Myopia protection intelligent glasses for judging eye health based on scene recognition

Abstract

The invention relates to the technical field of intelligent glasses, in particular to myopia protection intelligent glasses for judging eye health based on scene recognition, which are characterized in that a visual space of the intelligent glasses is divided into an internal visual space and an edge visual space, and light in the internal visual space and the edge visual space is recognized; determining a central light characteristic value and an edge light characteristic value, and combining the real-time scene with the central light characteristic value and the edge light characteristic value to determine a light compensation value; according to the invention, by setting the internal visual space and the edge visual space, a basis is provided for zoning processing of light in different areas, and the light condition in the visual space can be analyzed and adjusted more meticulously, so that the intelligent glasses can carry out light analysis and adjustment according to different actual scenes, and the user experience is improved. And the adaptability of the intelligent glasses to various complex environments is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of smart glasses, and in particular to myopia protection smart glasses based on scene recognition to judge eye health. Background Art

[0002] Myopia, hyperopia, astigmatism, amblyopia and other eye diseases can all lead to poor vision among teenagers. Among them, myopia is the main factor leading to poor vision among teenagers and has become a major health problem faced by teenagers in Asian countries or regions.

[0003] Prior art CN117076995A discloses a method for judging eye status applied to myopia protection smart glasses, including a glasses frame, and built-in sensor components, a microprocessor and memory, a protection component, a battery and corresponding circuits; the sensor components include but are not limited to wearing status sensors, acceleration and angular velocity sensors; the built-in intelligent algorithm software is solidified in the memory and runs on the microprocessor. The implementation method is to confirm that the glasses are in a wearing state through the wearing status sensor; the acceleration and angular velocity sensors continuously acquire a number of real-time data at fixed time intervals, calculate digital features through a microprocessor, and determine the eye status based on the calculation results of the acceleration and angular velocity numerical digital features;

[0004] However, in the process of identifying light, when there are multiple areas with different light in the visual space environment, if the myopia protective glasses cannot actively distinguish the central position of vision, the light between the strong light area and the weak light area will affect each other, which will lead to inaccurate light adjustment of the myopia protective glasses, reducing the intelligence and adjustment accuracy of the myopia protective glasses. Summary of the invention

[0005] The purpose of the present invention is to solve the problems in the background technology and to propose a pair of myopia protection smart glasses based on scene recognition to judge the health of the eyes.

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] A myopia protection smart glasses for judging eye health based on scene recognition, comprising:

[0008] The information collection terminal is used to collect the basic information of the wearer and transmit it to the light analysis terminal and the posture analysis terminal;

[0009] The light recognition end, based on the two rays emitted by the light detection device, marks the point where each ray falls on the object as an endpoint, and then sets a first edge plane, a second edge plane and a third edge plane based on the individual rays and the endpoints, and then divides the visual space of the smart glasses into an internal visual space and an edge visual space based on the first edge plane, the second edge plane and the third edge plane;

[0010] The maximum value of the light brightness value in the internal visual space is detected and marked as the central light feature value, and then the light brightness value in the edge visual space is detected. At the same time, based on the central light feature value, the light brightness value in the edge visual space is processed, and then the processed light brightness value is averaged to obtain the edge light feature value. After that, the light recognition end and the light analysis end are connected in a one-way communication manner;

[0011] The scene recognition end is used to recognize the real-time scene and transmit it to the light analysis end;

[0012] A light analysis end is used to receive a real-time scene, obtain a standard brightness value and a comfortable contrast value of the real-time scene, and perform a difference calculation between a central light characteristic value and an edge light characteristic value to obtain an absolute difference. Based on the absolute difference, a light consistency signal and a light difference adjustment signal are determined. When a light consistency signal is generated, the absolute difference is compared with the standard brightness value to determine a light compensation value. When a light difference adjustment signal is generated, a light compensation value is determined based on the central light characteristic value, the edge light characteristic value and the comfortable contrast value. Then, the light analysis end transmits the light compensation value to the parameter adjustment end.

[0013] The parameter adjustment end is used to adjust the light of the smart glasses according to the light compensation value.

