Infrared eye tracking device and eye tracking equipment

Abstract

This application provides an infrared eye-tracking device and an eye-tracking apparatus, relating to the field of eye-tracking technology. The device includes a single infrared light-emitting diode (LED), a collimating element, a metasurface device, an image sensor, and a control processing unit. The metasurface device is disposed in the light-emitting path of the collimating element and located between the infrared LED and the human eye's observation area. The metasurface device is a device designed based on a two-dimensional metasurface grating structure, configured to diffract and modulate the incident light beam into multiple spatially separated illumination beams to form multiple concentrically distributed light spots on the corneal surface of the target eyeball. The image sensor is used to acquire corneal reflection images containing multiple light spots. The control processing unit is communicatively connected to the image sensor and outputs gaze direction data based on the relative positional relationship between the multiple light spots in the corneal reflection image and the pupil center, thereby achieving higher precision and higher reliability eye-tracking performance while simplifying the hardware.

Description

Technical Field

[0001] This application relates to the field of eye-tracking technology, specifically to an infrared eye-tracking device and an eye-tracking equipment. Background Technology

[0002] In recent years, eye-tracking technology, as a key technology for perceiving human-computer interaction intentions, has demonstrated enormous application potential and value in fields such as Virtual Reality (VR), Augmented Reality (AR), and medical diagnosis. In near-eye display devices, achieving accurate and real-time gaze tracking is crucial for enhancing immersion, realizing foveated rendering, and enabling innovative interactions.

[0003] Currently, infrared eye-tracking technology based on the pupil-corneal reflection (Pulchin spot) principle is the mainstream solution. Traditional implementations typically rely on an optical system consisting of multiple infrared light-emitting diodes (LEDs) and a camera. Specifically, an illumination array of 6 to 12 infrared LEDs is arranged around the camera. These spatially separated point light sources create multiple discrete reflective spots on the corneal surface, and the gaze is calculated by analyzing the relative position changes of these spots with the center of the pupil. However, this approach has significant inherent drawbacks: a large number of system components, complex wiring, high overall power consumption, and a large physical footprint, which conflicts with the trend of VR / AR devices towards lightweight, compact, and long-lasting devices. To simplify the system architecture and reduce the number of infrared LEDs, related improvement technologies propose using diffractive optical elements to replace the multi-LED array. For example, multi-order phase gratings or Dammann gratings are used to split the beam emitted by a single infrared LED, thereby generating a regularly arranged array of spots on the image plane. While this approach alleviates hardware redundancy to some extent, it still has several drawbacks: higher-order diffraction efficiencies are generally low (usually below 60%), resulting in insufficient light energy utilization; the energy proportion of the zero-order direct beam (the unsplittered central spot) is too high (often exceeding 15%), creating strong background interference; furthermore, these gratings mostly produce regular dot matrix distributions, making it difficult to directly form a specific annular light field distribution that is more robust to eye movement. These drawbacks collectively lead to insufficient contrast and low signal-to-noise ratio of the effective corneal reflective signal spot, ultimately limiting the accuracy and reliability of eye tracking.

[0004] Therefore, there is an urgent need for a new eye-tracking solution that can reduce the number of light sources, enable a single light source to generate multiple high-quality ring-shaped illumination spots, reduce system power consumption and cost, and at the same time reduce the thickness of the eye-tracking module to achieve a lightweight and compact design for VR / AR devices. Summary of the Invention

[0005] This application provides an infrared eye-tracking device and an eye-tracking equipment to improve the problem that existing infrared eye-tracking solutions in the related technology are difficult to meet the requirements of lightweight and compact design of VR / AR devices in multiple key dimensions such as the number of devices, spot quality, system power consumption and size.

[0006] In a first aspect, this application provides an infrared eye-tracking device, comprising:

[0007] A single infrared LED is configured to emit an infrared illumination beam;

[0008] A collimating element is placed in the light-emitting path of an infrared LED to collimate and shape the infrared illumination beam.

[0009] A metasurface device is positioned on the light-emitting path of the collimating element and located between the infrared light-emitting diode and the human eye's observation area. The metasurface device is designed based on a two-dimensional metasurface grating structure and is configured to diffract and modulate the collimated and shaped incident beam into N spatially separated illumination beams to form N concentrically distributed light spots on the corneal surface of the target eye. Here, N is an even number not less than 2, and the position of each light spot on the corneal surface of the target eye corresponds to a different line-of-sight calibration feature point.

[0010] An image sensor, with its optical axis pointing towards the area observed by the human eye, is used to acquire corneal reflection images containing N light spots;

[0011] The control and processing unit is connected in communication with the image sensor and is used to output gaze direction data based on the relative positional relationship between N light spots in the corneal reflection image and the center of the pupil.

[0012] In one possible implementation, the metasurface device is a two-dimensional metasurface grating with a circular aperture. The arrangement of the microstructure array of the two-dimensional metasurface grating varies in different regions within the circular aperture: in the central circular region, the microstructures are arranged in a square lattice; in the annular region surrounding the central circular region, the microstructures are arranged in concentric rings. The diameter D_c of the central circular region satisfies the relationship 0.33≤D_c / D≤0.5 with respect to the aperture diameter D. The radial width W of the annular region satisfies the relationship 0.33≤W / D≤0.67 with respect to the aperture diameter D. The aperture diameter D is 2mm-4mm.

[0013] In one possible implementation, the two-dimensional metasurface grating includes a transparent substrate and a subwavelength microstructure array formed on the transparent substrate, wherein the subwavelength microstructure array is a nanopillar array; wherein the refractive index n of the nanopillar material at the wavelength of the infrared illumination beam satisfies n≥2.0, and the height H of the nanopillar satisfies 600nm≤H≤1000nm.

[0014] In one possible implementation, the subwavelength period of the subwavelength microstructure array is 550 nm to 650 nm.

[0015] In one possible implementation, the working surface of the two-dimensional metasurface grating is divided into multiple concentric annular modulation regions, each annular modulation region corresponding to the generation of a light spot; wherein, the first Phase distribution function of each ring modulation region satisfy:

[0016]

[0017] In the formula, Radial coordinates, It is the azimuth angle. This refers to the operating wavelength of an infrared LED. For the first The focusing distance corresponding to each light spot For the first Topological charge number of each ring modulation region and The first The inner and outer boundary radii of the ring modulation region.

[0018] In one possible implementation, the physical parameters of each nanopillar in the nanopillar array are designed according to the target phase value to be achieved at the location of each nanopillar; wherein, transmission phase modulation in the range of 0 to 2π is provided by changing the cross-sectional size of the nanopillar, and / or geometric phase modulation is provided by changing the rotational orientation angle of the nanopillar in the plane.

[0019] In one possible implementation, the cross-sectional dimensions of the nanopillars vary in the range of 300 nm to 550 nm to achieve transmission phase modulation coverage in the range of 0 to 2π; and / or, the rotational orientation angle of the nanopillars is optimized according to the target phase distribution and polarization error compensation requirements so that the difference in diffraction efficiency of the two-dimensional metasurface grating for transversely electrically polarized light and transversely magnetically polarized light is less than 5%.

