Head-mounted eye movement tracking device

Abstract

The invention relates to the technical field of eye feature detection and tracking, in particular to a head-mounted eye movement tracking device which comprises an eyeshade worn on the head, the eyeshade comprises a main body part, display lenses are symmetrically installed in the main body part, bandages are symmetrically installed at the two ends of the main body part, and an adjusting assembly is arranged between every two adjacent bandages; a film pressure sensor layer is arranged on the inner side wall of the bandage, and an infrared camera and a near-infrared LED are arranged on the inner side wall of the main body part; an eyeball tracking module, a calculation module and a guide module are arranged in the main body part, a visual stimulation task is presented through the display lens, a pupil image is collected by using the eyeball tracking module, and the calculation module analyzes a pupil trajectory and evaluates the attention level; when the pressure data is detected to be abnormal, the system can guide the child to adjust the tightness of the bandage. The child is guided to adjust autonomously by designing the guiding task, and the situation that an operator does not adjust the device in time, so that the child pulls the device autonomously, and consequently the detection data result is inaccurate is avoided.

Description

Technical Field

[0001] This invention relates to the field of eye feature detection and tracking technology, specifically to a head-mounted eye-tracking device. Background Technology

[0002] In eye-tracking studies targeting children, operators need to adjust the parameters of the eye-tracking device according to the child's cognitive task, attentional characteristics, and social interaction patterns. The accuracy of these adjustments often relies on human experience and intermittent observation of the child's state. Currently, the industry commonly uses head-mounted eye trackers in conjunction with basic fixation devices as the standard data acquisition method.

[0003] According to the paper "Eye-tracking data collection system based on portable mobile devices and a discriminant model for children with developmental disorders" (East China Normal University, 2024), "fixed headband devices have risks such as pressure regulation lag and local skin indentation. They also have fluctuations in wearing pressure, which can easily cause discomfort and resistance in children, making them prone to pulling on the headband device and resulting in poor accuracy of the detection data."

[0004] As can be seen from practical applications, existing head-mounted eye-tracking devices struggle to adjust the tightness to match children's tolerance levels. When monitoring multiple children, operators find it difficult to make timely adjustments. Therefore, the fundamental flaw in existing eye-tracking technologies for children lies in their insufficient ability to adapt to children's head features and the lack of adaptive adjustment and guidance design, leading to issues with tracking stability and data reliability. Therefore, it is necessary to propose a novel head-mounted safety eye-tracking device for children to address these problems. Summary of the Invention

[0005] To address the aforementioned issues, this invention provides a head-mounted eye-tracking device for real-time monitoring of a child's wearing status. By designing guided tasks, the device encourages children to adjust the device independently, preventing children from pulling on it due to delayed adjustments by operators. This ensures both wearing comfort and the stability and accuracy of eye-tracking data acquisition, thereby improving the accuracy of early screening data such as cognitive development assessments, attentional characteristic analysis, and autism spectrum disorders.

[0006] To achieve the above objectives, the technical solution of the present invention is as follows: A head-mounted eye-tracking device includes an eye mask worn on the head. The eye mask includes a main body, display lenses symmetrically mounted inside the main body, and straps symmetrically mounted at both ends of the main body. An adjustment component for adjusting the tightness of the straps is provided between adjacent straps. A thin-film pressure sensor layer for collecting pressure data of the straps on the head is mounted on the inner wall of the straps. An infrared camera and a near-infrared LED are mounted on the inner wall of the main body. A guidance and tracking system is also included.

[0007] The guidance tracking system includes an eye-tracking module, a computing module, and a guidance module.

[0008] The guidance module is used to control the display lenses to show different visual stimulus task animations.

[0009] The eye-tracking module is used to control the activation of near-infrared LEDs to illuminate the eyes when the child's eyes are focused on the visual stimulus task animation on the display lens, and to continuously acquire pupil-corneal reflection images of several eyes using an infrared camera, and send the pupil-corneal reflection images of the eyes to the computing module.

[0010] The calculation module receives the pupil-corneal reflection image sent by the eye-tracking module, extracts eye feature parameters from the pupil-corneal reflection image, and performs spatiotemporal correlation analysis between the visual stimulus task animation being displayed on the display lens and the eye feature parameters to generate attention evaluation indicators.

[0011] The calculation module is also used to set the pressure threshold range and receive pressure data collected by the thin-film pressure sensor layer; when the pressure data is not within the pressure threshold range, the calculation module generates a guide signal and sends the guide signal to the guide module.

[0012] The guidance module is also used to receive guidance signals and control the display lens to display adjustment task animations based on the guidance signals, guiding children to use the adjustment components to adjust the tightness of the straps.

[0013] The technical principles of the above solution are as follows:

[0014] The pressure data is collected in real time by a thin-film pressure sensor layer inside the strap. The calculation module compares this data with a preset pressure threshold to determine if the strap is properly fitted. If it is fitted, the guidance module outputs a visual stimulus task animation, and the eye-tracking module simultaneously activates the near-infrared LED and infrared camera to capture pupil-corneal reflection images and transmit them to the calculation module. The calculation module extracts eye feature parameters and correlates them with the spatiotemporal features of the animation to generate attention assessment indicators. If the pressure is not suitable, the calculation module generates a corresponding guidance signal, and the guidance module guides the child to adjust the strap independently by adjusting the task animation; once the pressure is within the acceptable range after adjustment, the stimulus task resumes.

