝐴𝐿 =
𝛽

EAR Controlled: EAR99

These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
FLI R B oson Adv an ced Rad iom et ry
A PPL I CA TI ON NOTE
𝑅𝐶𝐴𝐿 ∙ (1 − 𝛽)
𝑂𝐶𝐴𝐿 = 𝑂𝐹𝑎𝑐𝑡𝑜𝑟𝑦 −
𝑒𝑥𝑝( 𝐵𝐹𝑎𝑐𝑡𝑜𝑟𝑦 /293.15) − 𝐹𝐹𝑎𝑐𝑡𝑜𝑟𝑦

Figure 3: Example Tlinear output (top left) and Tlinear Error (bottom left) before (blue) and after (orange) adjusting the
calibration using the Two-Point Correction technique outlined in Section 5.2. This process creates modified Tlinear RBFO R
and O terms and must be done for each gain state (high and low) separately. Using at least two black body sources sufficient
to span the scene temperature region of interest, a single parameter is varied until the Tlinear output matches the known
scene temperatures. The adjustment pivots the camera Tlinear output around the calibration camera ambient temperature
(22 C), shown as a vertical red line. The Tlinear Error, computed as the difference between the Tlinear camera output and
the known scene temperature is shown falling within ± 2 C (light blue shaded region) after the adjustment (orange data
points). The right column of plots shows the modified TLinear response and Tlinear error (orange) after additionally adding
a Tlinear output offset of -1.5 (radiometrySetTlinearOffset in the SDK).

EAR Controlled: EAR99

These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
FLI R B oson Adv an ced Rad iom et ry
A PPL I CA TI ON NOTE
5.3 Adjusting the Radiometry Model- Temperature Stable Output
In this section, the radiometry model used to convert raw camera flux to temperature
stable flux is described. Generally, adjusting the radiometry model requires calibrating the
final conversion factors from temperature stable to temperature linear output. This
process requires multiple thermal references. One exception to this general rule is
described in the next section.

5.3.1 Scene Dynamic Range adjust- minor radiometry model change

Adjusting the Scene Dynamic Range (SDR) using the radiometry model GG_SCALE
parameter can be done without completely having to calibrate the temperature stable to
linear conversion RBFO factors. Change RBFO internal mode to handle scaling the R and O
terms appropriately to increase SDR. This also results in the compression of the Tlinear
counts per degree.

Example:
Noise trade w/ compressing SDR → counts/degC changes
radiometrySetGG_Scale_HG()
radiometrySetGG_Scale_LG()
Note: Tlinear LUT refresh required if using linear output

It is recommended to start utilizing AUTO gain feature if high SDR is needed w/ large
GG_scale setting. Since camera sensitivity will be reduced, this will help the user to select
which gain mode is needed.

For example, assuming RBFO Scaled Mode is checked, for an RBFO set consisting of
108000, 1524, 1.0, 15500 respectively when GG_SCALE = 1.0, changing to GG_SCALE=0.8
results in the R and O terms changing internally to 86400 and 1219.2.

Rnew = gg_scale * Rold

Onew = gg_scale * Oold

There will be no change to the Factory RBFO R and O values.

GG_SCALE can be adjusted independently for low and high gain modes of operation. Noise
will generally increase with lower values of GG_SCALE.

EAR Controlled: EAR99

These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
FLI R B oson Adv an ced Rad iom et ry
A PPL I CA TI ON NOTE
5.4 Thermal Model
Boson employs the use of a thermal model to compensate for the effects of the thermal
state of the camera to measure absolute temperature more accurately. The modification of
the parameters that make up the thermal model are not currently supported and will be
more thoroughly explained in this document at a later release. For more information on
this please contact applications engineering support at Flir through your sales channel.
There may be solutions for certain customers to further improve radiometric accuracy
through doing further calibration and characterization of the final system enclosure.

6 Isotherms
Beginning with Release 3.0, Boson provides an optional Isotherm feature, with new options for
colorizing pixels at the post-colorization video tap. Figure 4 illustrates some of the
potential uses of the Isotherms feature.

(a)Colorize hot objects

(b) Colorize hot object

(c)Colorize hot and cold objects (d) Colorize mid-range objects

EAR Controlled: EAR99

These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
FLI R B oson Adv an ced Rad iom et ry
A PPL I CA TI ON NOTE

(f) Yellow-hot (monochrome) palette

(e)Colorize hot objects, green-hot for all else
Figure 4: Example Imagery with Various Isotherm Settings

The Isotherms feature allows the full 16b range of the camera to be divided into 5 regions,
and different colorization schemes can be applied to each region. The boundaries between
regions are specified as a percentage of the 16b range (e.g. 0% = 0 counts, 100% = 65535).
For those configurations which support radiometry, boundaries can alternatively be
specified in temperature units (as illustrated in the images of Figure 4).

While using isotherms, FLIR recommends using AGC linear Percent (Section Error! Reference s
ource not found.) at or above 90% or 0.90. This will avoid regions of the isotherm bands
being disrupted by non-linear grey bands that are using a histogram equalization to set
grey levels, and subsequently colors.

The Colorization mode of each Isotherm region can be independently selected from one of
the following five options, all of which are shown in Figure 6:
• Linear-RGB mode: post-colorization output of pixels in a region with mode set to
Linear RGB mode is linearly interpolated from two specified endpoint colors. For
example, if the lower boundary is set to 30C and assigned an RGB color value of
(240,240,0) and the upper boundary is set to 40C and assigned a color value of
(240,0,40), then a pixel imaging a 35C object will have an output color value of
(240,120,20).
• Linear-HSV mode: identical to Linear-RGB mode except that interpolation is
performed in Hue Saturation Value (HSV) color space rather than in RGB color
space. Use of this color space provides some color mappings which are not possible
in RGB. An example is shown in Figure 5.

