🔬 physics

Optimized Calibration of Terahertz Polarimetric Imaging Systems With Imperfect Polarizers for Accurate Jones-Matrix Mapping

This paper presents an optimized in situ calibration framework for terahertz polarimetric imaging systems that models imperfect wire-grid polarizers and determines optimal rotation angles to accurately recover sample Jones matrices using a minimal set of measurements, thereby reducing systematic errors and enabling faster, reliable field-deployable imaging.

Original authors: M. Hassan Arbab

Published 2026-07-08

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: M. Hassan Arbab

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you have a high-tech camera that doesn't just take pictures of how bright an object is, but also captures the "twist" and "orientation" of light waves hitting it. This is called Terahertz (THz) polarimetric imaging. It's like having a superpower to see the hidden internal structure of materials (like crystals or biological tissues) by analyzing how they twist light.

However, to get a clear picture, the camera itself needs to be perfectly tuned. If the camera's internal lenses or mirrors are slightly off, the picture will be distorted. This process of tuning is called calibration.

Here is the story of how the researchers fixed a specific problem with their camera, using simple analogies:

1. The Problem: The "Leaky" Sunglasses

To calibrate this camera, the researchers use a special tool: a Wire-Grid Polarizer (WGP). Think of this like a pair of high-tech sunglasses with very fine, parallel wires.

The Ideal: In a perfect world, these sunglasses would let light pass through if it's vibrating up-and-down, but block all light vibrating side-to-side.
The Reality: In the real world (especially with Terahertz waves), these "sunglasses" aren't perfect. They are "leaky." Some of the "side-to-side" light sneaks through the gaps in the wires.
The Consequence: If you tell the camera, "My sunglasses are perfect," but they are actually leaky, the camera will calculate the wrong information about the object it's looking at. It's like trying to measure the temperature with a broken thermometer; the result will be consistently wrong (a "systematic error").

2. The Solution: Acknowledging the Leak

Instead of pretending the sunglasses are perfect, the researchers built a new mathematical model that admits, "Yes, these sunglasses leak a little bit."

They measured exactly how much light leaks through and how the phase (the timing) of that light changes, depending on the frequency (color) of the Terahertz wave.
By feeding this "leaky" data into their calibration math, they could cancel out the error. It's like telling a GPS, "My car's speedometer is off by 5 mph," so the GPS can correct the route calculation automatically.

3. The Optimization: The "Three-Point" Trick

Usually, to calibrate a complex system, you might need to take measurements at many different angles (like turning the sunglasses 18 times). This takes a long time.

The Math Puzzle: The researchers realized that the math used to solve for the camera's settings is like a puzzle. Some ways of arranging the puzzle pieces (measurement angles) make the puzzle very sensitive to small mistakes (like a wobbly table). Other arrangements make the puzzle very stable.
The Condition Number: They used a mathematical concept called the "condition number" to find the most stable arrangement. Think of it as finding the most stable way to stack blocks so they don't fall over if you nudge them.
The Result: They discovered that you don't need 18 measurements. You only need three specific angles (spaced out evenly) to get a result that is nearly as accurate as the long 18-step process.
- Analogy: Imagine trying to guess the shape of a circle. You could measure it at 18 points, but if you pick the three perfect points, you can draw a nearly perfect circle with just three dots.

4. The Proof: The Crystal Test

To prove their new method worked, they tested it on a Lithium Niobate crystal.

The Test: This crystal has two different "optical axes" (like two different roads light can travel on). The researchers rotated the crystal and measured how it reflected light.
The Comparison:
- When they used the "old" method (pretending the polarizer was perfect), the results were wrong, especially at higher frequencies.
- When they used their new "leaky" method with the optimized three angles, the results matched the theoretical predictions perfectly.
The Imaging: They then stacked two different crystals (Sapphire and Lithium Niobate) and took a picture. Their method successfully mapped out the unique properties of each crystal, even though they didn't know exactly how the crystals were oriented beforehand.

Summary

The paper presents a smarter, faster way to calibrate a Terahertz imaging camera.

Stop ignoring reality: They stopped pretending their calibration tools were perfect and instead mathematically accounted for their flaws (leakage).
Work smarter, not harder: They found that by choosing three specific angles to measure, they could get the same high-quality results as taking 18 measurements, saving significant time.
The Result: This allows for faster, more accurate imaging of materials in the field, ensuring that the "pictures" of light polarization are true to life.