[0014] As a further solution of the present invention, the method for determining the internal visual space and the edge visual space includes:

[0015] A light detection device is arranged on the smart glasses, wherein the light detection device is arranged as a matrix optical distance sensor;

[0016] The light detection device emits light to the outside, identifies the object to which the light is directed, and marks the point where the light falls on the object as an endpoint. The two endpoints are then connected to obtain a local line segment, wherein the ray emitted by the light detection device is perpendicular to the surface of the smart glasses;

[0017] Acquire the positions of two rays at the same time, construct two planes based on the individual rays, and mark them as the first edge plane and the second edge plane respectively, and then construct another plane based on the local line segment, and mark it as the third edge plane, wherein the first edge plane and the second edge plane are parallel to each other, and the third edge plane is parallel to the smart glasses, and further, the third edge plane is between the first edge plane and the second edge plane, and the third edge plane is perpendicular to the first edge plane and the second edge plane;

[0018] Select a first edge plane, identify the position of the ray on the first edge plane, take the ray position as the midpoint of the height threshold X1, and set a height edge point based on the height threshold X1, wherein the height edge point includes an upper edge point and a lower edge point, and then set an upper edge line and a lower edge line at the positions of the upper edge point and the lower edge point, respectively, wherein the upper edge line and the lower edge line are both on the first edge plane and are parallel to the ray on the first edge plane. Similarly, the upper edge line and the lower edge line are also set on the second plane according to the above method;

[0019] Connect the upper edge lines between the two planes to form an upper edge plane, and connect the lower edge lines between the two planes to form a lower edge plane;

[0020] The space enclosed by the first edge plane, the second edge plane, the third edge plane, the upper edge plane, the lower edge plane and the smart glasses is marked as the internal visual space, and the space outside the internal vision is marked as the edge visual space.

[0021] As a further solution of the present invention, when the two endpoints are not on the same plane, that is, the length distances of the two rays are inconsistent, the ray with the shorter length distance will be extended according to the original ray path so that the length distances of the two rays are consistent. At this time, the original ray with the shorter length distance will reset the endpoints, and then the two endpoints will be reconnected to obtain a local line segment, wherein the local line segment is parallel to the smart glasses.

[0022] As a further solution of the present invention, a method for determining the central light characteristic value and the edge light characteristic value includes:

[0023] Detecting the light brightness value in the internal visual space, identifying the maximum value of the light brightness in the internal visual space, and marking it as the central light characteristic value of the internal visual space;

[0024] Acquire the edge visual space, and detect the light brightness value in the edge visual space;

[0025] In the process of detecting the light brightness value in the edge visual space, the edge visual space is converted into a plane edge area according to the wearer's visual direction. Furthermore, each point in the plane edge area has a corresponding light brightness value.

[0026] Obtain the central light characteristic value of the internal visual space, compare the central light characteristic value with the light brightness value in the plane edge area, and if there is a light brightness value in the plane edge area that is consistent with the central light characteristic value, set a blank area to replace the area corresponding to the light brightness value in the plane edge area;

[0027] After the comparison of the light brightness values ​​in the plane edge area is completed, the light brightness values ​​in the plane edge area are detected again, and the average value of the light brightness values ​​in the plane edge area is calculated, and the obtained result is marked as the edge light characteristic value;

[0028] When calculating the edge light characteristic value, the light brightness value corresponding to the blank area in the plane edge area does not participate in the average value calculation process.

[0029] As a further solution of the present invention, a method for determining a light consistency signal and a light difference adjustment signal includes:

[0030] According to the real-time scene, a standard brightness value and a comfort contrast value corresponding to the real-time scene are obtained, wherein the comfort contrast value indicates a comfortable ratio between dark and bright light, and the comfort contrast values ​​in different activity scenes are different;

[0031] Obtain the wearer's light environment, first subtract the edge light characteristic value from the central light characteristic value in the light environment to obtain the light difference, take the absolute value of the light difference, and mark it as the absolute difference;

[0032] The absolute difference is compared with the difference threshold value X2. If the absolute difference is less than or equal to the difference threshold value X2, a light consistency signal is generated. Otherwise, if the absolute difference is greater than the difference threshold value X2, a light difference adjustment signal is generated.

[0033] As a further solution of the present invention, a method for determining the compensation light value includes:

[0034] When a light consistency signal is detected, the central light characteristic value is extracted, and the central light characteristic value is subtracted from the standard brightness value, and then the result is divided by the central light characteristic value to obtain the compensation light value;

[0035] When the light difference adjustment signal is detected, the central light characteristic value and the edge light characteristic value are obtained again, and the central light characteristic value is divided by the edge light characteristic value to obtain the light ratio value, and then the light ratio value is subtracted from the comfort contrast value to obtain the compensation light value.

