Sight drop point detection method

Abstract

The invention relates to the technical field of automatic driving, in particular to a sight falling point detection method. In the embodiment of the invention, the method comprises the following steps of: outputting a light source by using a light source device, generating a plurality of light spots through an optical module, projecting the light spots to eyeballs, and acquiring a plurality of reflected light spots reflected by the eyeballs by using an image acquisition device so as to form an acquired image, the electronic device analyzes a plurality of reflected light spots in the acquired image and the outer edge of the pupil of the eyeball through the application program, obtains the roundness of the plurality of light spots and the center coordinate of the pupil, selects a first light spot and a second light spot with the roundness greater than or equal to a threshold value, obtains a reference point according to the first light spot and the second light spot by using the electronic device, and displays the reference point. And performing a coordinate vector operation program by taking the reference point as an original point to obtain a mapping coordinate of the sight line drop point.

Description

Technical Field

[0001] This application relates to the field of autonomous driving technology, and in particular to a method for detecting the point of gaze. Background Technology

[0002] With the development of advanced driver assistance systems (ADAS) and smart cockpit technology, driver eye tracking has gradually become an important basis for assessing attention and safe driving status.

[0003] Current common line-of-sight detection technologies can be mainly divided into electrophysiological, imaging and optical types. Among them, the pupil-corneal reflection (PCCR) method is the mainstream method. It is widely used in human-machine interfaces and vehicle monitoring systems because of its non-contact, fast response speed and ability to operate in infrared environments.

[0004] The relevant PCCR method uses an infrared light source to illuminate the eyeball, forming a reflective spot (glint) on the corneal surface. A camera captures the image of the pupil and the reflective spot. By analyzing the relative displacement between the pupil center and the reflective spot through image algorithms, the gaze direction and the point of gaze can be calculated.

[0005] However, this method typically uses only a single or dual light source as a reflection reference. When the driver's head position or eye angle changes, the reflected light spot is prone to shift to the edge of the sclera or iris, causing deformation of the reflection point or a decrease in its roundness, which in turn leads to unstable tracking accuracy.

[0006] Furthermore, existing technologies often employ fixed physical light source configurations, such as placing two sets of infrared LEDs on either side of the camera. While this structure can stably form reflection points within a certain field of view, changes in the optical path angle when the driver's eyes deviate from the optical axis or the seat position changes can still lead to uneven distribution of reflected light spots, increasing the correction burden on image algorithms. This is especially true in the vehicle cabin environment, where the light source is constrained by the space of the dashboard or steering wheel structure, making the optical path design even more restrictive.

[0007] To improve the above problems, some studies have proposed to increase accuracy by arranging multiple light sources or using multiple cameras to track simultaneously. However, multi-light source systems require additional drive circuits and synchronization control modules, which not only increases cost and energy consumption, but also increases system size and integration difficulty, which is not conducive to the miniaturization design of automotive devices. At the same time, the overlapping of optical paths between multiple light sources can easily produce stray reflections or mutual interference, resulting in uneven image brightness and algorithm misjudgment.

[0008] Furthermore, traditional gaze tracking systems often use only a single reflection point as a reference coordinate, lacking a mechanism to discriminate the shape or roundness of the reflected light spot. When the light spot is deformed due to corneal curvature or ambient light interference, the system still forces the use of the reflection point for calculation, resulting in an offset error in the gaze placement point. This phenomenon is more pronounced during driving due to dynamic changes in posture, limiting the stability of existing PCCR methods in automotive environments.

