Single-Shot Fringe Projection Profilometry Based on LC-SLM Modulation and Polarization Multiplexing

Abstract

Fringe projection profilometry (FPP) is extensively utilized for the 3D measurement of various specimens. However, traditional FPP typically requires at least three phase-shifted fringe patterns to achieve a high-quality phase map. In this study, we introduce a single-shot FPP method based on common path polarization interferometry. In our method, the projected fringe pattern is created through the interference of two orthogonal circularly polarized light beams modulated by a liquid crystal spatial light modulator (LC-SLM). A polarization camera is employed to capture the reflected fringe pattern, enabling the simultaneous acquisition of four-step phase-shifting fringe patterns. The system benefits from advanced anti-vibration capabilities attributable to the common path self-interference optical path design. Furthermore, the utilization of a low-coherence LED light source results in reduced noise levels compared to a laser light source. The experimental results demonstrate that our proposed method can yield 3D measurement outcomes with high accuracy and efficiency.

1. Introduction

Structured light 3D measurement technology [1,2,3] plays a crucial role in fully recovering the 3D characteristics of objects while preserving their spatial information during the reconstruction process, which holds significant scientific value. Typically, this technology involves projecting a carrier-fringe pattern onto the object’s surface, capturing a deformed fringe pattern modulated by the object’s height from a different angle using an imaging device, and digitally demodulating and reconstructing a 3D digital image of the object based on the captured deformed fringe pattern [4,5,6]. In fringe projection profilometry (FPP), phase retrieval is a key process. Various single-shot methods based on integral transformations, such as Fourier Transform [7], Windowed Fourier Transform [8], and Wavelet Transform [9], have been proposed for phase calculation. However, transform-based approaches may suffer from low accuracy and subpar performance near discontinuities in phase maps [10]. Another single-shot phase calculation technique involves leveraging deep learning methods for phase recovery. Deep learning, with its powerful ability to comprehensively learn and automatically extract features, shows tremendous potential in effectively tackling the challenges faced in fringe projection [11]. C. Wang et al. [12] introduce a multi-scale feature fusion convolutional neural network (CNN) known as MSUNet++, which incorporates Discrete Wavelet Transform (DWT) in data preprocessing to capture high-frequency signals from fringe patterns as input for the network. This approach enhances the accuracy of reconstructing single-frame fringe patterns into 3D surfaces. S. Feng et al. [13] sequentially utilized two CNN networks to extract phase information from a single fringe image, pioneering the application of CNN networks in phase extraction. Creating a training dataset for fringe analysis can be challenging due to the requirement of preparing and scanning objects with diverse geometrical shapes. Furthermore, the applicability of a deep neural network may be limited to the specific system it was trained on, as different FPP systems can vary in optical configurations and camera–projector localizations, leading to generalization errors in supervised learning models. Although leveraging computer graphics (CG) technology [14] to create a virtual digital twin of the real world has expedited the process of generating training datasets, the dataset still needs to encompass complex and variable measurement environments. In general, deep learning’s powerful nonlinear fitting capabilities allow for the generation of high-quality phase maps using only a single fringe pattern with the assistance of deep learning techniques. However, the applicability of deep learning is somewhat restricted, which limits its broader usage across different scenarios.

To leverage the benefits of multi-shot phase-shifting techniques, researchers have explored wavelength or polarization multiplexing methods. In the realm of wavelength multiplexing, a color fringe projection approach has been introduced [15,16,17,18,19]. This method entails projecting sinusoidal colorful fringes onto an object’s surface using a projector. Subsequently, a color camera with red, green, and blue channels segregates the light with distinct phase shifts. While this strategy accelerates the phase-shifting process, the utilization of a color camera can lead to challenges associated with three-channel coupling, potentially diminishing measurement accuracy. In an alternative development, a polarization-based coded light illumination technique has been suggested to circumvent issues related to channel coupling. In this particular method, a polarization camera is employed to generate four phase-shifting fringe patterns with a single shot, effectively sidestepping channel coupling problems. Nonetheless, it is important to note that the optical path in this approach relies on a Michelson [20] or Mach–Zehnder [21] interferometer, which may compromise the system’s resilience to vibrations.