[0036] As a further solution of the present invention, when the parameter adjustment end adjusts the light of the smart glasses according to the compensation light value, the light compensation value includes a positive sign and a negative sign. When the sign in the light compensation value is positive, it indicates that the light transmittance effect of the smart glasses needs to be enhanced. When the sign in the light compensation value is negative, it indicates that the light transmittance effect of the smart glasses needs to be reduced.

[0037] As a further solution of the present invention, it also includes a posture recognition terminal, a posture analysis terminal and an intelligent reminder terminal;

[0038] The posture recognition end is used to set the visual center line according to the emission point of the light detection device, and obtain the center distance, then set the spatial coordinate system, collect the pitch angle between the visual center line and the spatial coordinate system, and transmit it to the posture analysis end;

[0039] The posture analysis end is used to collect the point distance in the internal visual space, compare the point distance with the center distance, and analyze the comparison result to determine the real eye distance, then generate a reminder signal according to the real eye distance, and transmit the reminder signal to the intelligent reminder end;

[0040] The intelligent reminder terminal generates corresponding sound and light reminder information based on the received reminder signal, and uses the terminal device to provide real-time reminders to the wearer.

[0041] As a further solution of the present invention, a method for determining the center distance and the pitch angle includes:

[0042] A multi-axial acceleration sensor is also provided on the smart glasses. The multi-axial acceleration sensor is used to detect the motion state of the wearer, and the multi-axial acceleration sensor is composed of at least a three-axial linear acceleration sensor. The multi-axial acceleration sensor is provided in the middle position of the smart glasses.

[0043] Set the spatial coordinate system according to the multi-axial acceleration sensor, then identify the emission point of the light detection device, connect the emission points to obtain the starting line segment, take the midpoint position of the starting line segment, and mark the midpoint position as the line of sight center;

[0044] Based on the rays emitted by the light detection device, a visual center line is also set at the center of the line of sight, wherein the visual center line and the two rays emitted by the light detection device are simultaneously on the same plane and are parallel to each other;

[0045] The distance length of the visual center line is collected and marked as the center distance. Based on the spatial coordinate system in the multi-axial acceleration sensor, the angle between the visual center line and the xoy plane in the spatial coordinate system is obtained and marked as the pitch angle θ.

[0046] As a further solution of the present invention, a method for determining the actual eye distance includes:

[0047] SS1: Acquire the internal visual space and collect the point distances in the internal visual space based on the matrix optical distance sensor;

[0048] SS2: Compare the point distances in the internal visual space with the center distance. If all point distances in the internal visual space are greater than the center distance, identify the minimum value in the point distances and mark it as the true eye distance d.

[0049] If there is a point distance less than or equal to the center distance, extract the pitch angle θ and the spatial coordinate system, and determine the component θ of the pitch angle θ in the spatial coordinate system xz , where θ xz is the pitch angle θ and the component on the xoz plane of the spatial coordinate system;

[0050] Identify the xoz plane and set the threshold angle θ in the fourth quadrant of the xoz plane xz-th , the component θ xz With the threshold angle θ xz-th For comparison, if θ xz <θ xz-th When , the visual center line is rotated downward by an angle α1, with the visual center fixed at the visual center. If θ xz-th ≤θ xz When α2=k2×(-θ xz ), when θ xz ≥0, the visual center line is rotated downward by an angle α3, α3 = k3 × θ xz , where k1, k2 and k3 are all reduction ratios;

[0051] Based on the center distance, the offset angle α and the pitch angle θ, the actual eye distance d is calculated using the trigonometric formula, and the offset angle α refers to α1, α2 and α3;

[0052] SS3: Compare the actual eye distance d with the standard sight interval Db. If d∈Db, a normal eye signal is generated. When the actual eye distance d is less than the minimum value of the standard visual range, the generated reminder signal is a close-distance eye signal; if the actual eye distance d is greater than the maximum value of the standard visual range, the generated reminder signal is a long-distance eye signal.

