Key Laboratory of Light Field Manipulation and Information Acquisition, Ministry of Industry and Information Technology, and Shaanxi Key Laboratory of Optical Information Technology, School of Physical Science and Technology, Northwestern Polytechnical University, Xi’an 710129, China
Extended depth-of-field (DoF) imaging enhances longitudinal clarity but traditionally fails to record polarization information. Here, we propose a liquid-crystal Pancharatnam–Berry (PB) phase element that extends DoF while sensing polarization ellipticity. Leveraging cubic-phase-induced defocus-insensitive point spread functions, a 10 cm DoF is achieved in a 10-cm-focal-length 4f system. Linear PB phases enable spin-orbit interactions to detect polarization states, overcoming irreversible polarization loss in conventional methods. This dual-functionality system integrates DoF extension with polarization sensing, enhancing multidimensional imaging capabilities and holding promise for advancing machine vision, microscopic imaging, and optical detection by merging depth clarity with polarization-based object differentiation.
【AIGC One Sentence Reading】:A liquid-crystal PB phase element is proposed. It extends DoF to 10 cm in a 4f system and senses polarization ellipticity via spin-orbit interactions, overcoming polarization loss. This enhances multidimensional imaging for machine vision and optical detection.
【AIGC Short Abstract】:Traditional extended depth-of-field (DoF) imaging lacks polarization info recording. We introduce a liquid-crystal PB phase element that extends DoF and senses polarization ellipticity. Using cubic-phase-induced defocus-insensitive point spreads, a 10 cm DoF is realized in a 4f system. Linear PB phases facilitate spin-orbit interactions for polarization state detection, avoiding polarization loss. This system enhances multidimensional imaging, benefiting machine vision, microscopy, and optical detection.
Note: This section is automatically generated by AI . The website and platform operators shall not be liable for any commercial or legal consequences arising from your use of AI generated content on this website. Please be aware of this.
Optical imaging with a large depth of field (DoF) has important applications in microscopic imaging, photography, machine vision, and satellite remote sensing[1–3]. Traditional approaches for extending DoF typically involve reducing the aperture size or employing defocused image fusion[4]. The former method significantly compromises system light throughput and degrades imaging resolution, while the latter demands extensive post-image processing, rendering it impractical for dynamic scenes. To address these limitations, on the one hand, focus engineering using specially designed phases emerges as a promising technique; structured focuses with Airy spots, optical needles, and multiple focal spots have been proposed to generate defocus-insensitive point spread functions (PSFs)[5–12]. On the other hand, introducing scattering media into optical systems provides a new strategy. Due to the memory effect of the scattering media, the PSFs at different defocused positions are correlated, enabling image restoration using a stacked or modified PSF[13,14]. Among these methods, an Airy spot generated by a cubic phase greatly simplifies the design process compared to other focus engineering methods and single PSF measurements without being stacked or modified[5–8]. However, these methods can only capture the intensity information of the scene.
Since polarization is an important property of an object, decoding polarization properties of scattered light can reveal valuable information about texture, orientation, and even the constituent materials of an object[15], while polarization-multiplexed dimensionality enables novel application scenarios[16,17]. Polarization imaging systems are often based on focal plane division or time-domain multiplexing detection of polarization components[18–23]. Planar liquid crystal (LC) optical elements have excellent properties such as large-scale fabrication, dynamic tunability, and efficient modulation of light fields[24–34], and have been used for extending DoF systems[35–40]. The phase modulation of LC elements is closely related to polarization, and a series of novel phenomena of optical spin-orbit interaction can be induced using the Pancharatnam–Berry (PB) phase conjugation characteristics of orthogonal polarization states[41–48]. Thanks to these properties, LC elements are well-suited for the combined control of phase and polarization, enabling DoF extension and polarization imaging. The integration of extending DoF with polarization imaging brings about a wide range of applications across biomedical diagnostics, remote sensing, and industrial inspection domains[49–53].
Here, we propose an LC PB-phase element that integrates cubic phase and tilted phase for extending the DoF and sensing the polarization ellipticity. Owing to the odd symmetry of the cubic phase and tilted phase, the LC PB-phase element can produce a pair of defocus-insensitive PSFs with orthogonal circular polarizations to implement DoF extension with a factor of 20. Furthermore, polarization ellipticity can be obtained according to the Stokes vector from a single shot of two orthogonal polarized images, and thereby objects can be successfully recognized from an additional discrimination dimension beyond intensity profiles.