[0020] In one possible implementation, the control processing unit includes: a microcontroller for generating drive signals to control an infrared LED to emit an infrared illumination beam; a memory for storing user eye-tracking parameter calibration data and an eye-tracking program; a digital signal processor, communicatively connected to the image sensor and the memory, for executing the eye-tracking program to process corneal reflection images, extract features of multiple light spots, and calculate the gaze direction based on the user eye-tracking parameter calibration data; and a communication interface, connected to the digital signal processor, for outputting the calculated gaze direction data to an external main processing device.

[0021] In one possible implementation, the infrared eye-tracking device further includes: a polarizer disposed in the optical path between the infrared light-emitting diode and the metasurface device; and a polarizing filter disposed in the optical path between the metasurface device and the image sensor; wherein the polarizer is used to make the light beam incident on the metasurface device linearly polarized, and the polarizing filter is used to polarize the reflected light from the eyeball after passing through the metasurface device.

[0022] In one possible implementation, the control processing unit is configured to enable a polarization anti-interference processing mode when the signal-to-noise ratio of the corneal reflection image is lower than a preset threshold or when a polarization interference environment is detected; wherein, the polarization anti-interference processing mode includes at least one of the following: boosting the gain signal of the image sensor, or applying a preset polarization compensation algorithm to process the corneal reflection image.

[0023] In one possible implementation, the infrared light-emitting diode, collimating element, and metasurface device are arranged sequentially along the same optical axis and integrated into a package, and the total thickness of the infrared eye-tracking device along the optical axis is less than 4 mm.

[0024] Secondly, this application provides an eye-tracking device that integrates an infrared eye-tracking device as described in any of the first aspects.

[0025] The infrared eye-tracking device and eye-tracking equipment provided in this application include: a single infrared light-emitting diode (LED) configured to emit an infrared illumination beam; a collimating element disposed on the light-emitting path of the infrared LED for collimating and shaping the infrared illumination beam; a metasurface device disposed on the light-emitting path of the collimating element and located between the infrared LED and the human eye observation area; the metasurface device is a device designed based on a two-dimensional metasurface grating structure, configured to diffract and modulate the collimated incident beam into N spatially separated illumination beams to form N concentrically distributed light spots on the corneal surface of the target eyeball; wherein N is an even number not less than 2, and the position of each light spot on the corneal surface of the target eyeball corresponds to a different gaze calibration feature point; an image sensor whose optical axis points to the human eye observation area for acquiring a corneal reflection image containing N light spots; and a control processing unit communicatively connected to the image sensor for outputting gaze direction data based on the relative positional relationship between the N light spots in the corneal reflection image and the pupil center.

[0026] This application, by employing a single infrared LED and metasurface devices, eliminates the redundant design of traditional solutions that rely on arrays of 6-12 infrared LEDs. This effectively eliminates the inherent drawbacks of multiple infrared LEDs, such as a large number of components, complex wiring, high power consumption, and large physical space occupation, achieving orders-of-magnitude optimization in size, power consumption, and cost for the eye-tracking module. Furthermore, by utilizing the subwavelength precision manipulation capabilities of metasurface devices, a single beam of infrared light is efficiently diffracted and modulated into N high-quality illumination spots distributed in a predetermined concentric ring shape in space. This effectively overcomes the inherent defects of solutions using multi-order gratings or Dammann gratings, such as low higher-order diffraction efficiency, strong interference from zero-order direct-emission light, and difficulty in generating specific ring-shaped light field distributions. The generated ring-shaped spots possess concentrated energy, uniform distribution, high contrast, and rotational symmetry, exhibiting excellent robustness to eye movement. They can stably provide multiple high signal-to-noise ratio corneal reflection feature points (Pulchin spots) over a large field of view, thus achieving higher precision and reliability in eye-tracking performance while greatly simplifying the hardware. Attached Figure Description

[0027] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0028] Figure 1 A schematic diagram of an infrared eye-tracking device provided as an exemplary embodiment of this application;

[0029] Figure 2 A schematic diagram illustrating the positional relationship between the infrared light-emitting diode and the metasurface device provided for an exemplary embodiment of this application;

[0030] Figure 3 A schematic diagram of the microstructure arrangement of a metasurface device provided as an exemplary embodiment of this application;

[0031] Figure 4 A schematic diagram of the device light path of the metasurface device provided in an exemplary embodiment of this application;

[0032] Figure 5 A schematic diagram of a microscopic part of a two-dimensional metasurface grating provided for an exemplary embodiment of this application;

[0033] Figure 6 A schematic diagram of the subwavelength period and nanopillar height of a metasurface microstructure array provided for an exemplary embodiment of this application;

[0034] Figure 7 A schematic diagram illustrating the division of the ring modulation region as provided in an exemplary embodiment of this application;

[0035] Figure 8A schematic diagram of the structure of the control processing unit provided in an exemplary embodiment of this application;

[0036] Figure 9 A schematic diagram of the eight-ring light spot distribution provided for an exemplary embodiment of this application;

[0037] Figure 10 A schematic diagram of the microstructure of a phase- and polarization-modulated nanopillar and its rotational orientation, provided for an exemplary embodiment of this application;

[0038] Figure 11 A schematic diagram of the structure of an eye-tracking device provided as an exemplary embodiment of this application;

[0039] Figure 12 A schematic diagram of an eye-tracking device provided as an exemplary embodiment of this application;

[0040] Figure 13 A schematic diagram of the distribution of annular light spots on the surface of the cornea provided in an exemplary embodiment of this application.

[0041] In the diagram, 10—infrared eye-tracking device; 11—infrared light-emitting diode; 12—collimating element; 13—metasurface device; 131—central two-dimensional metasurface grating; 132—annular two-dimensional metasurface grating; 14—image sensor; 15—control processing unit; 151—microcontroller; 152—memory; 153—digital signal processor; 154—communication interface; 110—eye-tracking device; 120—glasses; 121—cornea; 122—annular light spot; 123—visual axis; 124—exterior view.

[0042] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0043] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0044] The terms “first,” “second,” etc., used in the specification and claims of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented, for example, in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, products, or apparatus.

[0045] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with relevant laws, regulations and standards, and corresponding operation entry points are provided for users to choose to authorize or refuse.

[0046] First, some of the terms used in this application will be explained:

[0047] Metasurface: refers to an array of artificial microstructures with subwavelength thickness that can control the phase, amplitude, and polarization of electromagnetic waves.

[0048] Geometric phase (Pancharatnam-Berry phase): Polarization-dependent phase delay caused by gradual changes in the spatial orientation of microstructures, also known as PB phase.

[0049] Topological charge number: describes the vortex order of the spiral phase wavefront and determines the radius of the annular spot.

[0050] Purkinje images: Reflected images of a light source on the corneal surface, used for eye movement feature extraction.

[0051] TE / TM polarization: Transverse Electric / Magnetic, an orthogonal polarization state in which the electric / magnetic field direction is perpendicular to the incident plane.

[0052] Vector polarized light: A light beam in which the polarization state is not uniformly distributed at various points in space, such as radial polarization and angular polarization.