In this paper, a novel common path polarization interferometry-based fringe projection profilometry method is proposed. This approach involves utilizing a phase-only liquid crystal spatial light modulator (LC-SLM) to modulate the phase information of light, thereby enabling self-interference to generate fringes for projection. To address the nonlinear response of the LC-SLM, experimental corrections are applied through phase compensation. Consequently, a more precise 3D shape of the object is obtained post-phase unwrapping, achieved using the Transport of Intensity Equation (TIE) method [22]. Notably, the proposed method’s common path characteristic renders it unaffected by vibrations, enhancing its stability in practical applications.

This paper is structured as follows. In Section 2, we provide a concise overview of the optical system’s principles, including polarization phase-shifting and the phase calibration of the LC-SLM. Moving to Section 3, we present experimental results highlighting the importance of phase calibration for the LC-SLM. Additionally, we showcase successful measurements of two distinct specimens to demonstrate the efficacy of the proposed method. In Section 4, we present the phase error due to the depolarization by simulation. Lastly, Section 5 offers a conclusion to summarize the key findings and contributions of this paper.

2. Materials and Methods

2.1. Principle of Optical System

The polarization fringe projection profilometry (PFPP) system, as depicted in Figure 1, begins with unpolarized light emitted from a Light-Emitting Diode (LED). This light is then collimated by a lens (LO) and polarized at a 45° angle by a linear polarizer (P). The polarized beam is then reflected by a non-polarizing beam splitter (NPBS) toward a liquid crystal spatial light modulator (LC-SLM), where light in the 0° direction is not modulated and light in the 90° direction is modulated with a specific phase retardation ${\phi }_{SLM}$. These beams pass through a quarter-wave plate (QWP), transforming them into left- and right-handed circularly polarized light, respectively.

The polarized structured light, resulting from the interference between the modulated and unmodulated beams of the LC-SLM, is sensitive to the object’s height variations. The deformed fringe patterns generated by the object are captured by a polarization camera equipped with a micro-polarizer array (MPA), which features four different analyzer orientations (0°, 45°, 90°, and 135°) within adjacent $2\times 2$ pixel units, enabling synchronous phase-shifting.

Polarization phase-shifting [23,24,25] involves manipulating the polarization state of light, thereby altering the phase of the light and inducing phase-shifting. The LC-SLM is responsible for converting the linearly polarized light into two perpendicular linearly polarized lights, one of which is set with a specific phase retardation. The complex amplitude of these resultant orthogonal linearly polarized lights can be mathematically represented as follows:

$$\{\begin{array}{l}{E}_0\left(x,y\right)=A_0\mathit{exp}\left(i{\phi }_0\right)·\left[\begin{array}{c}1\\ 0\end{array}\right]\hfill \\ E_{90}\left(x,y\right)=A_{90}\mathit{exp}\left(i{\phi }_{90}\right)·\left[\begin{array}{c}0\\ 1\end{array}\right]\hfill \end{array}$$

(1)

In this context, the variables $A_0$, $A_{90}$, ${\phi }_0$, and ${\phi }_{90}$ represent the amplitude and phase modulation quantities of 0° and 90°, respectively, as modulated by the LC-SLM. Since the electric field in the 90° direction is not modulated by the LC-SLM, ${\phi }_{90}=0$. The amount of modulation of the electric field in the 0° direction by the LC-SLM is the phase loaded by the user, where ${\phi }_0\left(x,y\right)=0.1·x$. Given that the electric field directions of these two polarized light beams are orthogonal, no interference fringe is produced.

Assuming that the angle between the fast axis direction of the $QWP$ and the x-axis is 45°, its Jones matrix can be expressed as $G_{QWP}={\displaystyle \frac{1}{\sqrt{2}}}\left(\begin{array}{cc}1& -i\\ -i& 1\end{array}\right)$. After traversing the $QWP$, the complex amplitudes of the two light beams can be represented as follows:

$$\{\begin{array}{l}{E}_R\left(x,y\right)=G_{QWP}{E}_0={\displaystyle \frac{1}{\sqrt{2}}}{A}_{0}exp\left(i{\phi }_0\right)\left[\begin{array}{c}1\\ -i\end{array}\right]\hfill \\ E_L\left(x,y\right)=G_{QWP}{E}_{90}={\displaystyle \frac{1}{\sqrt{2}}}{A}_{90}exp\left(i{\phi }_{90}\right)\left[\begin{array}{c}-i\\ 1\end{array}\right]\hfill \end{array}$$

(2)