[0009] Furthermore, existing eye-tracking technologies still suffer from problems such as unstable light spots, insufficient roundness of reflection points, and complex multi-light source architectures in automotive applications. Therefore, it is necessary to solve the problem of how to improve an optical and algorithmic integration scheme that can generate multiple stable light spots with a single light source and combine roundness discrimination to select effective reflection points, so as to improve the accuracy of the eye-tracking point and the overall reliability of the system. Summary of the Invention

[0010] One objective of this application is to provide a method for detecting the gaze point. This method uses a single light source combined with an optical beam splitter to generate multiple sets of symmetrical light spots. Through image acquisition and algorithm analysis, the relative displacement between the pupil center and the average reference point of the multiple light spots is used to calculate the gaze direction and the gaze point position. This can effectively improve the stability and roundness of corneal reflection, reduce errors caused by head movement or light source deviation, and achieve a high-precision gaze tracking function suitable for automotive environments.

[0011] To achieve the aforementioned objectives, this application provides a method for detecting the point of gaze, comprising the following steps: using a light source device to output a light source, generating multiple light spots through an optical module, and projecting them onto the eyeball; using an image acquisition device to acquire multiple reflected light spots reflected from the eyeball to form an acquired image; an electronic device analyzing the multiple reflected light spots and the outer edge of the pupil in the acquired image through an application program, and obtaining the roundness of the multiple light spots and the center coordinates of the pupil, selecting a first light spot and a second light spot with a roundness greater than or equal to a threshold; using the electronic device to obtain a reference point based on the first light spot and the second light spot, and using the reference point as the origin to perform a coordinate vector calculation program to obtain the mapped coordinates of the point of gaze.

[0012] This application provides an embodiment in which the light source device uses an infrared LED, and the wavelength range of the light source is between 850nm and 940nm.

[0013] This application provides an embodiment in which the optical module includes: a collimating lens disposed on the light-emitting side of the light source device, the collimating lens being used to collimate the light source into a collimated light source; a porous grating disposed on one side of the collimating lens, the porous grating being used to divide the collimated light source into multiple light spots; a repeater disposed on one side of the porous grating, the repeater being used to guide and focus the multiple light spots; and a beam splitter disposed on one side of the repeater, the beam splitter being used to reflect the multiple light spots to the eyeball and transmit the multiple reflected light spots reflected by the eyeball to the image acquisition device; An objective lens is positioned between the eyeball and the beam splitter to focus multiple light spots onto the eyeball. An imaging lens is positioned between the beam splitter and the image acquisition device to focus multiple reflected light spots onto the image acquisition device. The light source is collimated by a collimating lens and split into multiple light spots by a multi-aperture grating. These multiple light spots are focused onto the eyeball by a relay lens and the objective lens, generating multiple reflected light spots on the eyeball. These reflected light spots pass through the beam splitter and the imaging lens, and the image acquisition device acquires these multiple reflected light spots.

[0014] This application provides an embodiment in which the porous grating is a two-hole grating or a four-hole grating.

[0015] This application provides an embodiment in which a light source device, a porous grating, a relay mirror, a beam splitter, an objective lens, and an imaging mirror are arranged along the same optical axis to form a modular structure that can be installed in a single location within a vehicle body.

[0016] This application provides an embodiment in which, in the steps of an electronic device analyzing multiple light spots and the outer edge of the pupil in an image and obtaining the roundness of the multiple light spots and the center coordinates of the pupil, and selecting a first light spot and a second light spot with a roundness greater than or equal to a threshold, the application analyzes the outer edge of the pupil and fits the pupil into a circle to obtain the center coordinates of the pupil.

[0017] This application provides an embodiment in which, in the step of an electronic device analyzing multiple light spots in an acquired image and the outer edge of the pupil of the eye through an application, and obtaining the roundness of the multiple light spots and the center coordinates of the pupil, and selecting a first light spot and a second light spot with a roundness greater than or equal to a threshold, the threshold is 0.7.

[0018] This application provides an embodiment in which, in the step of obtaining a reference point using an electronic device based on a first light spot and a second light spot, the coordinates of the reference point are the average of the first coordinates and the second coordinates of the first light spot and the second light spot.

[0019] This application provides an embodiment in which the step of obtaining the mapped coordinates of the gaze point by performing coordinate vector calculation with a reference point as the origin includes the following steps: calculating the pupil center displacement vector of the center coordinates based on the reference point at the first gaze time point and the second gaze time point, and obtaining the mapped coordinates.