[0053] In related technologies, traditional multi-infrared LED array solutions, while providing sufficient corneal reflection feature points, suffer from inherent problems such as component redundancy, complex wiring, high power consumption, and large physical space occupation, which conflict with the trend of VR / AR devices towards lightweight, compactness, and long battery life. To simplify system architecture and reduce the number of infrared LEDs, related improvement technologies employ multi-order phase gratings or Dammann gratings to split the beam emitted by a single infrared LED, thereby generating a regularly arranged array of light spots on the image plane. While this alleviates the hardware redundancy problem to some extent, the generally low higher-order diffraction efficiency, excessively high proportion of zero-order direct-emission light energy, and inability to form a specific annular light field distribution result in insufficient contrast of the corneal reflection spot, affecting tracking accuracy. Furthermore, other integration and improvement attempts in existing technologies have failed to systematically overcome the aforementioned core contradictions and have exposed new technical shortcomings. For example, some solutions attempt to compress module size through complex optical path folding or waveguide structures, but while pursuing compactness, this often comes with significant optical efficiency losses, affecting the intensity of the final illumination spot and system energy efficiency. Other approaches attempt to reuse light sources and simplify detector layout through specific reflective optical path designs, but the number or distribution of effective light spots generated is still not optimal, and may not provide sufficient and robust feature information when the eye moves over a wide range, thus limiting the accuracy and stability of line-of-sight calculation.

[0054] To address the aforementioned issues, this application provides an infrared eye-tracking solution. By introducing a metasurface device between an infrared LED light source and the eyeball, combined with a single infrared LED and a collimating element, an extremely simplified optical path is constructed. The introduced metasurface device includes a two-dimensional metasurface grating structure. Through a dual mechanism of phase modulation and polarization selection using a subwavelength artificial microstructure array, and employing a composite design of geometric and transmission phase modulation, it can reconstruct the incident spherical wavefront into multiple coaxial annular vector beams. These beams converge on the corneal surface of the target eyeball, forming high-quality illumination spots distributed in a predetermined concentric ring pattern, serving as high-contrast corneal reflection feature points (Pulchin spots). An image sensor simultaneously acquires reflection images containing these spots, and a control processing unit calculates the gaze direction in real time based on the dynamic geometric relationship between the spots and the pupil center. This solution achieves both the spatial illumination function of multiple LEDs and the beam splitting and shaping function of diffraction gratings in traditional solutions using a single metasurface device. This not only significantly reduces the number of components, lowers system power consumption and physical size, but also achieves better spot quality, light energy utilization and environmental robustness through active light field modulation based on metasurface devices. Thus, while greatly simplifying the hardware, it lays a new physical foundation for high-precision and high-reliability real-time eye tracking.

[0055] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0056] Figure 1 A schematic diagram of an infrared eye-tracking device provided as an exemplary embodiment of this application. Figure 1 As shown, the infrared eye-tracking device 10 includes a single infrared light-emitting diode 11, a collimating element 12, a metasurface device 13, an image sensor 14, and a control processing unit 15; wherein:

[0057] A single infrared light-emitting diode 11 is configured to emit an infrared illumination beam;

[0058] The collimating element 12 is disposed on the light output path of the infrared light-emitting diode 11 and is used to collimate and shape the infrared illumination beam.

[0059] The metasurface device 13 is disposed on the light output path of the collimating element 12 and located between the infrared light-emitting diode 11 and the human eye observation area. The metasurface device 13 is a device designed based on a two-dimensional metasurface grating structure. It is configured to diffract and modulate the collimated and shaped incident beam into N spatially separated illumination beams to form N concentric ring-shaped light spots on the corneal surface of the target eye. Here, N is an even number not less than 2, and the position of each light spot on the corneal surface of the target eye corresponds to a different line-of-sight calibration feature point.

[0060] Image sensor 14, whose optical axis is pointed to the human eye observation area, is used to acquire corneal reflection images containing N light spots;

[0061] The control processing unit 15 is communicatively connected to the image sensor 14 and is used to output gaze direction data based on the relative positional relationship between N light spots in the corneal reflection image and the center of the pupil.

[0062] In this design, a single infrared LED 11 is the sole active illumination source, emitting infrared light of a specific wavelength (e.g., 940nm). Its emitting surface size is, for example, 1mm × 1mm, and its half-power angle is, for example, ±60°. The collimating element 12 is, for example, an aspherical lens with a focal length of... For example, 3mm, numerical aperture NA is for example 0.6, used to shape the divergent beam emitted by infrared LED 11 into a collimated or nearly collimated plane wave beam to improve the efficiency of subsequent optical processing, and beam diameter D is for example 3mm.

[0063] For example, Figure 2A schematic diagram illustrating the positional relationship between an infrared light-emitting diode and a metasurface device, provided as an exemplary embodiment of this application. (See diagram below.) Figure 2 As shown, the propagation path of the infrared illumination beam emitted by the infrared light-emitting diode 11 points towards the metasurface device 13. The metasurface device 13 is a planar optical element with a two-dimensional metasurface grating structure. A metasurface refers to a thin film composed of subwavelength scale (i.e., feature size smaller than the working wavelength) artificial microstructures (such as nanopillars) arranged in a specific pattern, which can precisely control the phase, amplitude, polarization and other properties of light waves. In this embodiment, the grating structure is designed to perform complex wavefront modulation on the incident collimated beam. Its core function is to efficiently diffract and "replicate" a single incident beam into N (N is an even number not less than 2, such as N being 4, 6, or 8) spatially separated illumination sub-beams. These sub-beams propagate and are ultimately projected onto the corneal surface of the target eye, forming N high-quality illumination spots distributed in a specific concentric ring pattern. These spots undergo specular reflection on the cornea, forming a series of bright image points called Purkinje spots. The spatial position of each Purkinje spot serves as a unique geometric feature point for subsequent line-of-sight calculation. The optical axis of the image sensor 14 (typically an infrared camera) is aligned with the human eye's observation area to synchronously capture corneal reflection images containing the aforementioned N ring-shaped Purkinje spots and the pupil at a certain frame rate (e.g., 120fps). The control processing unit 15 is the "brain" of the device, electrically connected to the image sensor 14, and is used to receive each frame of image output by the image sensor 14 and run the embedded eye-tracking algorithm. The core task of this algorithm is to accurately identify and locate the centroids of N light spots and the pupil center from the image; then, based on the dynamic geometric relationship between these feature points (e.g., the distance and angle of each light spot relative to the pupil center), combined with pre-stored user eye physiological parameter calibration data, the algorithm calculates the precise direction of the user's gaze in real time through a mathematical model (e.g., usually represented by screen coordinates or a three-dimensional spatial gaze point), and outputs this gaze direction data for use by the VR / AR main system or other external devices for interaction and rendering.