[0015] The above approach has the following beneficial effects:

[0016] 1. This solution uses real-time pressure monitoring and animation-guided self-adjustment design to dynamically match the characteristics and tolerance of children's heads, avoid discomfort such as delayed pressure adjustment and local pressure marks, reduce the situation where children pull on the device, and improve wearing safety and compliance.

[0017] 2. This solution ensures the stability of the eye-tracking data acquisition environment through adaptive adjustment of the wearing status, effectively avoiding data deviations caused by loose or uncomfortable wearing, and improving the accuracy and reliability of data such as children's cognitive development assessment and early screening for autism spectrum disorder.

[0018] 3. This solution guides children to adjust the straps independently, reducing reliance on the operator's experience. It is especially suitable for batch testing scenarios with multiple children, reducing the operator's adjustment workload and improving testing efficiency. At the same time, the animated guidance format is in line with children's cognitive characteristics, further optimizing the user experience.

[0019] Furthermore, the adjustment assembly includes an adjustment box, a rotating shaft rotatably fitted on the inner side wall of the adjustment box, and a knob coaxially fixedly connected to the end of the rotating shaft away from the inner side wall of the adjustment box through the side wall of the adjustment box; a worm gear coaxially fixedly connected to the rotating shaft; a winding rod rotatably fitted on the inner side wall of the adjustment box, and a worm wheel coaxially fixedly connected to the winding rod, the worm wheel and the worm gear meshing; and the ends of the straps away from the main body all penetrate through the adjacent side wall of the adjustment box and extend into the adjustment box, and are all fixedly connected to the winding rod.

[0020] Beneficial effects: The worm gear meshing structure combines smooth transmission with self-locking characteristics, preventing the strap from loosening after adjustment and ensuring wearing stability. The knob operation is convenient and caters to children's self-adjustment needs; it can be tightened by turning forward or backward, adapting to children's operating abilities and reducing the difficulty of adjustment.

[0021] Furthermore, the straps are made of elastic fiber mesh fabric.

[0022] Beneficial effects: The elastic fiber mesh fabric combines elasticity and breathability, adapting to different children's head sizes and reducing localized pressure. The mesh structure enhances wearing comfort, reduces the risk of skin pressure marks, and improves the fit.

[0023] Furthermore, in the guidance module, the visual stimulus task animation includes:

[0024] Attention-maintaining task animation: A slowly moving animated character is generated on the display lens to test whether a child can continuously follow the animated character's gaze.

[0025] Note the task assignment animation: flashing patterns are generated simultaneously on both sides of the display lens to detect which side the child prioritizes looking at.

[0026] Social attention task animation: Generates a face image on a display lens to detect the proportion of a child's gaze at the eyes or mouth area of ​​the face image.

[0027] Visual search task animation: A target and several distractions are generated on the display lens. The speed and path of the child's search for the target are detected by searching among the distractions.

[0028] Beneficial effects: The multi-type visual stimulation tasks cover the core dimensions of children's attention, and can comprehensively test attention maintenance, allocation, social attention and search ability.

[0029] Furthermore, in the calculation module, the eye feature parameters include the pupil center coordinates and the eye rotation angle.

[0030] Beneficial effects: By selecting the pupil center coordinates and eye rotation angle as core features, the eye movement trajectory and fixation state can be obtained, enabling the calculation module to identify the pupil fixation point.

[0031] Furthermore, in the calculation module, the method for extracting ocular feature parameters from the pupil-corneal reflection image includes the following steps:

[0032] S101, Calculation of pupil center coordinates: An ellipse fitting algorithm is used to identify and fit the pupil region in the pupil-corneal reflection image, generating an elliptical pupil contour boundary, and the center of the ellipse is calculated as the pupil center coordinates. .

[0033] S102, Pupil Relative Position Calculation: Capture the coordinates of the corneal reflective point formed by near-infrared LED illumination of the cornea in the pupil-corneal reflection image, using the corneal reflective point as a reference coordinate point. Calculate the relative position of the pupil center coordinates in two consecutive pupil-corneal reflectance images. and The calculation formula is as follows:

[0034] (1).

[0035] (2).

[0036] in, , Let be the pupil center coordinates of the nth frame of the pupil-corneal reflection image. For the first The pupil center coordinates of the frame pupil-corneal reflection image; d is the fixed optical distance between the infrared camera and the child's cornea.

[0037] S103, Calculation of eyeball rotation angle: The eyeball rotation angle is calculated using a spatial geometry algorithm. and The calculation formula is as follows:

[0038] (3).

[0039] (4).

[0040] in, The angle of horizontal eye movement. The angle of vertical rotation of the eyeball.

[0041] Beneficial effects: By accurately locating the pupil center through an ellipse fitting algorithm, and combining the corneal reflective light point reference with spatial geometric calculations, the eyeball rotation angle can be solved.