From Equation (2), it is evident that after modulation by the $QWP$, the two linearly polarized light beams transform into right- and left-handed circularly polarized light, respectively. The polarized structured light, comprising both left- and right-handed circularly polarized light, is modulated by the surface morphology of the object. After this modulation, the beam that passes through the MPA undergoes a polarization phase shift. Considering the channel with a polarization direction of 45° as an example, both the left- and right-handed circularly polarized light pass through it and convert into two beams of linearly polarized light with identical vibration directions. The complex amplitudes of these two light beams can be represented as follows:

$$\{\begin{array}{l}{E}_{R\_45}=G_{P\_45}{E}_R={\displaystyle \frac{1}{2\sqrt{2}}}{A}_{0}exp\left(i{\phi }_0\right)\left[\begin{array}{c}1-i\\ 1-i\end{array}\right]\hfill \\ E_{L\_45}=G_{P\_45}{E}_L={\displaystyle \frac{1}{2\sqrt{2}}}{A}_{90}exp\left(i{\phi }_{90}\right)\left[\begin{array}{c}1-i\\ 1-i\end{array}\right]\hfill \end{array}$$

(3)

The interference of these two beams of linearly polarized light results in a specific intensity distribution, which can be mathematically expressed as follows:

$$\begin{array}{cc}{I}_{45}\left(x,y\right)\hfill & =\left(E_{R_{45}}+E_{L_{45}}\right)·{\left(E_{R_{45}}+E_{L_{45}}\right)}^{*}={A_0}^2+{A_{90}}^2+2A_0{A}_{90}\cos \left({\phi }_{90}-{\phi }_0\right)\hfill \\ & ={A_0}^2+{A_{90}}^2+2A_0{A}_{90}\cos \left({\phi }_{SLM}\right)\hfill \end{array}$$

(4)

Herein, ${A_0}^2+{A_{90}}^2$ denotes the background intensity, $2A_0{A}_{90}$ signifies the modulation intensity, and ${\phi }_{SLM}$ represents the preset phase retardation of the LC-SLM. The variable ${\phi }_{SLM}$ in the aforementioned equation signifies the phase distribution prior to object modulation. Following modulation by the surface morphology, the light intensity distribution in this specific polarization direction can be mathematically expressed as follows:

$$I^{45}\left(x,y\right)={A_0}^2+{A_{90}}^2+2A_0{A}_{90}\cos \left({\phi }_{SLM}+{\phi }_O\right)$$

(5)

Herein, ${\phi }_O$ represents the phase modulation by the surface of the object. Similarly, after passing through the linear polarizer in the four directions of the MPA, the light intensity distribution of the interference field formed can be represented as

$$\{\begin{array}{c}\begin{array}{l}{I}^0\left(x,y\right)={A_0}^2+{A_{90}}^2+2A_0{A}_{90}\cos \left({\phi }_{SLM}+{\phi }_O-{\displaystyle \frac{\pi }{2}}\right)\hfill \\ I^{45}\left(x,y\right)={A_0}^2+{A_{90}}^2+2A_0{A}_{90}\cos \left({\phi }_{SLM}+{\phi }_O\right)\hfill \\ I^{90}\left(x,y\right)={A_0}^2+{A_{90}}^2+2A_0{A}_{90}\cos \left({\phi }_{SLM}+{\phi }_O+{\displaystyle \frac{\pi }{2}}\right)\hfill \end{array}\\ I^{135}\left(x,y\right)={A_0}^2+{A_{90}}^2+2A_0{A}_{90}\cos \left({\phi }_{SLM}+{\phi }_O+\pi \right)\end{array}$$

(6)

According to the four-step phase-shifting algorithm [26], the demodulated phase information can be expressed as

$$\phi \left(x,y\right)={\phi }_{SLM}+{\phi }_O=ta{n}^{-1}\left({\displaystyle \frac{I_{90}-I_0}{I_{135}-I_{45}}}\right)$$

(7)

2.2. Phase Calibration of LC-SLM

The gamma correction of LC-SLM is typically performed by averaging the phase response values in a specific region of its liquid crystal surface to determine the phase response at the current grayscale level. Subsequently, an inverse interpolation method is utilized to establish a linear relationship between the grayscale values and the phase response quantities. However, owing to the broad-spectrum nature of LEDs and the non-ideal characteristics of the LC-SLM itself, discrepancies still exist between the loaded phases in local regions and the phase response despite gamma correction. In LC-SLM modulation, the amount of phase modulation is related to the refractive index of the LC-SLM in the 0° and 90° directions with the following expression.