[0064] The infrared eye-tracking device provided in this application, by employing a single infrared light-emitting diode and metasurface devices, abandons the redundant design of relying on an array of 6-12 infrared light-emitting diodes in traditional solutions. This not only effectively eliminates the inherent defects of multiple infrared light-emitting diodes, such as a large number of components, complex wiring, high power consumption, and large physical space occupation, but also enables the eye-tracking module to achieve orders of magnitude optimization in terms of size, power consumption, and cost. In addition, by utilizing the subwavelength precision manipulation capability of metasurface devices to diffract and modulate a single beam of infrared light into N high-quality illumination spots distributed in a predetermined concentric ring shape in space, it effectively overcomes the inherent defects of low higher-order diffraction efficiency, strong interference of zero-order direct-emission light, and difficulty in generating a specific ring-shaped light field distribution in solutions using multi-order gratings or Dammann gratings. The generated ring-shaped light spot has the characteristics of concentrated energy, uniform distribution, high contrast and rotational symmetry. It has excellent robustness to eye movement and can stably provide multiple high signal-to-noise ratio corneal reflection feature points (Pulchin spots) within a large field of view. Thus, while greatly simplifying the hardware, it achieves higher precision and higher reliability of eye tracking performance.

[0065] In some embodiments, the metasurface device is a two-dimensional metasurface grating with a circular aperture. The arrangement of the microstructure array of the two-dimensional metasurface grating varies in different regions within the circular aperture: in the central circular region, the microstructures are arranged in a square lattice; in the annular region surrounding the central circular region, the microstructures are arranged in concentric rings. The diameter D_c of the central circular region and the aperture diameter D of the circular aperture satisfy the relationship: 0.33≤D_c / D≤0.5; the radial width W of the annular region and the aperture diameter D satisfy the relationship: 0.33≤W / D≤0.67; the aperture diameter D is 2mm-4mm.

[0066] For example, Figure 3 This is a schematic diagram of the microstructure arrangement of a metasurface device provided as an exemplary embodiment of this application. For example... Figure 3 As shown, within the entire circular aperture, the microstructure of the central circular region (schematically represented by a lattice in the figure), namely the central two-dimensional metasurface grating 131, adopts a square lattice arrangement, while the microstructure of the outer annular region (schematically represented by a ring lattice in the figure), namely the annular two-dimensional metasurface grating 132, adopts a concentric annular arrangement. The area ratio of the central circular region to the annular region is configured to ensure that the difference in spot energy uniformity across diffraction orders is less than 20%.

[0067] It should be noted that the above proportional relationship was determined through repeated optimization via optical simulation and experimental verification. Its purpose is to rationally allocate the phase modulation "responsibilities" and contributions of the central region and the annular region within a limited aperture area to generate multiple spatially separated annular illumination spots, thereby collaboratively achieving high uniformity of energy distribution for each diffraction order (corresponding to each annular spot) in the final output light field. Accordingly, Figure 4 A schematic diagram of the device light path of the metasurface device provided in an exemplary embodiment of this application. For example... Figure 4 As shown, a single incident beam is converted into multiple (6 in the illustration) beams that are separated in space.

[0068] In some embodiments, the two-dimensional metasurface grating includes a transparent substrate and a subwavelength microstructure array formed on the transparent substrate, wherein the subwavelength microstructure array is a nanopillar array; wherein the refractive index n of the nanopillar material at the wavelength of the infrared illumination beam satisfies n≥2.0, and the height H of the nanopillar satisfies 600nm≤H≤1000nm.

[0069] For example, Figure 5 A schematic diagram of a microscopic part of a two-dimensional metasurface grating provided for an exemplary embodiment of this application. (See diagram below.) Figure 5 As shown, the two-dimensional metasurface grating includes a transparent substrate and an array of subwavelength microstructures formed on the transparent substrate. The transparent substrate is made of fused silica, such as JGS1 grade fused silica with high optical homogeneity; the refractive index of the substrate at a working wavelength of 940 nm is approximately 1.45, and the thickness t0 is, for example, 500 μm. Correspondingly, the microstructure array is specifically an array of periodically arranged nanopillars. To achieve efficient optical phase modulation, the refractive index n of the material constituting the nanopillars should be no less than 2.0 at the working wavelength of the infrared light-emitting diode (e.g., 940 nm). A high refractive index contrast (relative to the substrate and air) is the physical basis for the microstructure to generate sufficient phase modulation. In specific implementations, amorphous silicon (a-Si), silicon nitride (…), etc., can be used. Materials such as gallium nitride (GaN) or other similar materials are used to meet this requirement. The height H of a single nanopillar along the optical axis is designed to be in the range of 600 nm to 1000 nm. This height range is on the same order of magnitude as the operating wavelength and is key to ensuring that the nanopillar can provide full-phase modulation from 0 to 2π. In one specific implementation, the height of the nanopillar is, for example, 800 nm.

[0070] In some embodiments, the subwavelength period of the subwavelength microstructure array is 550 nm to 650 nm.

[0071] For example, Figure 6 A schematic diagram of the subwavelength period and nanopillar height of a metasurface microstructure array provided as an exemplary embodiment of this application. (See diagram below.) Figure 6As shown in the figure, the arrangement of nanopillars in the subwavelength microstructure array is illustrated, and the spacing between adjacent nanopillars is the subwavelength period (labeled as period Λ). In this embodiment, the subwavelength period is set in the range of 550 nm to 650 nm. This design follows and ensures the basic principle of metasurface device operation: in order to suppress unnecessary higher-order diffraction and ensure that the incident light energy is mainly coupled to the designed diffraction order (i.e., the order used to form the ring-shaped light spot), the period of the microstructure must be smaller than the equivalent wavelength (λ / n) of the working wavelength in the substrate material.

[0072] For example, taking an operating wavelength λ = 940 nm and a substrate refractive index n ≈ 1.45 as an example, the equivalent wavelength is approximately 648 nm; by designing the subwavelength period Λ to be 600 nm, the condition Λ < λ / n is fully satisfied. This design allows the grating to operate in the so-called "near field" or "abnormal diffraction" region, thereby achieving high single diffraction order efficiency and effectively suppressing stray light. Therefore, limiting the subwavelength period to the range of 550 nm to 650 nm is one of the key structural constraints that, from a physical perspective, ensure that metasurface devices can accurately and efficiently convert a single beam of incident infrared light into multiple predetermined spatially distributed annular illumination spots with high light energy utilization and low optical crosstalk.

[0073] In some embodiments, the working surface of the two-dimensional metasurface grating is divided into multiple concentric annular modulation regions, each annular modulation region corresponding to the generation of a light spot; wherein, the first Phase distribution function of each ring modulation region satisfy:

[0074]

[0075] In the formula, Radial coordinates, It is the azimuth angle. This refers to the operating wavelength of an infrared LED. For the first The focusing distance corresponding to each light spot For the first Topological charge number of each ring modulation region and The first The inner and outer boundary radii of the ring modulation region.

[0076] For example, Figure 7 This is a schematic diagram illustrating the division of a ring modulation region as provided in an exemplary embodiment of this application. Figure 7 As shown, the working surface of the two-dimensional metasurface grating is divided into multiple concentric annular modulation regions, each region being an annular modulation region.

[0077] Specifically, each annular modulation region is assigned an independent composite phase modulation function to control the shape and position of its corresponding generated annular light spot. In the aforementioned phase distribution function, the operating wavelength of the infrared LED is... For example, 940nm; in an embodiment with 4 rings, it can be set as follows: =30mm, =35mm, =40mm, =45mm, , , , For a grating with an aperture of Φ=3mm, the partition boundary can be set as follows: , , , .