[0042] Furthermore, in the calculation module, the method for generating attention evaluation metrics includes the following steps:

[0043] S201, Spatiotemporal Feature Calibration: Spatiotemporal feature calibration is performed on the visual stimulus task animation being displayed on the display lens. The calibration content includes: the screen area coordinates of the visual stimulus task animation, the movement trajectory of the animation elements, the task type label of the animation, and the standard gaze requirements of the corresponding task.

[0044] S202, Spatiotemporal correlation matching: Spatiotemporal correlation matching is performed between eye feature parameters and spatiotemporal feature calibration of visual stimulus task animation to calculate attention correlation parameters; attention correlation parameters include: the proportion of pupil fixation time on visual stimulus target, pupil fixation reaction delay time, the time spent by pupil to locate target in visual search task animation and the proportion of pupil fixation on the eye / mouth area of ​​face in social attention task animation.

[0045] S203, Quantitative scoring: Quantitatively score each attention-related parameter, and calculate the comprehensive attention score based on the quantitative scoring results using weighted averages.

[0046] Beneficial effects: Through spatiotemporal feature calibration, association matching, and quantitative scoring, precise correlation between visual stimulus tasks and eye features is achieved. Multi-dimensional attention association parameters and weighted comprehensive scores can objectively quantify children's attention levels.

[0047] Furthermore, in the calculation module, the guiding signals include a pressure-to-tightness rotary knob guiding signal, a pressure-to-looseness rotary knob guiding signal, and a pressure adaptation signal.

[0048] Beneficial effects: By subdividing the guidance signals into three categories, the accuracy of children's self-regulation is ensured, and the efficiency of adaptive regulation is improved.

[0049] Furthermore, the pressure-to-tighten rotary knob guidance signal is a signal generated by the calculation module when the pressure data collected by the thin-film pressure sensor layer exceeds the pressure threshold range.

[0050] The signal for guiding the rotary knob to loosen the pressure is a signal generated by the calculation module when the pressure data collected by the thin-film pressure sensor layer is less than the pressure threshold range.

[0051] The pressure adaptation signal is a signal generated by the calculation module when the pressure data collected by the thin-film pressure sensor layer is within the pressure threshold range.

[0052] Beneficial effects: It clarifies the generation conditions of various guidance signals, enabling a precise correspondence between pressure states and signal outputs. It simplifies the signal judgment logic of the calculation module, improving the timeliness and accuracy of signal generation.

[0053] Furthermore, in the guidance module, adjusting the task animation includes:

[0054] Animation for adjusting excessive pressure: When the guide module receives the excessive pressure rotary knob guidance signal, the guide module controls the display lens to show a red cartoon bubble model. When the pressure data approaches the threshold range, the color of the cartoon bubble model gradually changes from red to yellow.

[0055] Pressure too loose adjustment animation: When the guide module receives the pressure too loose rotary knob guidance signal, the guide module controls the display lens to display a blue cartoon bubble model. When the pressure data approaches the threshold range, the color of the cartoon bubble model gradually changes from blue to yellow.

[0056] Pressure Adaptation Confirmation Animation: When the guidance module receives the pressure adaptation signal, the guidance module controls the display lens to show a yellow cartoon bubble model smiley face. At the same time, the cartoon bubble model smiley face flashes for 3-5 seconds and then disappears, automatically switching to the visual stimulation task animation.

[0057] Beneficial effects: The animation design aligns with children's cognitive habits, the colors intuitively guide the adjustment direction, and the color gradients provide feedback on the adjustment progress. The adaptation confirmation animation offers clear feedback, helping children quickly complete the adaptation adjustment. Attached Figure Description

[0058] Figure 1 This is an isometric view of the head-mounted eye-tracking device of the present invention.

[0059] Figure 2 This is a front view of the main body of the head-mounted eye-tracking device of the present invention.

[0060] Figure 3 This is a frontal cross-sectional view of the adjustment box in the head-mounted eye-tracking device of the present invention.

[0061] Figure 4 This is a top sectional view of the adjustment box in the head-mounted eye-tracking device of the present invention.

[0062] Figure 5 This is a structural diagram of the guidance and tracking system in the head-mounted eye-tracking device of the present invention.

[0063] Figure 6 This diagram illustrates the steps of extracting ocular feature parameters from pupil-corneal reflection images in the head-mounted eye-tracking device of the present invention.

[0064] Figure 7 This diagram illustrates the steps of generating attention assessment metrics in the head-mounted eye-tracking device of the present invention.

[0065] The reference numerals in the accompanying drawings of the instruction manual include: 1. Main body; 2. Display lens; 3. Strap; 4. Infrared camera; 5. Near-infrared LED; 6. Adjustment box; 7. Rotating shaft; 8. Knob; 9. Worm gear; 10. Winding rod; 11. Worm wheel. Detailed Implementation

[0066] The technical solution of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0067] In the description of this invention, it should be noted that the terms "center," "upper," "lower," "left," "right," "vertical," "horizontal," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. Furthermore, the terms "first," "second," and "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0068] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0069] The following detailed description illustrates the specific implementation methods:

[0070] Implementation, for example Figure 1 and Figure 2The image shows a head-mounted eye-tracking device, comprising an eye mask worn on the head. The eye mask includes a main body 1, within which display lenses 2 are symmetrically mounted. Straps 3, made of elastic fiber mesh fabric, are symmetrically mounted at both ends of the main body 1. Adjustment components for adjusting the tightness of the straps 3 are provided between adjacent straps 3. A thin-film pressure sensor layer for collecting pressure data of the straps 3 on the head is mounted on the inner wall of the straps 3. An infrared camera 4 and a near-infrared LED 5 are mounted on the inner wall of the main body 1. A guidance and tracking system is also included.