EAR Controlled: EAR99

These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
FLI R B oson Adv an ced Rad iom et ry
A PPL I CA TI ON NOTE

(a) RGB interpolation (b) HSV interpolation

Figure 5: Example of RGB vs. HSV interpolation.

Notice the colors at the top and bottom are identical, but the interpolated colors in between are not.

• Standard mode: post-colorization output of pixels in a region set to Standard

mode appear the same with isotherms enabled and they do with isotherms disabled
if the same colorization LUT is chosen for each mode. That is, output value of each
pixel in a Standard region is based strictly on the AGC algorithm and colorization
LUT chosen.
• Non-Linear mode: post-colorization output of pixels in all such regions are
mapped non-linearly between colors, independently of the palette setting and the
colorization mode of other regions. This mode attempts to mimic the non-linear
AGC algorithm. A potential advantage of Non-Linear mode over Standard mode is
better utilization of available colors. In an image in which much of the image is in
one or more regions set to Linear RGB mode or Linear HSV mode, the regions set
to Standard mode may use only a small fraction of the grayshades, as illustrated in
Figure 6c. Conversely, the regions set to Non-Linear Gray mode use the full
spectrum of grayshades, as illustrated in Figure Figure 6d.
• Single-Color mode: post-colorization output of pixels in a region set to Single-
Color mode have the same luminance value as with isotherms disabled; however,
chrominance values are based on a single specified color. This mode readily
supports monochrome color palettes (e.g. red hot).

EAR Controlled: EAR99

These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
FLI R B oson Adv an ced Rad iom et ry
A PPL I CA TI ON NOTE

(a) Isotherms disabled, black-hot (b) Standard mode (hotter foreground)

palette and Single Color modes (cooler
background)

(c) Linear-HSV mode(hotter foreground) and (d) Linear-RGB mode(hotter foreground) and
Standard mode(cooler background) Non-Linear Gray mode (cooler background)

Figure 6: Examples of each Isotherm Mode

EAR Controlled: EAR99

These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
FLI R B oson Adv an ced Rad iom et ry
A PPL I CA TI ON NOTE
7 Radiometry Implementation Notes, Best Practices
When accessing the radiometric data stream via USB or CMOS, the only option for accessing
both AGC and Temperature data at the same time is to access both data streams
concurrently. It is possible to get USB video tap point independently from the CMOS video
mode options. The tap point used for USB video is chosen by using the
If this is not implemented as an option in the hardware configuration being used, an AGC
must be implemented to convert from 16-bit to 8 bit video.

8 Accuracy
Radiometric accuracy is dependent upon many variables, including the following:
• Target temperature
• Target emissivity and factors in the optical path affecting the signal, such as atmospheric
absorption.
• Target location in the field of view
• Target distance / subtense
• Camera temperature and temperature stability
• Change in temperature since previous FFC
• FFC type (internal vs. external)
• Mounting / installation considerations (see Section 2.2)

Consequently, accuracy requirements vary depending upon test conditions. Additional

radiometric optimization may need to be done to achieve peak accuracy in different system
level assemblies.

8.1.1 Test Conditions

For the test conditions defined below, radiometric accuracy shall be as follows:

• Steady-state operation (δTfpa/δt < 0.1C / min)

• Target is within the Central ROI of the FOV
• Tscene between 5C and 120C (high-gain state), 5C and 500C (low-gain state)
• TfpaK within 0.2C of value at previous FFC
• Camera heatsunk from the rear only (no other thermally conductive contacts to camera)
• No window, aperture, or other attenuating material (except air) between camera and
target, >99% atmospheric transmission in 8-14 micron spectral band.
• Blackbody emissivity > 98%

EAR Controlled: EAR99

These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021
FLI R B oson Adv an ced Rad iom et ry
A PPL I CA TI ON NOTE
Table 3: High-Gain Radiometric Accuracy

T Ambient [C]
High-Gain
-20 0 25 50
10 +/- 3 C +/- 3 C +/- 3 C +/- 5 C
T Scene
50 +/- 3 C +/- 3 C +/- 3 C +/- 5 C
[C]
100 +/- 4 C +/- 3 C +/- 3 C +/- 4 C

Table 4: Low-Gain Radiometric Accuracy

T Ambient [C]
Low-Gain
-20 0 25 50
30 +/- 4 C +/- 3 C +/- 3 C +/- 5 C
T Scene
50 +/- 4 C +/- 3 C +/- 3 C +/- 5 C
[C]
100 +/- 4 C +/- 3 C +/- 3 C +/- 4 C

Table 5: Low-Gain, High Scene Temperature Radiometric Accuracy

T Ambient [C]
Low-Gain
-20 0 25 50
T Scene 250 -- -- +/- 5 C --
[C] 500 -- -- +/- 10 C --

For technical support, visit the FLIR Support Center (web: [Link] or
request support through your sales channel.

EAR Controlled: EAR99

These commodities, technology or software are subject to the U.S. Export Administration Regulations (EAR).
Diversion contrary to U.S. law is prohibited.
TELEDYNE FLIR Proprietary-Confidential Copyright 2021
Doc. # 102-2013-100-08 Release 101, July 2021