[0078] Based on the above design, collimated 940nm infrared light is incident on the metasurface. Each annular modulation region applies different phase modulations to the incident wavefront, forming a ring-Airy pattern in the far field through diffraction integration. Due to the topological charge of each ring... The radius of the generated ring-shaped light spot is different. With transmission distance satisfy: Four concentric annular spots are formed on the target corneal plane (z≈35mm), with radii of 2.1mm, 3.2mm, 4.3mm, and 5.4mm, respectively, and a ring width Δρ≈0.3mm. The angular spacing between the spots is >5°, and the peak intensity difference between the spots is less than 15%, with uniform distribution. These high-contrast, rotationally symmetrical annular spots are reflected by the cornea and captured by an image sensor (e.g., an infrared camera with a frame rate of 120fps and a resolution of 640×480). Due to their rotational symmetry, even if the eyeball rotates within a range of ±40°, at least two effective reflected spots can still be stably maintained in the image, and the signal-to-noise ratio (signal-to-background intensity ratio) of the Puerchin spot is higher than 10dB, thus providing high-quality image features for subsequent large field-of-view and highly robust gaze resolution.

[0079] In some embodiments, the physical parameters of each nanopillar in the nanopillar array are designed according to the target phase value to be achieved at the location of each nanopillar; wherein, transmission phase modulation in the range of 0 to 2π is provided by changing the cross-sectional size of the nanopillar, and / or geometric phase modulation is provided by changing the rotational orientation angle of the nanopillar in the plane.

[0080] Accordingly, to achieve the precise wavefront defined by the aforementioned partitioned phase distribution function, the physical parameters of each nanopillar in the nanopillar array constituting the two-dimensional metasurface grating need to be customized. The core design principle is that the structural parameters (size and orientation) of each nanopillar are uniquely determined by the target phase value required to be achieved at its spatial location (corresponding to specific coordinates (r, θ) within the annular modulation region). Specifically, local phase modulation in the range of 0 to 2π is provided through two independent or synergistic mechanisms: 1) Transmission phase modulation: by changing the cross-sectional dimensions of the nanopillars (e.g., the width of a rectangular nanopillar). and The equivalent refractive index of the nanopillar is adjusted to change the phase delay generated after the light wave passes through it. By systematically adjusting the cross-sectional dimensions, any target phase value in the range of 0 to 2π can be achieved at this location. In specific designs, the width of the nanopillar can vary, for example, from 300 nm to 550 nm, to achieve full phase coverage. 2) Geometric phase modulation: Pancharatnam-Berry phase (PB phase) is introduced by changing the rotation orientation angle of the nanopillar in the substrate plane (i.e., the angle ψ between its major axis and the reference axis). The nanopillar rotation angle ψ and the resulting additional phase delay φ_PB satisfy the relationship: φ_PB = 2ψ.

[0081] It should be noted that in practical designs, one or both mechanisms (combined control of transmission phase and geometric phase) can be used to achieve a predetermined target phase distribution. For example, a large-scale phase profile can be achieved primarily by relying on the transmission phase, while the geometric phase is used for fine-tuning errors or to achieve specific polarization operations. This method of designing the parameters of each nanopillar in reverse based on the target phase value is the microscopic physical basis for metasurfaces to achieve complex optical functions (such as generating multi-ring vector beams).

[0082] This application provides a core physical means for achieving complex optical field manipulation in metasurface devices by inversely designing the physical parameters of each nanopillar based on the target phase value. Specifically, by changing the cross-sectional dimensions of the nanopillars to achieve transmission phase modulation, a direct and robust full-phase coverage from 0 to 2π can be provided, which is the basis for constructing the main wavefront profile. Geometric phase modulation by controlling the in-plane rotational orientation angle of the nanopillars makes it possible to introduce additional polarization-related phase degrees of freedom. These two mechanisms can work independently or be used in combination (composite control), resulting in extremely high design freedom. This allows for precise encoding on a single planar device with subwavelength accuracy, thus reliably and efficiently implementing the system-level function of "efficiently converting a single beam into a multi-ring vector beam" on a micro-nanostructure, ensuring the high quality, high precision, and design flexibility of the final generated light spot.

[0083] In some embodiments, the cross-sectional dimensions of the nanopillars vary in the range of 300 nm to 550 nm to achieve transmission phase modulation coverage in the range of 0 to 2π; and / or, the rotational orientation angle of the nanopillars is optimized according to the target phase distribution and polarization error compensation requirements so that the difference in diffraction efficiency of the two-dimensional metasurface grating for transversely electrically polarized light and transversely magnetically polarized light is less than 5%.

[0084] For example, to achieve high-efficiency, low-polarization-sensitivity, precise phase modulation, the key structural parameters of the nanopillars were synergistically optimized and quantitatively constrained. Specifically, to achieve full-range transmission phase modulation coverage from 0 to 2π, the width of the rectangular nanopillars was... The design range is limited to 300 nm to 550 nm. By systematically changing this width, the equivalent refractive index of the nanopillar can be tuned, thereby providing the required continuous phase retardation. In one specific embodiment, an 8-step quantization design can be used, that is, the phase range of 0-2π is divided into 8 discrete phase steps, each step corresponding to a specific nanopillar width, and the width increment Δw between adjacent steps is approximately 31 nm. This discretization design greatly simplifies the fabrication process of the nanopillar while ensuring phase accuracy.

[0085] Meanwhile, to achieve high-performance geometric phase modulation and improve the polarization stability of the device, the design of the in-plane rotation orientation angle ψ of the nanopillars must follow a rigorous optimization process. This rotation angle is not only used to introduce the target Pancharatnam-Berry phase (φ_PB=2ψ), but also to accurately correct the polarization-dependent phase error caused by the anisotropy of the nanopillar shape. During the design process, the value of the rotation angle ψ is not determined independently, but is globally optimized through an iterative algorithm. This algorithm takes the accurate realization of the target phase distribution as the core constraint, and minimizes the difference in the response of the nanopillars to different polarization states, such as transverse electric wave (TE) and transverse magnetic wave (TM). After this optimization, the finally fabricated two-dimensional metasurface grating can ensure that the difference in diffraction efficiency for TE and TM orthogonally polarized light is less than 5%, thereby achieving near polarization-independent operating characteristics and significantly enhancing the signal stability and reliability of the eye-tracking system in complex optical environments. By combining a structural design that controls the aspect ratio of the nanopillars to within 1.5, this embodiment achieves a diffraction efficiency exceeding 75% under both TE and TM polarized light, with a polarization extinction ratio better than 0.5 dB. This polarization-independent characteristic enables the eye-tracking device to adapt to complex polarization environments in ordinary user scenarios, ensuring the stability and reliability of the illumination spot signal.

[0086] Correspondingly, the aforementioned metasurface device can be fabricated using an exemplary micro / nano fabrication process, the main steps of which include: 1) precisely defining the planar pattern of nanopillars on a substrate coated with amorphous silicon using electron-beam lithography (EBL), with the critical dimension control accuracy reaching ±10 nm; 2) transferring the above pattern to the amorphous silicon layer using deep reactive-ion etching (DRIE), etching to form a nanopillar array with a height of approximately 800 nm (process deviation controlled within ±20 nm); 3) uniformly depositing a layer of aluminum oxide with a thickness of approximately 150 nm on the device surface using atomic layer deposition (ALD) technology. As an antireflection layer, the thin film can reduce the surface reflectivity of the device in the operating wavelength range to, for example, below 0.5%, thereby effectively improving the overall transmission efficiency of the optical system.