[0071] like Figure 1 and Figure 3 As shown, specifically, the adjustment assembly includes an adjustment box 6, a rotating shaft 7 rotatably fitted on the inner side wall of the adjustment box 6, a knob 8 coaxially fixed to the left end of the rotating shaft 7 through the side wall of the adjustment box 6 by screws; a worm gear 9 is integrally formed coaxially on the rotating shaft 7.

[0072] like Figure 4 As shown, a winding rod 10 is rotatably fitted on the inner wall of the adjusting box 6. A worm gear 11 is coaxially fixed to the winding rod 10 by screws, and the worm gear 11 meshes with the worm 9.

[0073] The ends of the straps 3 away from the main body 1 all penetrate the side wall of the adjacent adjustment box 6 and extend into the adjustment box 6, and are all fixedly connected to the winding rod 10 by screws.

[0074] like Figure 2 and Figure 5 As shown, specifically, the guidance tracking system includes an eye-tracking module, a computing module, and a guidance module.

[0075] The guidance module is used to control the display lens 2 to display different visual stimulus task animations.

[0076] In the guidance module, the visual stimulus task animation includes:

[0077] Attention-maintaining task animation: A slowly moving animated character is generated on display lens 2 to test whether the child can continuously follow the animated character's gaze.

[0078] Note the task assignment animation: flashing patterns are generated simultaneously on both sides of display lens 2 to detect which side the child prioritizes looking at.

[0079] Social attention task animation: Generate a face image on display lens 2 to detect the proportion of a child's gaze at the eye or mouth area of ​​the face image.

[0080] Visual search task animation: A target and several distractions are generated on display lens 2. The child searches for the target among the distractions to detect the speed and path of the child's search for the target.

[0081] The eye-tracking module is used to control the near-infrared LED 5 to illuminate the eyeball when the child's eyes are focused on the visual stimulus task animation on the display lens 2, and to continuously acquire several pupil-corneal reflection images of the eyeball using the infrared camera 4, and send the pupil-corneal reflection images of the eyeball to the computing module.

[0082] The calculation module receives the pupil-corneal reflection image sent by the eye-tracking module, extracts eye feature parameters from the image, and performs spatiotemporal correlation analysis between the visual stimulus task animation displayed on the display lens 2 and the eye feature parameters to generate attention assessment indicators. The eye feature parameters include the pupil center coordinates and the eye rotation angle.

[0083] like Figure 6 As shown, specifically, the method for extracting ocular feature parameters from pupil-corneal reflection images includes the following steps:

[0084] S101, Calculation of pupil center coordinates: An ellipse fitting algorithm is used to identify and fit the pupil region in the pupil-corneal reflection image, generating an elliptical pupil contour boundary, and the center of the ellipse is calculated as the pupil center coordinates. .

[0085] S102, Pupil Relative Position Calculation: Capture the coordinates of the corneal reflective point formed by the near-infrared LED 5 illuminating the cornea in the pupil-corneal reflective image, using the corneal reflective point as a reference point. Calculate the relative position of the pupil center coordinates in two consecutive pupil-corneal reflectance images. and The calculation formula is as follows:

[0086] (1).

[0087] (2).

[0088] in, , Let be the pupil center coordinates of the nth frame of the pupil-corneal reflection image. For the first The pupil center coordinates of the frame pupil-corneal reflection image; d is the fixed optical distance between the infrared camera 4 and the child's cornea.

[0089] S103, Calculation of eyeball rotation angle: The eyeball rotation angle is calculated using a spatial geometry algorithm. and The calculation formula is as follows:

[0090] (3).

[0091] (4).

[0092] in, The angle of horizontal eye movement. The angle of vertical rotation of the eyeball.

[0093] like Figure 7 As shown, specifically, the method for generating attention evaluation metrics in the calculation module includes the following steps:

[0094] S201, Spatiotemporal feature calibration: Perform spatiotemporal feature calibration on the visual stimulus task animation being displayed on the display lens 2. The calibration content includes: the screen area coordinates of the visual stimulus task animation, the movement trajectory of the animation elements, the task type label of the animation, and the standard gaze requirements of the corresponding task.

[0095] S202, Spatiotemporal correlation matching: Spatiotemporal correlation matching is performed between eye feature parameters and spatiotemporal feature calibration of visual stimulus task animation to calculate attention correlation parameters; attention correlation parameters include: the proportion of pupil fixation time on visual stimulus target, pupil fixation reaction delay time, the time spent by pupil to locate target in visual search task animation and the proportion of pupil fixation on the eye / mouth area of ​​face in social attention task animation.

[0096] S203, Quantitative scoring: Quantitatively score each attention-related parameter, and calculate the comprehensive attention score based on the quantitative scoring results using weighted averages.