[0020] This application can effectively reduce the errors caused by spot offset, deformation and reflection angle differences in single-light source PCCR, and avoid the complexity of synchronous control and hardware integration in multi-light source systems. With the help of optical structure and algorithm, this application can achieve high-precision eye tracking function with low power consumption and high integration in confined spaces such as automotive cockpits, and has good stability and mass production feasibility. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating steps S10 to S50 of an embodiment of this application.

[0022] Figure 2 This is a flowchart illustrating step S31 of an embodiment of this application;

[0023] Figure 3 This is a flowchart illustrating step S51 of an embodiment of this application;

[0024] Figure 4 This is a schematic diagram of the structure of an optical module according to an embodiment of this application;

[0025] Figure 5 This is a schematic diagram of an eyeball image showing a single reflected light spot according to an embodiment of this application;

[0026] Figure 6 This is a schematic diagram of the line of sight landing point of a single reflected light spot according to an embodiment of this application;

[0027] Figure 7 This is a schematic diagram of an eye image showing two reflected light spots according to an embodiment of this application;

[0028] Figure 8 This is a schematic diagram of the line-of-sight landing point of two reflected light spots according to an embodiment of this application.

[0029] Explanation of reference numerals in the attached figures:

[0030] 10: Light source device; 20: Optical module; 21: Collimating lens; 22: Multi-aperture grating; 23: Relay lens; 24: Beam splitter; 25: Objective lens; 26: Imaging lens; 30: Eyeball; 40: Image acquisition device; 91: Light source; 92: Light spot; 93: Reflected light spot; S10, S20, S30, S31, S40, S50, S51: Steps. Detailed Implementation

[0031] To gain a better understanding of the features and effects of this application, preferred embodiments and detailed descriptions are provided below.

[0032] The related pupil-corneal reflex (PCCR) method often uses a single or dual physical light source configuration. The position of the reflection point is easily affected by the driver's head posture, eye rotation angle and ambient light, resulting in insufficient roundness of the reflected light spot. This leads to an unstable relative relationship between the pupil center and the reflection point. When the reflected light falls at the junction of the iris and sclera, the shape of the light spot is distorted, and the error of the line of sight landing point is significantly increased. In addition, the use of multiple light sources or multiple camera architectures to improve detection accuracy can partially improve the problem, but it complicates the overall optical system, increases costs and makes the calibration process cumbersome, which is not conducive to the integration and long-term stable operation of automotive systems.

[0033] This design utilizes a single infrared light source combined with optical fibers, a collimating lens, and a multi-aperture grating to split the output beam into multiple symmetrical rays, forming several stable reflective spots on the corneal surface. By analyzing the roundness of these spots, the two most stable spots are selected as calculation reference points. The system then uses the average position of these two spots as a baseline, along with the pupil center, for normalization and vector calculation to determine the precise location of the line of sight.

[0034] In the following description, various embodiments of this application will be illustrated with the aid of the accompanying drawings to provide a detailed account. However, the concepts of this application may be embodied in many different forms and should not be construed as limited to the exemplary embodiments set forth herein.

[0035] First, please refer to Figure 1 , Figure 1 This is a flowchart illustrating steps S10 to S50 of an embodiment of this application, as shown below. Figure 1 As shown, the detection method for the landing point of the line of sight in this embodiment includes the following steps:

[0036] Step S10: The light source output by the light source device generates a light spot through the optical module and projects it onto the eyeball;

[0037] Step S20: Use an image acquisition device to acquire the reflected light spot from the eyeball to form an acquired image;

[0038] Step S30: The electronic device analyzes the reflected light spot and the outer edge of the pupil in the acquired image through the application program, and obtains the roundness of the reflected light spot and the center coordinates of the pupil, and selects the first light spot and the second light spot with a roundness greater than or equal to the threshold.