[0087] In some embodiments, the control processing unit includes: a microcontroller for generating drive signals to control an infrared LED to emit an infrared illumination beam; a memory for storing user eye-tracking parameter calibration data and an eye-tracking program; a digital signal processor, communicatively connected to the image sensor and the memory, for executing the eye-tracking program to process corneal reflection images, extract features of multiple light spots, and calculate the gaze direction based on the user eye-tracking parameter calibration data; and a communication interface, connected to the digital signal processor, for outputting the calculated gaze direction data to an external main processing device.

[0088] For example, Figure 8 This is a schematic diagram of the structure of a control processing unit provided for an exemplary embodiment of this application. Figure 8As shown, the control processing unit 15 includes a microcontroller 151, a memory 152, a digital signal processor 153, and a communication interface 154. The microcontroller 151 serves as the system's main controller and timing core; for example, it can be a low-power microcontroller based on an ARM Cortex-M4 core, operating at a 168MHz clock frequency. Its main responsibility is to generate precise drive signals, specifically pulse-width modulation (PWM) modulated current signals (e.g., at a frequency of 20kHz) to control the switching and luminous intensity of a single infrared LED 11. This modulation frequency is higher than the persistence of vision in the human eye, effectively reducing visible flicker interference from infrared illumination. The digital signal processor 153, as the core of real-time image processing and algorithm execution, typically integrates an embedded image processing unit, running, for example, an optimized Hough circle detection algorithm to accurately identify and extract the centroid coordinates of N annular Pulchin spots and the pupil center position from each frame of image in real time. For example, the centroids of four annular light spots can be extracted within 5ms. Subsequently, based on the geometric relationship between the extracted features of multiple light spots and the pupil center, and combined with user calibration data from memory 152, a gaze resolution algorithm is executed to finally obtain the gaze direction data. Memory 152 typically uses non-volatile memory (such as 512KB Flash) to persistently store two types of key information, enabling the system to quickly adapt to different users or restore its working state. These two types of key information include: 1) user eye-tracking parameter calibration data, such as the physiological parameters of different users' eyes (such as personalized parameter tables for corneal curvature radius, Kappa angle, etc.); 2) firmware code and system configuration parameters of the eye-tracking program. The communication interface 154 serves as a data bridge between this device and an external main system (such as the main processor of a VR / AR device). It can adopt a high-speed serial standard, such as the MIPI CSI-2 interface, with a transmission rate of up to 1Gbps. This ensures that the gaze data stream derived from the 120fps high frame rate image is transmitted to the main processing device without frame loss and with low latency, thereby realizing a closed-loop interaction between gaze and display content.

[0089] Through the collaborative work of the above modules, a complete, efficient and reliable eye-tracking information processing chain is realized, from "hardware driving and synchronization" to "image acquisition and real-time feature processing", then to "personalized parameter calculation" and "high-speed result reporting".

[0090] It should be noted that all specific structural parameters involving metasurface devices (including but not limited to the size, period, rotation angle, material, and phase distribution function of the nanopillars) refer to preset static values ​​determined during the device design phase to achieve specific optical functions. Once fabricated, the physical structure of a metasurface device is fixed, becoming a passive, static optical element that does not possess any ability to actively change its physical form or structural parameters.

[0091] Considering that the unpolarized light emitted by traditional infrared LEDs in related technologies undergoes random polarization changes during corneal reflection, making it susceptible to interference from polarization-sensitive optical surfaces such as eyeglasses and sunglasses worn by users, leading to a significant decrease in eye-tracking success rate, and that existing improvement solutions lack adaptability to such polarization interference environments, some embodiments of the infrared eye-tracking device further include: a polarizer disposed in the optical path between the infrared LED and the metasurface device; and a polarization filter disposed in the optical path between the metasurface device and the image sensor; wherein the polarizer is used to make the light beam incident on the metasurface device linearly polarized, and the polarization filter is used to select the polarization of the reflected light from the eyeball after passing through the metasurface device.

[0092] The polarizer is positioned in the optical path between the infrared LED and the metasurface device to convert the unpolarized infrared light emitted by the LED into high-purity linearly polarized light with a specific polarization direction (e.g., TM) before it is incident on the metasurface device. In one specific implementation, a linear polarizer with an extinction ratio greater than 20 dB can be integrated with the infrared LED.

[0093] A polarizing filter is placed in the optical path between the metasurface device and the image sensor, with its transmission axis orthogonal to the transmission axis of the polarizer (e.g., transmitting TE light). Its main function is to selectively transmit light that has passed through the metasurface device and been reflected back by the cornea of ​​the eye.

[0094] For example, this metasurface device employs a design structure with high polarization selectivity. For instance, its microstructure can utilize high-refractive-index materials such as gallium nitride (GaN) with nanopillars having an aspect ratio >2.5, and the aspect ratio of the nanopillars is designed to be greater than 2.5. This structure produces localized resonant enhancement for specific linearly polarized light (such as transversely magnetically polarized light TM), resulting in a diffraction efficiency higher than 85%, while the diffraction efficiency for orthogonally polarized light (transversely electrically polarized light TE) is less than 10%. This achieves intrinsic polarization filtering at the device level, enhancing the polarization characteristics of the illumination light from the source.

[0095] In terms of phase distribution, the device operating surface is divided into eight concentric rings. The topological charge number corresponds to each ring. Design focal length between 3 and 10. The beam is gradient-distributed within a range of 25mm to 60mm. This design allows the emitted beam to form eight spatially separated annular spots on the target corneal plane, with an inter-annular spacing of approximately 0.8mm between each spot. The overall beam covers a corneal area with a diameter of approximately 12mm, providing sufficient and optimized feature points for high-precision gaze tracking over a large field of view. For example, Figure 9A schematic diagram of the eight-ring light spot distribution provided for an exemplary embodiment of this application.

[0096] To further enhance resistance to ambient light interference, this embodiment also adds spatially varying vector polarization modulation to each annulus. Specifically, this is achieved by designing the principal axis rotation angle of the nanopillars. The spatial distribution is achieved by following the following rules:

[0097]

[0098] in, For the first The reference rotation angle of each ring (e.g., can be set to) , , ..., This design ensures that the emitted beam from each ring is radially-angularly hybrid vector-polarized light. After reflection from the cornea, the reflected light carrying eye movement information maintains a specific polarization state distribution preset by the metasurface. When this reflected light passes again through a polarizing filter (whose transmission axis is orthogonal to the polarizer in front of the light source) placed in the receiving optical path, stray background light from non-target surfaces such as the environment and lenses (whose polarization state is not specially modulated) is greatly suppressed (suppression ratio can exceed 25dB), while the useful signal light from the cornea only experiences limited attenuation (e.g., about 3dB). Through this closed-loop polarization management mechanism, the system signal-to-noise ratio can be significantly improved by more than 8 times, thus effectively overcoming the problem of high failure rate of traditional solutions in strong interference scenarios such as when users wear polarized sunglasses. For example, Figure 10 A schematic diagram of the microstructure of a phase- and polarization-modulated nanopillar and its rotational orientation, provided as an exemplary embodiment of this application.