[0097] Combination Figure 1 and Figure 3 As shown, first, the child places the main body 1 of the eye mask against their eyes, aligning the display lens 2 with their eyes, and then wraps the elastic fiber mesh fabric strap 3 around their head. At this time, the thin film pressure sensor layer on the inner wall of the strap 3 collects the pressure data of the strap 3 on the head in real time. The child can adjust the tightness of the strap 3 through the adjustment component. Turning the knob 8 on the outside of the adjustment box 6 drives the rotating shaft 7 and the coaxially integrated worm gear 9 to rotate. Because the worm gear 9 meshes with the worm wheel 11 on the winding rod 10, it drives the winding rod 10 to rotate, thus tightening the strap 3 by wrapping it around the winding rod 10. Due to the self-locking characteristic of the meshing of the worm wheel 11 and the worm gear 9, the strap 3 will not loosen on its own after adjustment, and tightening can be achieved by turning the knob 8 in either the forward or reverse direction, which is suitable for children's operating habits.

[0098] Combination Figure 2As shown, in this embodiment, the device is factory-set to have an infrared camera 4 with a frame rate of 30 frames per second, an illumination intensity of 500 lux for the near-infrared LED 5, and a fixed optical distance d = 15 cm between the infrared camera 4 and the child's cornea. The eye-tracking module controls the near-infrared LED 5 to activate and illuminate the eye, aligning it with the eye area. At this time, the guidance module controls the display lens 2 to sequentially present different visual stimulation task animations:

[0099] In this embodiment, the guidance module controls the display lens 2 to sequentially display four types of visual stimulus task animations, each task lasting 60 seconds, with a 10-second rest between tasks.

[0100] Attention maintenance task animation: A slowly moving cartoon rabbit animation character is generated in the central area of ​​display lens 2, with a movement speed of 5cm / s (corresponding to the pixel movement speed on display lens 2), and the movement trajectory is a horizontal reciprocating motion.

[0101] Social attention task animation: A standardized frontal image of a human face is displayed in the center of display lens 2, with the image size occupying 30% of the area of ​​display lens 2, and the coordinate range of the eye region in the face image is clearly marked. ) and mouth area coordinate range .

[0102] During the execution of various visual stimulation tasks, the eye-tracking module synchronously controls the near-infrared LED 5 to start, illuminating the child's eyes. The infrared camera 4 continuously acquires pupil-corneal reflection images at a frame rate of 30 frames / second and sends the acquired pupil-corneal reflection images to the calculation module in real time. The calculation module extracts eye feature parameters according to the steps S101-S103 mentioned above and completes the calculation by combining formulas (1) to (4). The specific process is as follows:

[0103] S101, Pupil Center Coordinate Calculation: The calculation module uses an ellipse fitting algorithm to identify the pupil region contour for each frame of the pupil-corneal reflection image, and generates an elliptical pupil contour boundary through least squares fitting, thereby calculating the pupil center coordinates of each frame of the pupil-corneal reflection image. For example, the pupil center coordinates were obtained by fitting the pupil-corneal reflection image of frame 10. The pupil-corneal reflectance image of frame 11 was fitted to obtain ,in To display a 2-pixel unit for the lens, which is the factory default setting for the device. This corresponds to an actual visual angle of 0.01° and an actual distance of 0.1mm.

[0104] S102, Pupil Relative Position Calculation: The calculation module captures the coordinates of the corneal reflective spot formed by the near-infrared LED 5 illuminating the cornea in each frame of the pupil-corneal reflective image. The corneal reflective point is the reflection point of the cornea after being illuminated by the infrared LED 5 and projected into the infrared camera 4. It is a fixed reference point (since the relative positions of the near-infrared LED 5 and the infrared camera 4 are fixed, and the corneal position is also fixed, the coordinates of the corneal reflective point are constant after the wear is stabilized. In this embodiment, the coordinates of the corneal reflective point are ( = , = The relative movement distance of the pupil center in two consecutive pupil-corneal reflection images is calculated according to formulas (1) and (2).

[0105] In this embodiment, the pupil-corneal reflection images in frames 10 and 11 are analyzed. ; Converting pixel distance to actual spatial distance, then the actual .

[0106] S103, Calculation of eyeball rotation angle: [Calculate the actual...] Substituting the fixed optical distance d=15cm into formulas (3) and (4), the horizontal and vertical rotation angles of the eyeball are calculated using the arctangent function: The calculation results show that within 1 / 30th of a second between frames 10 and 11, the child's eyeballs rotated horizontally by about 0.076° and vertically by about 0.038°, corresponding to the movement trajectory of the cartoon rabbit character.

[0107] Spatiotemporal correlation analysis and attention evaluation index generation: The calculation module completes the analysis according to steps S201-S203, and generates attention evaluation indexes by combining the extracted eye feature parameters. The specific process is as follows:

[0108] S201, Spatiotemporal Feature Calibration: The currently displayed visual stimulus task animation is calibrated. In this embodiment, taking the attention maintenance task as an example, the movement trajectory coordinates of the cartoon rabbit animation character (a sequence of pixel coordinates that changes over time) are calibrated, the task type label is "attention maintenance", and the standard gaze requirement is "the pupil center follows the movement of the animation character, the gaze duration accounts for no less than 80%, and the reaction delay time does not exceed 100ms". Taking the social attention task animation as an example, the pixel coordinate range of the eye area and mouth area in the face image is calibrated, the task type label is "social attention", and the standard gaze requirement is "the gaze percentage of the eye area is no less than 60%".