[0053] Compared with the prior art, the advantages of the present invention are:

[0054] The present invention sets a light detection device, and sets a first edge plane, a second edge plane, and a third edge plane based on the rays emitted by the light detection device and the endpoints of each ray falling on the object. Based on the first edge plane, the second edge plane, and the third edge plane, the visual space of the smart glasses is divided into an internal visual space and an edge visual space, which provides a basis for the separate processing of light in different areas, and helps to analyze and adjust the light conditions in the visual space more carefully. Then, the light in the internal visual space and the edge visual space is identified, and the central light characteristic value and the edge light characteristic value are determined. At the same time, the real-time scene is combined with the central light characteristic value and the edge light characteristic value to determine the light compensation value, so that the smart glasses can analyze and adjust the light according to different actual scenes, thereby improving the adaptability of the smart glasses to various complex environments.

[0055] The present invention also calculates the wearer's actual eye distance, measures and analyzes the pitch angle, and sets different offset angles accordingly. The wearer's final actual eye distance is determined by the offset angle, which can more accurately evaluate the wearer's visual fatigue state and further improve the adaptability of the myopia protective glasses and the accuracy of fatigue monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] Figure 1 It is a schematic diagram of the system structure of the present invention;

[0057] Figure 2 This is a schematic diagram for demonstrating the actual eye distance of the present invention. DETAILED DESCRIPTION

[0058] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0059] Embodiment 1, refer to Figure 1 , a myopia protection smart glasses based on scene recognition to judge eye health, including an information collection end, a light recognition end, a scene recognition end, a light analysis end and a parameter adjustment end;

[0060] The information collection terminal is used to collect basic information of the wearer, wherein the basic information includes the identity information of the wearer and the health information corresponding to the identity. Furthermore, the identity information includes the age, gender and occupation of the wearer, and the health information refers to the healthy eye data corresponding to the age stage of the wearer, including blinking frequency and healthy eye posture, etc., and then the information collection terminal transmits the collected basic information to the light analysis terminal;

[0061] The light recognition end is used to recognize the light environment of the wearer, wherein the light environment includes characteristic light values ​​in the internal visual space and characteristic light values ​​in the edge visual space. Specifically, the light environment recognition method includes:

[0062] A light detection device is provided on the smart glasses. In this embodiment, the light detection device is provided as a matrix optical distance sensor. It should be further explained that one light detection device is provided at each of the left and right ends of the smart glasses, and the two light detection devices are used to determine the real-time line of sight center of the wearer;

[0063] The light detection device emits light to the outside, identifies the object to which the light is directed, and marks the point where the light falls on the object as an endpoint. The two endpoints are then connected to obtain a local line segment, wherein the ray emitted by the light detection device is perpendicular to the surface of the smart glasses;

[0064] Acquire the positions of two rays at the same time, construct two planes based on the individual rays, and mark them as the first edge plane and the second edge plane respectively, and then construct another plane based on the local line segment, and mark it as the third edge plane, wherein the first edge plane and the second edge plane are parallel to each other, and the third edge plane is parallel to the smart glasses, and further, the third edge plane is between the first edge plane and the second edge plane, and the third edge plane is perpendicular to the first edge plane and the second edge plane;

[0065] It should be further explained that when the two endpoints are not on the same plane, that is, when the lengths of the two rays are inconsistent, the ray with the shorter length is extended according to the original ray path to make the lengths of the two rays consistent. At this time, the endpoint of the original ray with the shorter length will be reset, and then the two endpoints will be reconnected to obtain a local line segment, where the local line segment is parallel to the smart glasses;

[0066] Select a first edge plane, identify the position of the ray on the first edge plane, take the ray position as the midpoint of the height threshold X1, and set a height edge point based on the height threshold X1, wherein the height edge point includes an upper edge point and a lower edge point, and then set an upper edge line and a lower edge line at the positions of the upper edge point and the lower edge point, respectively, wherein the upper edge line and the lower edge line are both on the first edge plane and are parallel to the ray on the first edge plane. Similarly, the upper edge line and the lower edge line are also set on the second plane according to the above method, wherein the specific value of the height threshold X1 is set by those skilled in the art according to big data experience;

[0067] Then, the upper edge lines between the two planes are connected to form an upper edge plane, and the lower edge lines between the two planes are connected to form a lower edge plane;