[0039] Step S40: Obtain a reference point using an electronic device based on the first and second light spots;

[0040] Step S50: Use the reference point as the origin to perform coordinate vector calculations to obtain the mapped coordinates of the line of sight.

[0041] In this embodiment, please refer to the following: Figure 4 , Figure 4 This is a schematic diagram of the structure of an optical module according to an embodiment of this application, as shown below. Figure 4 As shown in steps S10 to S20, the light source 91 output by the light source device 10 generates multiple light spots 92 through the optical module 20 and projects them onto the eyeball 30. The image acquisition device 40 acquires multiple reflected light spots 93 reflected from the eyeball 30 to form an acquired image (not shown).

[0042] In this embodiment, the light source device 10 uses an infrared LED, and the wavelength range of the light source 91 is between 850nm and 940nm. The infrared LED is selected to avoid glare for the driver and improve image contrast.

[0043] In this embodiment, the optical module 20 includes a collimating lens, a multi-aperture grating, a relay lens, a beam splitter, an objective lens, and an imaging lens. The collimating lens 21 is disposed on the light-emitting side of the light source device 10 to collimate the light source 91 into a collimated light source.

[0044] In this embodiment, the porous grating 22 is disposed on one side of the collimating lens 21. The porous grating is used to divide the collimating light source into multiple light spots 92. Furthermore, in this embodiment, the relay lens 23 is disposed on one side of the porous grating. The relay lens 23 is used to guide and focus the multiple light spots 92.

[0045] In this embodiment, the beam splitter 24 is disposed on one side of the relay lens 23. The beam splitter 24 is used to reflect multiple light spots 92 to the eyeball 30 and transmit the multiple reflected light spots 93 reflected by the eyeball 30 to the image acquisition device 40.

[0046] The objective lens 25 is positioned between the eyeball 30 and the beam splitter 24. The objective lens 25 is used to focus multiple light spots 92 onto the eyeball 30. The imaging lens 26 is positioned between the beam splitter 24 and the image acquisition device 40. The imaging lens 26 is used to focus multiple reflected light spots 93 onto the image acquisition device 40.

[0047] In this embodiment, the light source 91 is collimated by the collimating lens 21 and divided into multiple light spots 92 by the porous grating 22. The multiple light spots 92 are focused on the eyeball 30 by the relay lens 23 and the objective lens 25, and multiple reflected light spots 93 are generated in the eyeball 30. The multiple reflected light spots 93 pass through the beam splitter 24 and the imaging lens 26, so that the image acquisition device 40 can acquire the multiple reflected light spots 93.

[0048] The light source device 10, the porous grating 22, the relay mirror 23, the beam splitter 24, the objective lens 25 and the imaging mirror 26 are arranged along the same optical axis to form a modular structure that can be installed in a single location within the vehicle body.

[0049] In this embodiment, the porous grating is a two-aperture grating or a four-aperture grating, and the porous grating produces two or four symmetrical light spots depending on the type of grating.

[0050] Next, in this embodiment, as described in step S30, the electronic device analyzes the multiple reflected light spots 93 and the outer edge of the pupil of the eyeball 30 in the acquired image through the application, and obtains the roundness of the multiple reflected light spots 93 and the center coordinates of the pupil, and selects the first light spot and the second light spot with a roundness greater than or equal to the threshold.

[0051] In this embodiment, the electronic device includes, but is not limited to, personal computers, tablet computers, and smartphones.

[0052] In this embodiment, please refer to Figure 2 , Figure 2 This is a flowchart illustrating step S31 of an embodiment of this application, as shown below. Figure 2 As shown, step S31 is included:

[0053] Step S31: The application analyzes the outer edge of the pupil and fits the pupil into a circle to obtain the center coordinates of the pupil.

[0054] In other words, the application in the electronic device analyzes the outer circle of the pupil and fits the pupil into a circle to obtain the center coordinates (Xp, Yp) of the pupil.