[0109] S202, Spatiotemporal Association Matching: The extracted pupil center coordinates and eye rotation angle feature parameters are matched with the calibrated animation spatiotemporal features to calculate attention association parameters. For example, in an attention maintenance task, the duration for which the pupil center falls within the coordinate range of the cartoon rabbit animation character within 60 seconds is 46 seconds, so the fixation duration percentage = 46s / 60s ≈ 76.7%; by calculating the time difference between when the pupil center starts following the animation character's movement and when the animation character starts moving, the reaction delay time is obtained as 120ms; in a social attention task animation, the duration for which the pupil center falls in the eye area within 60 seconds is 32 seconds, and the duration for which it falls in the mouth area is 10 seconds, so the fixation percentage for the eye area = 32s / (32s+10s) ≈ 76.2%.

[0110] S203, Quantitative Scoring and Comprehensive Evaluation: Set the weights and scoring standards for various attention-related parameters: In this embodiment, the weight of gaze duration percentage is set at 0.4, with a full score of 100 (100 points for 80% or above, and 2 points deducted for each 1% decrease); the weight of reaction delay time is set at 0.3, with a full score of 100 (100 points for 100ms or below, and 5 points deducted for each 10ms increase); the weight of eye area gaze percentage in social attention task animation is set at 0.3, with a full score of 100 (100 points for 60% or above, and 3 points deducted for each 1% decrease).

[0111] Based on the above calculations, a quantitative score was calculated: The attention maintenance task's fixation time accounted for 76.7% of the total, 3.3% lower than the 80% standard, resulting in a score of 100 - 3.3 × 2 = 93.4 points; the reaction delay was 120ms, 20ms higher than the 100ms standard, resulting in a score of 100 - (20 ÷ 10) × 5 = 90 points; the social attention task's eye area fixation during animation accounted for 76.2% of the total, meeting the 60% standard, earning a score of 100. The weighted average attention score was 93.4 × 0.4 + 90 × 0.3 + 100 × 0.3 = 37.36 + 27 + 30 = 94.36 points, indicating that the child performed well in this test.

[0112] In some embodiments, the computing module includes a communication unit and a terminal APP client. The communication unit is used to send the comprehensive attention score to the terminal APP client for display via a 4G network.

[0113] Once the test is complete, the communication unit will send the comprehensive attention score to the terminal APP client (such as a computer or mobile phone) so that users can intuitively view the child's attention status.

[0114] like Figure 2As shown, specifically, the calculation module is also used to set the pressure threshold range and receive pressure data collected by the thin-film pressure sensor layer; when the pressure data is not within the pressure threshold range, the calculation module generates a guide signal and sends the guide signal to the guide module.

[0115] The guidance module is also used to receive guidance signals and control the display lens 2 to display the adjustment task animation based on the guidance signals, guiding the child to adjust the tightness of the strap 3 using the adjustment components.

[0116] The guidance signals include the pressure too tight rotary knob 8 guidance signal, the pressure too loose rotary knob 8 guidance signal, and the pressure adaptation signal. The pressure too tight rotary knob 8 guidance signal is a signal generated by the calculation module when the pressure data collected by the thin-film pressure sensor layer exceeds the pressure threshold range.

[0117] The signal from rotary knob 8, indicating that the pressure is too loose, is generated by the calculation module when the pressure data collected by the thin-film pressure sensor layer is less than the pressure threshold range.

[0118] The pressure adaptation signal is a signal generated by the calculation module when the pressure data collected by the thin-film pressure sensor layer is within the pressure threshold range.

[0119] Specifically, in the guidance module, adjusting the task animation includes:

[0120] Animation for adjusting excessive pressure: When the guide module receives the guidance signal from the excessive pressure rotary knob 8, the guide module controls the display lens 2 to display a red cartoon bubble model. When the pressure data approaches the threshold range, the color of the cartoon bubble model gradually changes from red to yellow.

[0121] Pressure too loose adjustment animation: When the guide module receives the pressure too loose rotary knob 8 guide signal, the guide module controls the display lens 2 to display a blue cartoon bubble model. When the pressure data approaches the threshold range, the color of the cartoon bubble model gradually changes from blue to yellow.

[0122] Pressure Adaptation Confirmation Animation: When the guidance module receives the pressure adaptation signal, the guidance module controls the display lens 2 to display a yellow cartoon bubble model smiley face. At the same time, the cartoon bubble model smiley face flashes for 3-5 seconds and then disappears, automatically switching to the visual stimulation task animation.