Sign up for Chinese Optics Letters TOC Get the latest issue of Advanced Photonics delivered right to you!Sign up now
2. Principle and Design
2.1. Principle of extending DoF and polarization imaging
As illustrated in Fig. 1(a), the traditional system can only generate a clear image of an object located at the front focal plane and cannot acquire polarization information. The defocus sensitivity of the system’s PSF causes it to degrade into a blurred disk when the object deviates from the focal plane, leading to significant deterioration in imaging quality. When adding an LC PB-phase element into the spatial spectrum plane, as shown in Fig. 1(b), it empowers the system to extend DoF and acquire the object’s polarization information. The underlying principle is illustrated in Fig. 1(c). Due to the spin-orbit interaction induced by the LC PB-phase element, objects located in different planes will present coded images convolved with two orthogonal circularly polarized PSFs in the image plane of the system. Because of the defocus-insensitive property of PSFs, two clear images with orthogonal spin states will be obtained by the Wiener deconvolution algorithm using the same PSF, and consequently, the DoF can be extended. The polarization ellipse information of the objects can be further obtained by the reconstructed spin component images.
Figure 1.Schematic diagram of extending DoF and polarization imaging. (a) Traditional 4f system. Only a clear intensity image of the object in the front focal plane can be obtained. (b) 4f system with embedded LC PB-phase element. (c) Reconstructions of objects and their polarization information. Two spin component images of objects are reconstructed by the Wiener deconvolution algorithm using a defocus-insensitive PSF, from which clear images and polarization ellipticities of objects located in different planes can be obtained.
To perform DoF extension and spin separation, two functional phase modulations are integrated into a single LC element. Limited by the phase conjugate locking of two spin components obtained from an LC PB-phase element, an odd symmetric cubic phase that can produce a non-diffractive Airy spot is used to achieve defocus-insensitive PSFs, of which the phase profile is where represents half of the aperture length and represents the attenuation factor used to control the non-diffractive characteristics and the shape of the focal spot[54]. To achieve polarization imaging, a tilted phase is used to split the left-handed circular polarization (LCP) and right-handed circular polarization (RCP) PSFs. Because of the odd symmetries of these two phases and the conjugate feature of the PB phase, this LC element can transform the incident LCP and RCP components into opposite polarization states with spatially separated Airy spots, which enables the combination of extending DoF and polarization imaging.
3. Results
The LC element was integrated into the spatial spectrum plane of a system comprising two lenses with a 10 cm focal length and a 25.4 mm aperture diameter to validate its dual functionality. For this system, we selected and to ensure a defocus-insensitive PSF across a spatial range. The wavelength and for splitting LCP and RCP were further optimized to achieve a suitable field of view. The LC element is shown in Fig. 2(a) with its local polarized micrograph in Fig. 2(b), fabricated using sulfonic azo-dye SD1-based photoalignment technology, and a digital micro-mirror device (DMD, 1080 pixel × 1080 pixel with each pixel size of ) based dynamic exposure process. When adding the LC PB-phase element to the spatial spectrum plane of the system, as illustrated in Fig. 2(c), light from a point source located at the front focal plane traverses this optical setup; spin-dependent PSF pairs are generated, exhibiting minimal variation as the point source shifts across different depths. For PSF measurement, a 50 µm diameter pinhole was employed to approximate a point source, illuminated by a 620 nm light-emitting diode (LED) with an optical bandwidth of 10 nm. As demonstrated in Fig. 2(d), the PSFs for LCP and RCP incident lights remained remarkably consistent across point source distances ranging from to 5 cm wherein the correlation of the PSFs exceeded 0.9 as shown in Fig. 2(e), indicating that the same PSF can be effectively utilized for deconvolving the coded images. Furthermore, to demonstrate the robustness of our imaging system at different object distances and polarization states, we measured the imaging efficiency (IE) and polarization splitting ratio, where IE is defined by . The imaging efficiency remains stable around 41% across varying object distances shown in Fig. 2(f), while the polarization splitting ratio aligns with theoretical predictions in Fig. 2(g).