[0055] In this embodiment, the first and second light spots with a roundness greater than or equal to a threshold are selected, wherein the threshold is 0.7.

[0056] In this embodiment, as described in steps S40 to S50, an electronic device is used to obtain a reference point based on the first light spot and the second light spot, and the reference point is used as the origin to perform a coordinate vector calculation program to obtain the mapped coordinates of the line of sight.

[0057] The coordinates of the reference point mentioned above are the average of the first coordinates (X1, Y1) and the second coordinates (X2, Y2) of the first and second light spots.

[0058] In step S50, please refer to Figure 3 , Figure 3 This is a flowchart illustrating step S51 of an embodiment of this application, as shown below. Figure 3 As shown, the steps include:

[0059] Step S51: Calculate the pupil center displacement vector based on the reference point at the first fixation time point and the second fixation time point, and obtain the mapped coordinates.

[0060] In this embodiment, the pupil center displacement vector, Δ(X, Y), is calculated based on the reference point using the first and second fixation time points, respectively, and the mapped coordinates are obtained.

[0061] The coordinate vector calculation program in this embodiment is coordinate normalization. It uses the coordinates of the reference point as the origin and translates the coordinates of the pupil center so that all gaze images are compared with the same reference. This can eliminate the influence of overall image offset caused by slight head movements of the driver or differences in camera installation angle, and make the relative displacement between the pupil center and the reference point directly reflect the direction of gaze change. After normalization, the pupil movement vector Δ(X, Y) can be calculated in the same standardized coordinate system to further calculate the corresponding gaze landing point position, thereby improving the stability and accuracy of the tracking results.

[0062] The advantage of this embodiment is that it only requires a single light source to generate multiple sets of symmetrical light spots through optical beam splitting and lens group design. Combined with roundness discrimination and reference point averaging algorithms, it effectively improves the stability of the light spots and the accuracy of the line-of-sight landing point. By using coordinate normalization and vector calculation, it reduces errors caused by head movements, camera position shifts, or changes in ambient light, maintaining high accuracy and reproducibility. Compared to typical multi-light source or multi-camera architectures, this application features low hardware complexity, low power consumption, and high integration, making it suitable for in-vehicle gaze detection and driving behavior monitoring systems, while also ensuring module miniaturization and mass production feasibility.

[0063] Next, in order to make the form of this embodiment clearer, practical application examples are given here, and a comparative experiment is conducted on the vertical arrangement of a single light source and two light sources.

[0064] Both sets of experiments were conducted under identical conditions in a simulated vehicle cockpit environment. Infrared cameras were used to capture eye images of the subjects in different gaze directions, and the deviation of the gaze point was calculated to evaluate the overall accuracy and stability.

[0065] First, in the single-reflection spot experimental conditions, please refer to... Figure 5 , Figure 5 This is a schematic diagram of an eyeball image with a single reflected light spot according to an embodiment of this application, as shown below. Figure 5 As shown, the light spot formed by the optical engine projecting onto the eye is the calculated position and the connection with the pupil. When only one set of infrared light sources is used to illuminate the eye, only one reflected light spot is formed on the cornea, and its position will shift significantly as the eyeball moves.

[0066] In some cases, the reflection point falls at the junction of the iris and sclera, causing light spot distortion and insufficient roundness. This leads to accumulated errors in the system's line-of-sight vector calculation. Please refer to this information as well. Figure 6 , Figure 6 This is a schematic diagram of the line-of-sight landing point of a single reflected light spot according to an embodiment of this application, as shown below. Figure 6 As shown, Figure 6 In the experiment, it was assumed that there were nine target points in front of us. Each person looked at those nine points, and the difference between the predicted point obtained by the algorithm and the calculated point was shown. The results showed that the deviation directions of different measurement points were scattered and the distances were large, and the landing point stability was insufficient. The statistical analysis is shown in Table 1 below. Table 1 is the statistical result of the line-of-sight landing point error of a single reflected light spot.