$${\phi }_{SLM}={\displaystyle \frac{2\pi }{\lambda }}{{\displaystyle \int }}_0^d\left(n_0\left(\theta \right)-n_{90}\right)dz$$

(8)

Herein, $d$ represents the thickness of the liquid crystal molecular layer. $n_0\left(\theta \right)$ represents the equivalent refractive index of LC-SLM in the 0° direction, and $n_{90}$ represents the equivalent refractive index of LC-SLM in the 90° direction. $n_0\left(\theta \right)$ can be expressed as

$${\displaystyle \frac{1}{{n^2}_0\left(\theta \right)}}={\displaystyle \frac{co{s}^2\theta }{{n^2}_0}}+{\displaystyle \frac{si{n}^2\theta }{{n^2}_{90}}}$$

(9)

$\theta$ represents the deflection angle of the LC-SLM liquid crystal molecules, which can be expressed as

$$\theta ={\displaystyle \frac{\pi }{2}}-2ta{n}^{-1}\left[exp\left(-{\displaystyle \frac{V_{rms}-V_C}{V_O}}\right)\right]$$

(10)

Herein, $V_{rms}$ is the operating voltage, $V_C$ is the threshold voltage, and $V_O$ is the overload voltage.

In practice, since LEDs are broadband light sources, the wavelength distribution in space is variable, which results in $n_0\left(\theta \right)$ being changed with spatial location. In addition, due to the manufacturing error of LC-SLM, for the same grayscale, the turning angle of the liquid crystal molecules at different positions is different, so Equation (8) can be rewritten as

$${\phi }_{SLM}={\displaystyle \frac{2\pi }{\lambda }}{{\displaystyle \int }}_0^d\left(n_0\left(\theta \left(x,y\right),\lambda \left(x,y\right)\right)-n_{90}\right)dz$$

(11)

In this study, we introduce a pixel-by-pixel phase correction technique aimed at further minimizing the discrepancy between the loaded phase and the phase response.

The principle of LC-SLM phase compensation is illustrated in Figure 2. Initially, the LC-SLM is loaded with the initial phase, and the polarization camera captures and reconstructs the true phase, depicted in Figure 2(a1). Subsequently, in Figure 2(b2), aligning area (A) with area (B) using the four marker points (identified by the green area representing the pre-designed marker points) enables the extraction of the loaded phase region (highlighted by the red dashed box area). Following this, in Figure 2(d2), the phase error is determined by unwrapping the wrapped phases and calculating the differences between the corresponding point phases, as depicted in Figure 2(c2).

$$\Delta \phi \left(x,y\right)={{\phi }^{\prime }}_{SLM}\left(x,y\right)-{{\phi }^{\prime }}_{POL}\left(x,y\right)$$

(12)

Herein, $x$, $y$ represent the horizontal and vertical pixel coordinates, respectively; ${{\phi }^{\prime }}_{SLM}$ and ${{\phi }^{\prime }}_{POL}$ represent the unfolding phases of ${\phi }_{SLM}$ and ${\phi }_{POL}$, respectively. As shown in Figure 2(e2), the corrected loaded phase is given by

$${{\phi }^{″}}_{SLM}\left(x,y\right)={\phi }_{SLM}\left(x,y\right)+\Delta \phi \left(x,y\right)$$

(13)

As shown in Figure 2, the blue dashed box portion represents the gamma correction of LC-SLM and the red dashed box portion represents the phase compensation of LC-SLM. Nonlinear correction involves the adjustment of the loading grayscale–voltage curve of the entire liquid crystal layer within the LC-SLM, aiming to constrain the phase response range uniformly. On the other hand, phase compensation is targeted at finely tuning the phase modulation of individual pixels within the LC-SLM to enhance precision. By implementing these processes in conjunction, a linear relationship is established between the loaded grayscale values and phase modulation across the entire system.

