Characterization and Compensation of the Size-of-Source Effect in Thermography for Accurate Temperature Measurements

Thermography is a non-contact and non-invasive temperature measurement method that determines the distribution of the radiance emitted from the surface of a measurement object. Infrared cameras with special lenses and detectors are used for this purpose. The camera detectors initially provide uncalibrated digital values proportional to radiance (raw values), which are converted into temperature values using a calibration function derived from Planck's law.

As the parameters of the calibration function depend on the specific camera, each device is measured and calibrated by the manufacturer. In this process, a radiation source with a known temperature and defined geometry (e.g. a circle with a diameter of 1 inch) is used to record the detector's raw digital values in response to various defined temperature levels. This measurement data is then used to calculate the parameters of the calibration function.

Similar to camera technology in the visual wavelength range, the optical representation of objects in thermography can exhibit deviations. For example, the resolution of small structures (e.g. line grids) is limited by the optics and detector. Unlike in photography, where the absolute brightness or contrast often only plays a qualitative role, this optical limitation in thermography means that the measured value is directly dependent on the object size. This phenomenon therefore has a direct effect on the quantitative accuracy of the temperature values and is referred to as the "size-of-source effect" (SSE). This leads to one of the most significant systematic measurement deviations in infrared thermography and depends on the imaged size of the measurement object. The SSE can essentially be attributed to two causes:

• For small objects, the SSE is mainly associated with diffraction phenomena.
• For large objects, optical scattering phenomena are the main cause.

The SSE varies depending on the configuration of the camera (lens, detector and possibly optical filters) and is therefore a systematic error that is unique to each camera.

As part of a funded project together with InfraTec GmbH, a calibration procedure was developed that enables geometry-independent temperature measurements with thermal imaging cameras. Regardless of the camera configuration and the radiance temperature, the systematic measurement deviations caused by SSE were significantly reduced with the help of a data-based compensation filter. 
Building on this, this research project systematically records the SSE for a wide variety of camera configurations. A central focus is on the investigation of mathematical approaches for modeling the SSE and the resulting compensation methods. Since the properties of thermographic images (e.g. the noise-contrast ratio and the shape and size of thermal hot or cold spots) differ considerably, the applicability of the individual model approaches is largely determined by the respective scenario. It is therefore necessary to investigate the advantages and disadvantages of these models and the resulting compensation methods in order to determine their respective limits. In this way, the dependence of the measured temperature on the object size is to be minimized for any thermographic camera and thus a significant increase in the measurement accuracy of the absolute temperature measurement is to be achieved.
 

Miguel Méndez, M.Sc.

January 2025 - December 2027

1. Miguel Mendez; Jannik Ebert; Lars Sommerlade; Robert Schmoll; Andreas Kroll: Comparative Assessment of the Size-of-Source Effect in Middle and Long-Wavelength Infrared Cameras, QIRT Asia 2023 - The 4th Quantitative Infrared Thermography Conference, doi:10.21611/qirt.2023.16, 2023
2. Miguel Mendez; Robert Schmoll; Andreas Kroll: Rapid measurement of the Size-of-Source Effect by using an Iris aperture, Sensor and Measurement Science International Conference (SMSI) 2025, doi:10.5162/SMSI2025/D3.1, 2025
3. Jannik Ebert; Miguel Mendez; Robert Schmoll; Andreas Kroll: Assessing the Size-of-Source Effect in Thermography by Measuring the MTF with Consideration of Scattering Effects, 2025 IEEE International Instrumentation and Measurement Technology Conference (I2MTC), doi:10.1109/I2MTC62753.2025.11079201, 2025
4. Miguel Mendez; Jannik Ebert; Lars Sommerlade; Robert Schmoll; Andreas Kroll: The size of source effect for middle- and long-wavelength infrared cameras: Assessment by direct measurements and a modulation transfer function approach, Quantitative Infrared Thermography Journal, 1 - 18, doi:10.1080/17686733.2025.2589712, 2025
5. Miguel Mendez; Robert Schmoll; Andreas Kroll: High-Resolution Characterization of the Size-of-Source Effect via a Continuously Variable Aperture, Journal of Sensors and Sensor Systems (JSSS), submitted, 2025
6. Jannik Ebert; Miguel Mendez; Robert Schmoll; Andreas Kroll: Modeling the Size-of-Source Effect in Thermography Using a Measured MTF Extended by a Parametric Scattering Model, The 18th International Conference on Quantitative InfraRed Thermography, accepted, 2026
7. Miguel Mendez; Robert Schmoll; Andreas Kroll: Modeling the Size-of-Source Effect of Thermography Cameras: A Data-Driven PSF Approach, XL. Metrology Symposium of the Association of University Professors of Metrology (AHMT) - AHMT Symposium 2026, accepted, 2026