[0067] Table 1. Statistical results of the line-of-sight landing point error of a single reflected light spot

[0068]

[0069] Please refer to Table 1. The average errors (ΔX, ΔY) under single light source conditions are 0.44 and 0.27, respectively, and the standard deviations are 0.29 and 0.21, respectively. This shows that the results not only have a high error amplitude, but also have large fluctuations in repeated tests, making it difficult to maintain stable and accurate gaze estimation.

[0070] Next, an experiment was conducted on the two reflection spot conditions. In this application, the light source device 10 is split into two infrared beams, one above the other, by the optical module 20, so that two symmetrical reflection spots appear on the corneal surface of the eyeball 30.

[0071] Please refer to the experimental results. Figure 7 , Figure 7 This is a schematic diagram of an eyeball image with two reflected light spots according to an embodiment of this application, and please also refer to... Figure 8 , Figure 8 This is a schematic diagram of the line-of-sight landing point of the two reflected light spots according to an embodiment of this application, as shown below. Figure 7 and Figure 8 As shown, under different viewing angles, the positions of the two light spots remain symmetrical and their shapes are stable. The algorithm can take their average position as a reference point and use it in conjunction with the pupil center of the eyeball (30°) to calculate the line-of-sight vector, which can effectively reduce the landing point error caused by light spot offset. Figure 8 The corresponding line-of-sight error distribution diagram shows that the deviation has been significantly reduced and the distribution is concentrated. The statistical analysis is shown in Table 2 below. Table 2 is the statistical result of the line-of-sight error of the two reflection spots.

[0072] Table 2. Statistical results of the error in the line-of-sight landing point of the second reflection spot.

[0073]

[0074] Statistical results show that the average errors (ΔX, ΔY) of the two-reflection light spots decreased to 0.28 and 0.16, respectively, and the standard deviations also decreased to 0.14 and 0.07, respectively. The overall error range was reduced by about 59% to 63% compared with the single light source.

[0075] Based on the aforementioned experimental results, this application, through a single infrared light source combined with a grating and optical lens group design, can form multiple sets of stable and symmetrical reflective spots on the corneal surface of the eye. The roundness screening method automatically selects suitable reflection points as the basis for calculation. This design effectively overcomes the problem of line-of-sight error caused by reflection spot offset, deformation or incident angle difference in the traditional single-source PCCR method.

[0076] Furthermore, the multi-spot reference point calculation and normalized vector algorithm of this application can maintain high stability in automotive environments and can still accurately estimate the gaze direction under conditions of driver posture changes or light interference. Compared with existing dual-light source or multi-camera systems, this application realizes multi-beam function with a single module, which has the advantages of low hardware complexity, low cost and high integration, and is suitable for use in embedded in vehicle lens modules or dashboard structures.

[0077] Therefore, this application can significantly improve the detection accuracy and reliability of automotive eye tracking systems, while also taking into account miniaturization and energy consumption control. It has high industrial value and practical feasibility in smart cockpit, safe driving monitoring and human-machine interface applications.

[0078] The embodiments described above, in this application, provide a method for detecting the gaze point. This method utilizes a single light source in conjunction with an optical beam splitter to generate multiple symmetrical light spots. After acquiring images of the pupil and the light spots via an image acquisition device, an algorithm analyzes the relative displacement between the pupil center and the average reference point of the multiple light spots to calculate the gaze direction and the gaze point location. This improves the stability of corneal reflection and the roundness of the light spots, and effectively suppresses errors caused by head movements or changes in the incident angle of the light source. Thus, it achieves high-precision gaze tracking technology suitable for automotive environments.

[0079] However, the above description is merely a preferred embodiment of this application and is not intended to limit the scope of protection of this application. Any equivalent substitutions and modifications made based on the shape, structure, features, and spirit described in the claims of this application should be included within the scope of protection of this application. Therefore, the scope of protection of this patent application should be determined by the appended claims.