3. Results

3.1. Optical System

As illustrated in Figure 3, the experimental setup incorporates an LED (Daheng Optics, China, GCI-060401, operating at a wavelength of 620 $\mathrm{nm}$), a polarization camera (Baumer, German VCXU.2-50MP POL, USB3.0, featuring a pixel resolution of $2448\times 2048$ and a pixel size of $3.45 \mathsf{\mu }\mathrm{m}$), and an LC-SLM (UPOLabs, China, HDSLM80R, with a pixel size of $8.0 \mathsf{\mu }\mathrm{m}$). The collimating lens positioned behind the LED is a plano-convex lens with a focal length of $150 \mathrm{mm}$. Both lens 3 and lens 4 in front of the polarization camera have focal lengths of $150 \mathrm{mm}$, while lens 1 and lens 2 have focal lengths of $250 \mathrm{mm}$. The system’s field of view spans $7 \mathrm{mm}\times 8 \mathrm{mm}$, with an imaging magnification of 1.0×.

3.2. LC-SLM Nonlinear Correction Experiment

A set of 256 grayscale images, with grayscale values ranging from 0 to 255, are generated by the simulation and sequentially loaded into the LC-SLM. These images are then demodulated to yield the actual phase response. The average phase value is considered as the phase modulation at that particular grayscale, as depicted in Figure 4a. The phase obtained when the LC-SLM is loaded with the initial gamma value is compared with the ideal phase. An inverse linear interpolation method [27,28] is employed to derive the gamma curve that corresponds to the ideal phase, thereby achieving global nonlinear correction for the LC-SLM, as illustrated in Figure 4b.

The phase loaded to the LC-SLM in the experiment is a primary phase ${\phi }_{\mathrm{SLM}}\left(x,y\right)=-0.1·x$. After removing the carrier, the theoretically recovered phase is a constant. Due to the presence of errors, the phase distribution along the y-axis direction recovered before and after LC-SLM calibration is shown in Figure 5.

Due to factors such as the poor collimation of the LED, which is a broad-spectrum low-coherence light source, and the less-than-ideal angle of the quarter-wave slice, there is still some phase response error in LC-SLM after gamma correction. In practical measurements, the loaded carrier needs to be wrapped from $-\pi$ to $\pi$, which has a certain regularity, and therefore, the phase response error also has a certain regularity, which looks like water ripples, as shown in Figure 5a. Figure 5b represents the phase distribution after LC-SLM calibration, and due to the pixel-by-pixel phase correction, the phase response error is further corrected and the periodic phase error is significantly reduced.

Upon completion of the global nonlinear correction of the LC-SLM, a planar reflector is employed to facilitate phase compensation. This is achieved by loading an initial phase onto the LC-SLM and compensating for the initial phase to derive a new phase, in accordance with the method delineated in Figure 2. This new phase can be reloaded to accomplish the nonlinear correction of the LC-SLM. The morphological results of silicon wafers, both prior to and following the application of LC-SLM phase compensation, are depicted in Figure 6.

Figure 6a and Figure 6d showcase 3D images depicting the phase distribution before and after LC-SLM phase compensation, respectively. In contrast, Figure 6b and Figure 6e show 2D images illustrating the phase distribution before and after LC-SLM phase compensation, respectively. Lastly, Figure 6c,f specifically focus on the phase distribution within the red delineated area pre- and post-phase compensation. It is evident from the figure that phase compensation markedly enhances the measurement outcomes. The significant reduction in the fringe-like error following the phase compensation clearly demonstrates the necessity of this phase compensation process.

3.3. Sample Measurement

To demonstrate the effectiveness of the proposed method on real objects, measurements were conducted on a silicon wafer with a $50.0 \mathsf{\mu }\mathrm{m}$ deep groove and a metal plate with a circular pit.

The silicon wafer under study is depicted in Figure 7a. In the polarization interferometry setup, polarized interference fringes are created and then projected onto the wafer’s surface. By adjusting the wafer’s position, full-aperture projection is achieved. The projected polarized interference fringes are modulated across the wafer’s entire aperture. A polarization camera, positioned at a 37° angle to the projection optical axis, captures a single-frame deformed pattern modulated by the wafer, as shown in Figure 7b. This deformed fringe pattern includes pixelated four-step phase-shifting. Through extraction operations on adjacent $2\times 2$ pixel units, four phase-shifted fringe patterns (each with a 90° phase shift) can be separated and extracted from the single-frame deformed fringe image, as illustrated in Figure 7c. (1) through (4) in subfigure c represent phase-shifted fringes of the silicon wafer from 0 to $1.5\pi$, respectively, with intervals of $0.5\pi$. Using the four-step phase-shifting demodulation algorithm, a wrapped phase is obtained and shown in Figure 7d. Figure 7e represents the absolute phase distribution post-phase unwrapping. Additionally, the experimentally measured depth of the silicon wafer groove, as depicted in Figure 7f, is established to be $50.6 \mathsf{\mu }\mathrm{m}$ after undergoing phase-to-height mapping calibration. Notably, the measurement error stands at approximately $1.2\%$, indicating a small margin of error in the measurements.