Technical Summary: Optimized Calibration of Terahertz Polarimetric Imaging Systems With Imperfect Polarizers

Problem Statement
Accurate terahertz (THz) polarimetric imaging requires robust system calibration to recover the Jones matrix of a sample. Existing calibration procedures for THz systems, such as the portable PHASR Scanner, often rely on wire-grid polarizers (WGPs) as reference components. However, in the broadband THz regime, commercially available WGPs are imperfect optical components. When the wire period is comparable to the operating wavelength, these polarizers exhibit frequency-dependent leakage of the polarization component parallel to the wires. Ignoring this "leakage" in calibration models introduces systematic errors that bias the recovered Jones matrix parameters. Furthermore, because calibration parameters are estimated from a finite set of angle-dependent measurements, the numerical conditioning of the resulting system of equations significantly affects how measurement noise propagates into calibration errors. Current methods often require extensive measurement sets (e.g., 18 angles) to mitigate these issues, which is inefficient for field-deployable scenarios.

Methodology
The authors propose an optimized in situ calibration framework for the THz PHASR Scanner that explicitly accounts for imperfect, leaky polarizers while minimizing the number of required measurements.

Leaky Polarizer Modeling: The calibrating WGP is modeled not as an ideal blocker but as a lossless, leaky element characterized by a frequency-dependent extinction coefficient ( $\eta$ ) and phase retardance ( $\delta$ ). These parameters were experimentally characterized using independent transmission measurements on free-standing tungsten WGPs.
In Situ Calibration Setup: A rotating WGP is placed in front of a reference mirror at the scanner's focal plane. The system records the reflected electric field as the WGP rotates, generating a known polarization transformation.
Mathematical Formulation: The authors derive two equivalent calibration formulations linking the measured electric field to the system's Jones matrix ( $J_2$ $J_{2}$ ) and incident field ( $E_1$ $E_{1}$ ).
- Formulation A: Based on an orthogonal harmonic-basis formulation, which allows for a Fourier series interpretation.
- Formulation B: An alternative matrix equation.
  Both formulations separate known angular coefficients from unknown calibration parameters.
Optimization via Condition Number: To reduce the measurement burden without sacrificing accuracy, the authors analyze the condition numbers ( $\kappa$ $κ$ ) of the system matrices. They determine that minimizing the condition number reduces the sensitivity of the solution to measurement noise.
- For Formulation A, minimizing $\kappa$ requires non-uniformly spaced angles.
- For Formulation B (the preferred formulation), minimizing $\kappa$ is achieved by selecting rotation angles that are uniformly spaced modulo $\pi$ (e.g., $\Delta\phi = 60^\circ$ ).
Sample Characterization: Once calibrated, the system extracts the sample's Jones matrix. For birefringent samples with unknown principal axes, the authors utilize eigendecomposition of the measured Jones matrix to recover intrinsic reflection coefficients along the crystal's optical axes, independent of the sample's orientation.

Key Contributions

Leaky Polarizer Model: The development of a calibration framework that incorporates the frequency-dependent leakage of WGPs, preventing systematic bias in the recovered Jones matrix.
Optimized Angle Selection: The derivation of an optimal set of three rotation angles (specifically $45^\circ, 113^\circ, 157^\circ$ or $23^\circ, 67^\circ, 135^\circ$ depending on the formulation, with the equidistant set being preferred) that minimizes the condition number of the calibration equations.
Minimal Measurement Set: Demonstration that a minimal set of three optimized angles yields results comparable to a full set of 18 angles, significantly reducing acquisition time.
Robust Eigendecomposition Method: A validated procedure to map the Jones matrices of birefringent samples with unknown orientations using only two sample rotation angles.

Results

Calibration Accuracy: Using the optimized three-angle set with the leaky polarizer model, the relative error in the calibrated incident field and system Jones matrix remained below 10% compared to a reference calibration using 18 angles.
Error Mitigation: Validation using a rotating x-cut lithium niobate ( $LiNbO_3$ ) crystal showed that neglecting WGP leakage (using a simple ideal model) resulted in systematic deviations from theoretical reflection coefficients, particularly at higher frequencies. Incorporating the leaky model eliminated this bias, producing results consistent with theoretical predictions.
Imaging Capability: The system successfully mapped the Jones matrices of a stack of birefringent crystals (sapphire on $LiNbO_3$ ). Eigendecomposition of the measured matrices accurately recovered the reflection coefficients along the ordinary and extraordinary axes, matching theoretical values despite the unknown principal axes of the samples.
Efficiency: The use of the ECOPS technique allows for rapid acquisition, with each THz image at a specific WGP angle acquired in less than one minute, making the three-angle calibration highly practical.

Significance
The paper claims that this optimized calibration technique enables faster and more reliable THz polarimetric imaging, particularly in field-deployable scenarios where minimizing the number of required calibration measurements is critical. By explicitly modeling the imperfections of standard wire-grid polarizers and mathematically optimizing the measurement angles, the method ensures high-fidelity Jones-matrix mapping without the need for ideal components or extensive measurement times. This approach addresses a specific limitation in broadband THz polarimetry where component non-ideality has previously been a source of uncorrected systematic error.

1. The Problem: The "Leaky" Sunglasses

2. The Solution: Acknowledging the Leak

3. The Optimization: The "Three-Point" Trick

4. The Proof: The Crystal Test

Summary

More like this