Claims

1. A method for detecting the point of gaze, characterized in that, The method for detecting the point of view includes: The light source is output by a light source device, which generates multiple light spots through an optical module and projects them onto the eyeball; Multiple reflected light spots from the eyeball are captured using an image acquisition device to form an acquired image; The electronic device analyzes the multiple reflected light spots in the acquired image and the outer edge of the pupil of the eyeball through an application program, and obtains the roundness of the multiple reflected light spots and the center coordinates of the pupil, and selects the first light spot and the second light spot whose roundness is greater than or equal to a threshold. The electronic device is used to obtain a reference point based on the first light spot and the second light spot; and The reference point is used as the origin to perform coordinate vector calculations to obtain the mapped coordinates of the point where the line of sight falls.

2. The method for detecting the point of gaze as described in claim 1, characterized in that, The light source device uses an infrared LED, and the wavelength range of the light source is between 850nm and 940nm.

3. The method for detecting the point of gaze as described in claim 1, characterized in that, The optical module includes: A collimating lens is disposed on the light-emitting side of the light source device, and the collimating lens is used to collimate the light source into a collimated light source; A porous grating is disposed on one side of the collimating lens, and the porous grating is used to divide the collimating light source into the plurality of light spots; A relay mirror is disposed on one side of the porous grating, and the relay mirror is used to guide and focus the multiple light spots; A beam splitter is disposed on one side of the relay lens. The beam splitter is used to reflect the plurality of light spots to the eyeball and transmit the plurality of reflected light spots reflected by the eyeball to the image acquisition device. An objective lens, disposed between the eyeball and the beam splitter, is used to focus the plurality of light spots onto the eyeball; and An imaging mirror is disposed between the beam splitter and the image acquisition device, and the imaging mirror is used to focus the plurality of reflected light spots onto the image acquisition device; The light source is collimated by the collimating lens and divided into multiple light spots by the porous grating. The multiple light spots are focused onto the eyeball by the relay lens and the objective lens, and multiple reflected light spots are generated on the eyeball. The multiple reflected light spots pass through the beam splitter and the imaging lens, and the image acquisition device acquires the multiple reflected light spots.

4. The method for detecting the point of gaze as described in claim 3, characterized in that, The porous grating is a two-hole grating or a four-hole grating.

5. The method for detecting the point of gaze as described in claim 3, characterized in that, The light source device, the porous grating, the relay mirror, the beam splitter, the objective lens, and the imaging mirror are arranged along the same optical axis to form a modular structure that can be installed in a single location within the vehicle body.

6. The method for detecting the point of gaze as described in claim 1, characterized in that, The step of the electronic device analyzing the multiple light spots and the outer edge of the pupil of the eyeball in the acquired image through an application, obtaining the roundness of the multiple light spots and the center coordinates of the pupil, and selecting the first light spot and the second light spot whose roundness is greater than or equal to a threshold includes the following steps: The application analyzes the outer edge of the pupil and fits the pupil into a circle to obtain the center coordinates of the pupil.

7. The method for detecting the point of gaze as described in claim 1, characterized in that, In the step of the electronic device analyzing the multiple light spots in the acquired image and the outer edge of the pupil of the eyeball through an application, obtaining the roundness of the multiple light spots and the center coordinates of the pupil, and selecting the first light spot and the second light spot whose roundness is greater than or equal to a threshold, the threshold is 0.

7.

8. The method for detecting the point of gaze as described in claim 1, characterized in that, In the step of obtaining a reference point using the electronic device based on the first light spot and the second light spot, the coordinates of the reference point are the average of the first coordinates and the second coordinates of the first light spot and the second light spot.

9. The method for detecting the point of gaze as described in claim 1, characterized in that, The step of obtaining the mapped coordinates of the line-of-sight point by performing coordinate vector calculations using the reference point as the origin includes the following steps: At the first and second fixation time points, the pupil center displacement vectors of the center coordinates are calculated based on the reference point, and the mapped coordinates are obtained.

Images