In the experiment involving another specimen, illustrated in Figure 8a, a rough metal plate with a circular pit was examined. The original image captured by the polarization camera is displayed in Figure 8b. Figure 8c presents the four phase-shifted fringe patterns of the pit. (1) through (4) in subfigure c represent phase-shifted fringes of the metal circular pit from 0 to $1.5\pi$, respectively, with intervals of $0.5\pi$. The phase distribution on the pit is depicted in Figure 8d and Figure 8e, respectively. The depth distribution of the pit is shown in Figure 8f.

The deterioration in the quality of the wrapped phase of the metal plate compared to that of the silicon wafer, as observed in Figure 7d and Figure 8d, can be attributed to the presence of depolarization in the metal plate. When polarized light is reflected off a metal surface, the metal exhibits different absorption rates for p- and s-light. As a result, the circularly polarized light becomes elliptically polarized after reflection, impacting the intensity distribution following the polarization phase-shifting and consequently affecting the phase distribution. This phenomenon is crucial to consider when dealing with metal surfaces in interferometric measurements, especially in cases where depolarization can introduce inaccuracies in the measured results.

4. Discussion

The simulation conducted to demonstrate the depolarization effect provides insights into how a metal surface modifies the polarization state of incident light, consequently influencing the observed effects on phase and intensity distributions in subsequent measurements. Figure 9a visually represents how circularly polarized light transforms into elliptically polarized light upon reflection by a rough metal surface. In Equation (2), it is noted that the Jones matrix of left- and right-handed circularly polarized light, generated after passing through a quarter-wave plate, exhibits equal strength in both electric field directions. In fact, due to the different attenuation coefficients of the object for each direction of the orthogonal electric field, the Jones matrix of the left- and right-handed circularly polarized light after reflection by the object can be expressed as

$$\{\begin{array}{l}{E}_R\left(x,y\right)=G_{QWP}{E}_0={\displaystyle \frac{1}{\sqrt{2}}}{A}_0\exp \left(i{\phi }_0\right)\left[\begin{array}{c}1\times a\\ -i\times b\end{array}\right]\hfill \\ E_L\left(x,y\right)=G_{QWP}{E}_{90}={\displaystyle \frac{1}{\sqrt{2}}}{A}_{90}\exp \left(i{\phi }_{90}\right)\left[\begin{array}{c}-i\times a\\ 1\times b\end{array}\right]\hfill \end{array}.$$

(14)

Herein, $a$ and $b$ represent the attenuation coefficients of the material for the electric field in the $x$ and $y$ directions, respectively. For aluminum metal, when the angle of incidence of light is 37°, the values of a and b are 0.93 and 0.87, respectively. After simulation, the intensity distributions of the interference fringes for the four polarization directions are shown in Figure 9b.

In Figure 9c, a phase error is evident in the reconstructed phase, attributed to depolarization effects on the metal surface. Consequently, if the object under measurement lacks a sufficiently extensive planar surface for phase compensation, the phase error induced by depolarization cannot be fully rectified. This unresolved error can lead to inaccuracies in the water waveform phase, emphasizing the critical importance of addressing depolarization effects to ensure precise and reliable measurement results.

5. Conclusions

In conclusion, we introduce a polarization fringe projection profilometry (PFPP) method that facilitates the simultaneous execution of four-step phase-shifting within a single-frame fringe pattern. This approach primarily utilizes an LC-SLM to generate projected polarization self-interference fringes, which are then recorded by a polarization camera as modulated single-frame deformed fringe patterns. Consistent experimental findings validate the efficacy of polarization fringe projection in achieving synchronized phase-shifting demodulation of single-frame fringe patterns. This capability enables the precise reconstruction of fringe patterns captured within a single frame. By employing this synchronous phase-shifting technique, we can mitigate the impact of air disturbances and system vibrations encountered in traditional four-step phase-shifting methods, thereby enhancing the accuracy of phase demodulation. Moreover, this method exhibits promising prospects for applications in dynamic measurements.