Figure 2.LC element design and characterization. (a) Photograph of the LC element. (b) Local polarized micrograph of the LC element. Scale bar: 1 mm. (c) Schematic diagram of PSF formation. When a point source passes through the system with the LC element, two spin-dependent PSFs are generated. (d) Experimentally measured PSFs for LCP and RCP components from z = −5 to 5 cm. (e) Correlation between PSF at z = 0 cm and PSFs from z = −5 to 5 cm. (f) Imaging efficiency (IE) of the system at different object distances. (g) Measured and theoretical polarization splitting ratios of the system.
We subsequently characterized this approach by imaging a 1951 USAF target (Thorlabs, R1DS1N, Newton, New Jersey, United States) to validate the DoF extension capability. The experimental setup is shown in Fig. 3(a), where an LC element is positioned in the spatial spectrum plane of a system. The raw image is shown in Fig. 3(b), displaying the LCP component in the upper right corner and the RCP component in the lower left corner. The polarization component associated with the PB phase is strongly dependent on the birefringent retardation of the LC molecules, which can be modulated by adjusting the applied voltage to control switchable activation and deactivation of the LC element[45]. Theoretically, we can modulate the voltage to the half-wave voltage, thereby suppressing the central zero-order image, which means we can achieve a larger field of view. However, due to the non-uniformity in the fabrication of the LC element, we only achieved a reduction in the zero-order image, but did not eliminate it completely. However, as the defocus increases, the intensity of the central image decreases, which means that we can still achieve a larger field of view. The first row of Fig. 3(c) displays the captured images when the device is deactivated, which rapidly blur as the focus deviates. In contrast, the second and third rows present the reconstructed images for the LCP and RCP components when the device is activated, demonstrating a significant extension of the DoF. The intensity profiles along the blue and red dotted lines in Fig. 3(c) are plotted in Fig. 3(d), where the blue (red) curve corresponds to the intensity distribution of the LCP (RCP) component under positive (negative) defocus conditions. Notably, we can clearly resolve the sixth element of the third group on the target, indicating a resolution of 35.08 µm for both LCP and RCP components, which indicates that the system achieves the maximum imaging resolution of the CCD camera without the need for additional microscopy components. Remarkably, a 20-fold DoF extension has been realized at this resolution, underscoring the exceptional capability for DoF extension in the system. Furthermore, leveraging the imaging properties of the system, the lateral magnification of the image remains unity, ensuring minimal distortion and maximal preservation of the spatial information.
Figure 3.Experiment for testing the capability of extending DoF. (a) Experimental setup for measurement. Imaging of a 1951 USAF target illuminated by LED light by the 4f system with added LC elements. L1, L2, L3, lenses; P, polarizer. (b) Raw image directly captured by CCD. Scale bar: 1 mm. The images in the upper right corner and lower left corner are the LCP and RCP images, respectively, and the center is the zero-order light. When the zero-order light becomes blurred, the two untreated images remain relatively clear. (c) Images when the LC element is turned off, as well as the deconvolved LCP and RCP component images. (d) Intensity distributions along the lines in (c).
The system can also record polarization information, thereby enhancing the functionality of traditional systems. We tested the polarization imaging capability by observing an LC sample with a designed polarization state shown in Fig. 4(a), and the polarized micrograph of an LC sample under LCP light is shown in Fig. 4(b). Reconstructed intensities of LCP and RCP components are shown in Fig. 4(c), and mapped distribution of the Stokes vector for this sample is calculated by . We can see that compared to the previous work that only extended the DoF, we recovered both the intensity and polarization information of the object, which provides a good basis for polarization imaging with extended DoF.
Figure 4.Experiment for polarization imaging with extended DoF. (a) LC sample with designed Stokes vector S3. (b) Micrograph of LC element under illumination of LCP light. (c) Reconstructed LCP and RCP images and Stokes vector S3 with different object distances.
Beyond the controlled sample tests, we further validated the system’s practicality by imaging real-world objects. In this part, we performed imaging experiments with a plastic ruler at , 0, 5 cm. As shown in Fig. 5(a), we imaged two areas, which have different polarization responses. Figures 5(b) and 5(c) are the raw images, which clearly demonstrate that the two parts have different polarization responses. Subsequent reconstruction of intensity profiles and Stokes vector shown in Figs. 5(d) and 5(e) enable unambiguous discrimination of these areas through combined intensity and polarization state analysis.
Figure 5.Real-world object measurement by imaging a plastic ruler. (a) Photo of the plastic ruler. (b), (c) Raw images of two areas directly captured by a CCD. (d), (e) Reconstructed images of LCP and RCP components and Stokes vector S3.