[0123] Combination Figure 1 and Figure 2 As shown, the calculation module presets a pressure threshold range and receives pressure data collected in real time from the thin-film pressure sensor layer inside the strap 3. If the collected pressure data exceeds the pressure threshold range, the calculation module immediately generates a corresponding guidance signal and sends it to the guidance module; if the pressure data is within the pressure threshold range, a pressure adaptation signal is generated to ensure that the test is performed under stable wearing conditions. The specific adjustment guidance process is as follows:

[0124] In the case of excessive pressure: if the pressure data collected by the thin-film pressure sensor layer exceeds the pressure threshold range, the calculation module generates a "pressure too tight rotation knob 8 guidance signal" and sends it to the guidance module. Upon receiving the signal, the guidance module immediately pauses the current visual stimulus task animation and controls the display lens 2 to display a red cartoon bubble model. As the child rotates knob 8 on the adjustment box 6 according to the animation guidance, the thin-film pressure sensor layer provides real-time feedback on the pressure data. When the pressure data drops to the pressure threshold range, the color of the cartoon bubble model gradually changes from red to yellow. If the child rotates it in the opposite direction to continue tightening, the pressure data continues to rise, and the color of the corresponding red cartoon bubble model continues to deepen. When the pressure data is within the pressure threshold range, the calculation module generates a pressure adaptation signal, and the guidance module immediately switches to displaying a pressure adaptation confirmation animation—a yellow cartoon bubble model smiley face. The smiley face flashes for 4 seconds and then disappears automatically, synchronously resuming the paused visual stimulus task animation to ensure test continuity.

[0125] In the case of excessively loose pressure: if the pressure data collected by the thin-film pressure sensor layer is less than the pressure threshold range, the calculation module generates a "pressure too loose rotation knob 8 guidance signal" and sends it to the guidance module. The guidance module controls the display lens 2 to display the pressure too loose adjustment animation—a blue cartoon bubble model. The child follows the animation guidance to turn knob 8, the winding rod 10 tightens the strap 3, and the pressure data gradually increases; when the pressure data is within the pressure threshold range, the blue bubble turns into a yellow smiley face confirmation animation, and after flashing, the visual stimulation task resumes.

[0126] This solution's head-mounted eye-tracking device not only assesses children's attention through visual stimulation tasks but also ensures wearing stability through a pressure monitoring and guided adjustment mechanism. 4G communication enables remote transmission and visualization of test results. It allows for real-time monitoring of the child's wearing status and guides children to adjust the device independently through designed tasks, preventing children from pulling on the device due to delayed adjustments by operators. This ensures both wearing comfort and the stability and accuracy of eye-tracking data acquisition, thereby improving the accuracy of early screening data such as children's cognitive development assessments, attention characteristic analysis, and autism spectrum disorders.

[0127] Obviously, the above embodiments are merely illustrative examples for clear explanation and are not intended to limit the implementation. Those skilled in the art will recognize that other variations or modifications can be made based on the above description. It is neither necessary nor possible to exhaustively list all possible implementations here. However, obvious variations or modifications derived therefrom are still within the scope of protection of this invention.

Claims

1. A head-mounted eye-tracking device, comprising an eye mask worn on the head, the eye mask comprising a main body (1), display lenses (2) symmetrically mounted inside the main body (1), and straps (3) symmetrically mounted at both ends of the main body (1), characterized in that, It also includes a guidance tracking system; An adjustment component for adjusting the tightness of the straps (3) is provided between adjacent straps (3); a thin film pressure sensor layer for collecting pressure data of the straps (3) on the head is installed on the inner wall of the straps (3); an infrared camera (4) and a near-infrared LED (5) are installed on the inner wall of the main body (1). The guidance and tracking system includes an eye-tracking module, a computing module, and a guidance module; The guidance module is used to control the display lens (2) to display different visual stimulus task animations; The eye-tracking module is used to control the near-infrared LED (5) to illuminate the eyeball when the child's eyeball is focused on the visual stimulation task animation on the display lens (2), and to continuously acquire several pupil-corneal reflection images of the eyeball using the infrared camera (4), and send the pupil-corneal reflection images of the eyeball to the computing module. The calculation module is used to receive the pupil-corneal reflection image sent by the eye tracking module, extract the eye feature parameters from the pupil-corneal reflection image, and perform spatiotemporal correlation analysis between the visual stimulus task animation being displayed on the display lens (2) and the eye feature parameters to generate attention evaluation indicators. The calculation module is also used to set the pressure threshold range and receive pressure data collected by the thin-film pressure sensor layer; when the pressure data is not within the pressure threshold range, the calculation module generates a guide signal and sends the guide signal to the guide module. The guidance module is also used to receive guidance signals and control the display lens (2) to display adjustment task animation based on the guidance signals, guiding the child to adjust the tightness of the strap (3) using the adjustment components.

2. The head-mounted eye-tracking device according to claim 1, characterized in that, The adjustment assembly includes an adjustment box (6), a rotating shaft (7) is rotatably fitted on the inner wall of the adjustment box (6), and a knob (8) is coaxially fixedly connected to the end of the rotating shaft (7) away from the inner wall of the adjustment box (6); a worm gear (9) is coaxially fixedly connected to the rotating shaft (7). A winding rod (10) is rotatably fitted on the inner wall of the adjusting box (6). A worm wheel (11) is coaxially fixedly connected to the winding rod (10). The worm wheel (11) meshes with the worm (9). The end of the strap (3) away from the main body (1) extends through the side wall of the adjacent adjustment box (6) into the adjustment box (6) and is fixedly connected to the winding rod (10).