[0099] In some embodiments, the control processing unit is configured to enable a polarization anti-interference processing mode when the signal-to-noise ratio of the corneal reflection image is lower than a preset threshold or when a polarization interference environment is detected; wherein, the polarization anti-interference processing mode includes at least one of the following: boosting the gain signal of the image sensor, or applying a preset polarization compensation algorithm to process the corneal reflection image.

[0100] For example, the control processing unit is configured to continuously monitor the signal quality of the corneal reflection image transmitted by the image sensor; analyze the signal-to-noise ratio of the image in real time, or detect the presence of strong polarization interference (e.g., the user wearing polarized sunglasses) by analyzing the polarization state characteristics of the light spot in the image; and automatically activate the built-in polarization anti-interference processing mode when the signal-to-noise ratio is detected to be lower than a preset safety threshold, or when polarization interference is clearly detected. This polarization anti-interference processing mode includes one or more of the following strategies: 1) Activating high-gain mode: automatically controlling the image sensor to increase its analog or digital gain, for example, by 12dB, to effectively amplify the corneal reflection signal that has become weak due to attenuation by polarizing lenses, overcoming the problem of insufficient signal strength. 2) Applying a polarization compensation algorithm: calling a dedicated polarization compensation algorithm pre-stored in memory. This algorithm performs real-time digital signal processing on the acquired image, and can specifically correct or compensate for distortions such as feature point deformation and positional shift caused by polarization interference, thereby restoring the accurate pupil-Purchin spot geometry at the signal level and ensuring that the accuracy of the line-of-sight calculation is not affected. Through the closed-loop control of "real-time monitoring - automatic judgment - intelligent response (gain enhancement + algorithm compensation)", it can dynamically adapt to the switching from normal environment to strong polarization interference environment, reducing the failure rate of more than 35% in traditional solutions to an extremely low level, achieving seamless optimization of user experience and a leap in system reliability.

[0101] In some embodiments, to support multi-user sharing of the device, the memory is also configured to store multiple sets (e.g., 10 sets) of independent user eye-tracking parameter calibration data. The control processing unit can receive externally input user identification switching commands through its communication interface. Upon receiving the command, it can load the complete configuration parameters of the corresponding user from the memory in a very short time (e.g., within 0.5 seconds), quickly complete personalized adaptation, and achieve seamless switching between different users.

[0102] Based on the aforementioned intelligent control, by optimizing the driving strategy, the driving current of a single infrared LED can be reduced to 15mA. Combined with the single-source architecture, the power consumption of the entire eye-tracking illumination system is reduced by approximately 70% compared to traditional multi-LED solutions, significantly improving the battery life of mobile devices. In summary, this embodiment achieves a balance between high performance, high convenience, and high energy efficiency by integrating adaptive polarization processing, fast multi-user switching, and low-power driving.

[0103] In some embodiments, the infrared light-emitting diode, collimating element, and metasurface device are arranged sequentially along the same optical axis and integrated into a package, and the total thickness of the infrared eye-tracking device along the optical axis is less than 4 mm.

[0104] For example, the three core optical components—a single infrared LED, a collimating element, and a metasurface device—are precisely arranged in a straight line along the same optical axis, fixed by a high-precision mechanical structure or active alignment process, and finally integrated into a compact module using integrated packaging technology. This coaxial linear arrangement effectively eliminates the extra space wasted due to optical path deflection and component misalignment in traditional solutions. This is achieved by selecting ultra-thin models of each device and optimizing their thickness; for example, using miniaturized infrared LED packages with a thickness of approximately 0.8 mm, aspherical collimating lenses with short focal lengths and thicknesses of approximately 1.5 mm, and ultra-thin metasurface devices with thicknesses of only 0.5 mm to 0.9 mm.

[0105] Through the above design and integration, the total thickness of the entire infrared illumination and beamforming subsystem along the optical axis has been successfully controlled to within 4mm (e.g., a total thickness of 3.2mm can be achieved). This size represents a reduction of more than 50% compared to traditional eye-tracking illumination modules (typically 8-10mm thick), providing significant physical possibilities and engineering feasibility for integrating high-performance eye-tracking functionality into the narrow temples or frame space of VR / AR glasses.

[0106] The above embodiments illustrate the implementation of the infrared eye-tracking device. Next, specific embodiments will be used to illustrate the application of the infrared eye-tracking device.

[0107] For example, Figure 11 A schematic diagram of the structure of an eye-tracking device provided as an exemplary embodiment of this application. Figure 11 As shown, the eye-tracking device 110 integrates an infrared eye-tracking device 10 as described in any of the above embodiments. The device 110 can function as a standalone integrated eye tracker with independent power supply and data processing capabilities, and can be directly used for psychological experiments, user experience research, or as a peripheral input device for a computer.

[0108] Accordingly, Figure 12 A schematic diagram of an eye-tracking device provided as an exemplary embodiment of this application. (See diagram below.) Figure 12 As shown, the specific form of this eye-tracking device is, for example, a pair of wearable glasses, such as augmented reality glasses. The eye-tracking device 10 is miniaturized and distributedly integrated into the frame structure of the glasses 120. For example, the three core optical components—a single infrared LED, a collimating element, and a metasurface device—are precisely aligned along the same optical axis, fixed by a high-precision mechanical structure or active alignment process, and finally integrated into a compact module using integrated packaging technology. This integrated module and the image sensor are optically coplanarly arranged compactly on the inside of the glasses frame (e.g., near the bridge of the nose), with their optical path pointing towards the wearer's eyeball. For example, Figure 13A schematic diagram of the distribution of annular light spots on the surface of the cornea provided in an exemplary embodiment of this application.

[0109] It should be noted that, Figure 12 and Figure 13 The compact arrangement of the integrated module (or metasurface device 13) and the image sensor 14 in an optically coplanar manner within the frame of the glasses 120 is merely an example. In practical applications, the arrangement, optical path, and integration method of the above modules can be adaptively adjusted and modified according to the specific form, structural design constraints, and performance optimization goals of the eye-tracking device. For example, if space permits, the modules can also be arranged in other positions within the frame.

[0110] In summary, this application has at least the following advantages:

[0111] I. By adopting a "single infrared LED + metasurface" architecture, the system hardware is fundamentally simplified. Specifically, a single LED replaces the traditional array of 4-8 LEDs, reducing the number of light sources by 75%-87.5%. This not only directly reduces the bill of materials cost but also reduces surface mount technology (SMT) steps by 60%. In terms of physical dimensions, this architecture, combined with ultra-thin components and coaxial packaging, compresses the total thickness of the eye-tracking module from the traditional 8-10 mm to less than 3-4 mm, a reduction of over 50%, allowing for perfect integration into AR glasses where the temple thickness is typically less than 5 mm. Simultaneously, the operating current of a single LED can be reduced to 15-20 mA, and the total system power consumption is less than 50 mW, a reduction of approximately 70% compared to multi-LED solutions, effectively extending the battery life of VR all-in-one devices and other mobile devices by approximately 20%.