Notably, our approach enables multidimensional discrimination capability in DoF-extended imaging by integrating polarization information. To demonstrate this advantage, we observed two objects with different defocus positions in the same scene. As shown in Fig. 6(a), two samples with ‘flower’ and ‘snowflake’ were located at and 5 cm, respectively, both of which had different polarization distributions. The raw image is shown in Fig. 6(b), and the reconstructed intensity and Stokes vector are shown in Fig. 6(c). This polarization-driven contrast provides an extra discrimination dimension to intensity profiles, overcoming limitations of traditional DoF-extended systems that rely solely on spatial features. Such multispectral-polarimetric fusion establishes the potential for target recognition in DoF-extended imaging.
Figure 6.Multi-object target recognition integrating the polarization dimension. (a) Schematic diagram of the scene. (b) Raw image directly captured by a CCD. Scale bar: 1 mm. (c) Reconstructed images of LCP and RCP components and Stokes vector S3.
Table 1 presents a comparative analysis of our method’s performance relative to alternative approaches, evaluating four key parameters: DoF range, resolution, processing speed, and polarization imaging capability. Intuitively, our method achieves polarization imaging while maintaining comparable DoF and imaging capabilities to other approaches. Furthermore, since both the geometric phase element and its spin-dependent separation are wavelength-insensitive, this architecture is suitable for broadband applications across various wavelengths. However, dispersion effects and spin-splitting efficiency must also be considered. In practice, our system is capable of microscopic imaging, where implementing extended DoF polarized microscopy could significantly enhance pathological diagnostics by providing depth-invariant polarization contrast. The integration of deep learning[7,55–57] for image reconstruction could mitigate poor recovery caused by PSF distortion while accelerating processing speed. Although the system’s fixed unity magnification limits field scalability, incorporating front-mounted imaging optics can expand the operational depth range from centimeter to kilometer scales. This capability facilitates high-value target identification in extreme-depth scenarios, including aerial reconnaissance and geological surveying. Furthermore, leveraging the multi-band response of the LC PB-phase element[45] in conjunction with wavelength-specific filters enables multispectral imaging across various bands.
Table 1. Comparison of Different LC-Based Extended DoF Systems
Table 1. Comparison of Different LC-Based Extended DoF Systems
Method
DoF range
Resolution
Processing speed
Polarization imaging
Reference
Dual-focal microlens array
3.13–34.87 mm
—
Fast
✗
[35]
Image fusion
13–71 cm
—
Slow
✗
[36]
20–160 cm
—
Slow
✗
[37]
Cubic phase spatial light modulator (SLM)
Millimeter level
30 lp/mm
Middle
✗
[38]
Annular SLM
Millimeter level
13 lp/mm
Fast
✗
[39]
Axicon
27–50 cm
—
Fast
✗
[40]
Our work
5–15 cm
14.25 lp/mm
Middle
✓
In summary, we demonstrated an LC PB-phase element that synergistically enables DoF extension and polarization imaging. By leveraging cubic and tilted phase modulation, the system generates defocus-insensitive PSF pairs for orthogonal circular polarizations, maintaining cross-correlation coefficients exceeding 0.9 over a 10 cm axial range. Furthermore, we achieved multiplexed imaging of multi-plane objects, where polarization-state divergence provided an additional discrimination dimension beyond intensity profiles. This breakthrough addresses the critical challenge of irreversible polarization loss in conventional DoF-extended systems. The proposed LC PB-phase element has broad application prospects and will play an important role in advancing machine vision, microscopic imaging, and optical detection.
[55] L. Xu, J. S. J. Ren, C. Liu et al. Deep convolutional neural network for image deconvolution. Proceedings of the 28th International Conference on Neural Information Processing Systems - Volume 1, 1790(2014).
[56] S. Nah, T. H. Kim, K. M. Lee. Deep multi-scale convolutional neural network for dynamic scene deblurring. 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 257(2017).
[57] K. Zhang, W. Luo, Y. Zhong et al. Deblurring by realistic blurring. 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2734(2020).
Tools
Get Citation
Copy Citation Text
Xuanyu Wu, Xuanguang Wu, Xiaofeng Xu, Weiqi Chen, Bingyan Wei, Sheng Liu, Xuetao Gan, Jianlin Zhao, Peng Li, "Dual-mode liquid-crystal element for polarization imaging with an extended depth of field," Chin. Opt. Lett. 23, 100006 (2025)