3. The head-mounted eye-tracking device according to claim 2, characterized in that, The strap (3) is made of elastic fiber mesh fabric.

4. The head-mounted eye-tracking device according to claim 3, characterized in that, In the guidance module, the visual stimulus task animation includes: Attention-maintaining task animation: A slowly moving animated character is generated on the display lens (2) to test whether the child can continuously follow the animated character's gaze; Note the task assignment animation: flashing patterns are generated simultaneously on both sides of the display lens (2) to detect which side the child prioritizes looking at; Social attention task animation: A face image is generated on the display lens (2) to detect the proportion of a child’s gaze at the eye or mouth area on the face image; Visual search task animation: A target and several distractions are generated on the display lens (2), and the target is searched among the distractions to detect the speed and path of the child's search for the target.

5. The head-mounted eye-tracking device according to claim 4, characterized in that, In the calculation module, the eye feature parameters include the pupil center coordinates and the eye rotation angle.

6. The head-mounted eye-tracking device according to claim 5, characterized in that, In the calculation module, the method for extracting ocular feature parameters from the pupil-corneal reflection image includes the following steps: S101, Calculation of pupil center coordinates: An ellipse fitting algorithm is used to identify and fit the pupil region in the pupil-corneal reflection image, generating an elliptical pupil contour boundary, and the center of the ellipse is calculated as the pupil center coordinates. ; S102, Pupil Relative Position Calculation: Capture the coordinates of the corneal reflective point formed by the near-infrared LED (5) illuminating the cornea in the pupil-corneal reflective image, using the corneal reflective point as the reference coordinate. Calculate the relative position of the pupil center coordinates in two consecutive pupil-corneal reflectance images. and The calculation formula is as follows: （1）； （2）； in, , Let be the pupil center coordinates of the nth frame of the pupil-corneal reflection image. For the first The pupil center coordinates of the frame pupil-corneal reflection image; d is the fixed optical distance between the infrared camera (4) and the child's cornea; S103, Calculation of eyeball rotation angle: The eyeball rotation angle is calculated using a spatial geometry algorithm. and The calculation formula is as follows: （3）； （4）； in, The angle of horizontal eye movement. The angle of vertical rotation of the eyeball.

7. The head-mounted eye-tracking device according to claim 6, characterized in that, In the calculation module, the method for generating attention evaluation metrics includes the following steps: S201, Spatiotemporal feature calibration: Spatiotemporal feature calibration is performed on the visual stimulus task animation being displayed on the display lens (2). The calibration content includes: the screen area coordinates of the visual stimulus task animation, the movement trajectory of the animation elements, the task type label of the animation, and the standard gaze requirements of the corresponding task. S202, Spatiotemporal correlation matching: The eye feature parameters are spatiotemporally correlated with the spatiotemporal feature calibration of the visual stimulus task animation to calculate the attention correlation parameters. The attention correlation parameters include: the proportion of pupil fixation time on the visual stimulus target, the pupil fixation reaction delay time, the time spent by the pupil to locate the target in the visual search task animation, and the proportion of pupil fixation on the eye / mouth area of ​​the face in the social attention task animation. S203, Quantitative scoring: Quantitatively score each attention-related parameter, and calculate the comprehensive attention score based on the quantitative scoring results using weighted averages.

8. The head-mounted eye-tracking device according to claim 7, characterized in that, In the calculation module, the guiding signals include the pressure too tight rotary knob (8) guiding signal, the pressure too loose rotary knob (8) guiding signal and the pressure adaptation signal.

9. The head-mounted eye-tracking device according to claim 8, characterized in that, The pressure is too tight. The rotary knob (8) guide signal is the signal generated by the calculation module when the pressure data collected by the thin film pressure sensor layer is greater than the pressure threshold range; the pressure is too loose. The rotary knob (8) guide signal is the signal generated by the calculation module when the pressure data collected by the thin film pressure sensor layer is less than the pressure threshold range; the pressure adaptation signal is the signal generated by the calculation module when the pressure data collected by the thin film pressure sensor layer is within the pressure threshold range.

10. The head-mounted eye-tracking device according to claim 9, characterized in that, In the guidance module, adjusting the task animation includes: Animation for adjusting pressure too tight: When the guide module receives the guidance signal from the pressure too tight rotary knob (8), the guide module controls the display lens (2) to display a red cartoon bubble model. When the pressure data approaches the threshold range, the color of the cartoon bubble model gradually changes from red to yellow. Pressure too loose adjustment animation: When the guide module receives the pressure too loose rotary knob (8) guide signal, the guide module controls the display lens (2) to display a blue cartoon bubble model. When the pressure data approaches the threshold range, the color of the cartoon bubble model gradually changes from blue to yellow. Pressure Adaptation Confirmation Animation: When the guidance module receives the pressure adaptation signal, the guidance module controls the display lens (2) to display a yellow cartoon bubble model smiley face. At the same time, the cartoon bubble model smiley face flashes for 3-5 seconds and then disappears, automatically switching to the visual stimulation task animation.

Images