[0112] II. In terms of optical performance, the solution presented in this application represents a qualitative leap. The multi-ring illumination spot generated by the metasurface exhibits excellent quality, with a ring width of less than 0.3 mm, a contrast ratio exceeding 10 dB, and an edge sharpness (sharpness factor) three times higher than traditional dot-matrix spot solutions, enabling sub-pixel-level (<0.1 pixel) positioning accuracy for corneal reflection features. Thanks to the rotational symmetry of the ring spot, the system's effective tracking field of view expands to ±45°. Even when the eyeball rotates to its extreme position, at least two effective spots remain stable in the image, thus reducing the tracking blind zone area by 90%. Furthermore, by introducing polarization selection and vector modulation design, the solution presented in this application possesses strong background light suppression capabilities, reducing ambient light interference by more than 25 dB. This ensures reliable eye tracking even in strong outdoor light (>50,000 lux) environments, greatly expanding its application potential in complex scenarios such as outdoor AR devices.

[0113] Third, by using a single light source, the necessary intensity consistency matching calibration step in traditional multi-LED solutions is effectively eliminated, significantly reducing the calibration time for a single device on the production line from 120 seconds to 30 seconds, thus improving production efficiency. In terms of thermal management, the heat source is concentrated from multiple dispersed LEDs to a single point, effectively reducing the module's peak temperature (approximately 15°C), thereby improving the long-term operational reliability of the entire eye-tracking module, with its mean time between failures (MTBF) expected to increase by 40%.

[0114] In summary, the solution proposed in this application not only achieves extreme simplification, thinness, and low power consumption at the device level, but also surpasses the core optical performance, environmental adaptability, production efficiency, and system reliability, providing a high-performance, highly reliable, and easily integrated eye-tracking solution for next-generation near-eye display devices.

[0115] In the above embodiments, the descriptions of each embodiment have their own emphasis. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments. The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combination of these technical features does not contradict each other, it should be considered within the scope of this specification.

[0116] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the following claims.

[0117] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. An infrared eye tracking device, characterized by include: A single infrared LED is configured to emit an infrared illumination beam; A collimating element is disposed on the light emission path of the infrared light-emitting diode and is used to collimate and shape the infrared illumination beam; A metasurface device is disposed on the light-emitting path of the collimating element and located between the infrared light-emitting diode and the human eye observation area. The metasurface device is a device designed based on a two-dimensional metasurface grating structure and is configured to diffract and modulate the collimated and shaped incident light beam into N spatially separated illumination beams to form N concentrically distributed light spots on the corneal surface of the target eye. Wherein, N is an even number not less than 2, and the position of each light spot on the corneal surface of the target eye corresponds to a different line-of-sight calibration feature point. An image sensor, whose optical axis is pointed to the human eye's observation area, is used to acquire corneal reflection images containing N light spots; The control processing unit is communicatively connected to the image sensor and is used to output gaze direction data based on the relative positional relationship between the N light spots in the corneal reflection image and the center of the pupil. The metasurface device is a two-dimensional metasurface grating with a circular aperture. The arrangement of the microstructure array of the two-dimensional metasurface grating varies in different regions within the circular aperture: in the central circular region, the microstructures are arranged in a square lattice; in the annular region surrounding the central circular region, the microstructures are arranged in concentric rings. Wherein, the diameter D_c of the central circular region and the aperture diameter D of the circular aperture satisfy the relationship: 0.33≤D_c / D≤0.5; the radial width W of the annular region and the aperture diameter D satisfy the relationship: 0.33≤W / D≤0.67; the aperture diameter D is 2mm-4mm.

2. The infrared eye-tracking device of claim 1, wherein, The two-dimensional metasurface grating includes a transparent substrate and a subwavelength microstructure array formed on the transparent substrate. The subwavelength microstructure array is a nanopillar array. The refractive index n of the nanopillar material at the infrared illumination beam wavelength satisfies n≥2.0, and the height H of the nanopillar satisfies 600nm≤H≤1000nm. And / or, the subwavelength period of the subwavelength microstructure array is 550 nm to 650 nm.

3. The infrared eye-tracking device according to claim 1, characterized in that, The working surface of the two-dimensional metasurface grating is divided into a plurality of concentric annular modulation regions, each of which corresponds to generate one of the light spots; wherein the phase distribution function of the first annular modulation region satisfies: ​​ In the formula, Radial coordinates, It is the azimuth angle. The operating wavelength of the infrared light-emitting diode is [wavelength]. For the first The focusing distance corresponding to each light spot For the first Topological charge number of each ring modulation region and The first The inner and outer boundary radii of the ring modulation region.

4. The infrared eye-tracking device according to claim 2, characterized in that, The physical parameters of each nanopillar in the nanopillar array are designed according to the target phase value required to be achieved at the location of each nanopillar; wherein, transmission phase modulation in the range of 0 to 2π is provided by changing the cross-sectional size of the nanopillar, and / or geometric phase modulation is provided by changing the rotational orientation angle of the nanopillar in the plane.

5. The infrared eye-tracking device according to claim 4, characterized in that, The cross-sectional dimensions of the nanopillars vary in the range of 300 nm to 550 nm to achieve transmission phase modulation coverage in the range of 0 to 2π. And / or, the rotation orientation angle of the nanopillar is optimized according to the target phase distribution and polarization error compensation requirements, so that the difference in diffraction efficiency of the two-dimensional metasurface grating for transversely electrically polarized light and transversely magnetically polarized light is less than 5%.

6. The infrared eye-tracking device according to any one of claims 1 to 5, characterized in that, The control processing unit includes: A microcontroller is used to generate drive signals to control the infrared light-emitting diodes to emit the infrared illumination beam; The memory is used to store user eye movement parameter calibration data and eye tracking programs; A digital signal processor, communicatively connected to the image sensor and the memory, is used to execute the eye-tracking program to process the corneal reflection image, extract features of multiple light spots, and calculate the gaze direction based on the user's eye-tracking parameter calibration data; A communication interface, connected to the digital signal processor, is used to output the calculated line-of-sight direction data to an external main processing device.

7. The infrared eye-tracking device according to any one of claims 1 to 5, characterized in that, The infrared eye-tracking device further includes: a polarizer disposed in the optical path between the infrared light-emitting diode and the metasurface device; and a polarizing filter disposed in the optical path between the metasurface device and the image sensor; wherein the polarizer is used to make the light beam incident on the metasurface device linearly polarized, and the polarizing filter is used to polarize the reflected light from the eyeball after passing through the metasurface device; The control processing unit is configured to enable a polarization anti-interference processing mode when the signal-to-noise ratio of the corneal reflection image is lower than a preset threshold or when a polarization interference environment is detected; wherein, the polarization anti-interference processing mode includes at least one of the following: boosting the gain signal of the image sensor, or applying a preset polarization compensation algorithm to process the corneal reflection image.

8. The infrared eye-tracking device according to any one of claims 1 to 5, characterized in that, The infrared light-emitting diode, the collimating element, and the metasurface device are arranged sequentially along the same optical axis and integrated into a package. The total thickness of the infrared eye-tracking device along the optical axis is less than 4 mm.

9. An eye-tracking device, characterized in that, It integrates an infrared eye-tracking device as described in any one of claims 1 to 8.

Images