[0068] Mark the space enclosed by the first edge plane, the second edge plane, the third edge plane, the upper edge plane, the lower edge plane and the smart glasses as the internal visual space, and mark the space outside the internal vision as the edge visual space;

[0069] Detecting the light brightness value in the internal visual space, identifying the maximum value of the light brightness in the internal visual space, and marking it as the central light characteristic value of the internal visual space;

[0070] Acquire the edge visual space, and detect the light brightness value in the edge visual space, wherein, in the process of detecting the light brightness value in the edge visual space, the edge visual space is converted into a plane edge area according to the visual direction of the wearer, and further, each point in the plane edge area has a corresponding light brightness value;

[0071] Then, the central light characteristic value of the internal visual space is obtained, and the central light characteristic value is compared with the light brightness value in the plane edge area. If there is a light brightness value in the plane edge area that is consistent with the central light characteristic value, a blank area is set as the area corresponding to the light brightness value in the plane edge area to replace it;

[0072] After the comparison of the light brightness values ​​in the plane edge area is completed, the light brightness values ​​in the plane edge area are detected again, and the average value of the light brightness values ​​in the plane edge area is calculated, and the obtained result is marked as the edge light characteristic value;

[0073] When calculating the edge light characteristic value, the light brightness value corresponding to the blank area in the plane edge area does not participate in the average value calculation process;

[0074] The light recognition end then transmits the wearer's light environment to the light analysis end;

[0075] The scene recognition end is used to recognize the real-time scene of the smart glasses, where the real-time scene refers to the wearer's current activity scene, including using electronic devices, watching TV, reading paper data, working or studying on a computer, etc. The scene recognition end then transmits the real-time scene to the light analysis end;

[0076] The light analysis end is used to identify the real-time scene, identify the real-time scene with the wearer's light environment, and determine the light compensation value. The specific method for determining the light compensation value includes:

[0077] According to the real-time scene, a standard brightness value and a comfortable contrast value corresponding to the real-time scene are obtained, wherein the comfortable contrast value indicates a comfortable ratio between dark and bright light, and the comfortable contrast values ​​in different activity scenes are different. Specifically, the comfortable contrast value in each activity scene is obtained by a person skilled in the art after big data calculation;

[0078] Then, the wearer's light environment is obtained, and the central light characteristic value in the light environment is first subtracted from the edge light characteristic value to obtain the light difference value, and the absolute value of the light difference value is taken and marked as the absolute difference value;

[0079] The absolute difference is compared with the difference threshold value X2. If the absolute difference is less than or equal to the difference threshold value X2, a light consistency signal is generated. Otherwise, if the absolute difference is greater than the difference threshold value X2, a light difference adjustment signal is generated. The specific value of the difference threshold value X2 is obtained by a person skilled in the art after big data calculation.

[0080] When a light consistency signal is detected, the central light characteristic value is extracted, and the central light characteristic value is subtracted from the standard brightness value, and then the result is divided by the central light characteristic value to obtain the compensation light value;

[0081] When the light difference adjustment signal is detected, the central light characteristic value and the edge light characteristic value are obtained again, and the central light characteristic value is divided by the edge light characteristic value to obtain a light ratio value, and then the light ratio value is subtracted from the comfort contrast value to obtain a compensation light value;

[0082] Afterwards, the light analysis end transmits the compensation light value to the parameter adjustment end, and the parameter adjustment end adjusts the light of the smart glasses according to the compensation light value. Furthermore, the light compensation value includes a positive sign and a symbol. When the symbol in the light compensation value is a positive sign, it indicates that the light transmission effect of the smart glasses needs to be enhanced. When the symbol in the light compensation value is a negative sign, it indicates that the light environment is strong, and the light transmission effect of the smart glasses needs to be reduced.

[0083] Embodiment 2: Based on Embodiment 1, the present invention is different from Embodiment 1 in that the present embodiment further includes a gesture recognition terminal, a gesture analysis terminal, and an intelligent reminder terminal.

[0084] The gesture recognition terminal is used to collect the real-time actions of the wearer, wherein the method for collecting the real-time actions of the wearer includes:

[0085] A multi-axial acceleration sensor is also provided on the smart glasses. The multi-axial acceleration sensor is used to detect the motion state of the wearer, and the multi-axial acceleration sensor is composed of at least three-axial linear acceleration sensors. In this embodiment, the multi-axial acceleration sensor is provided in the middle position of the smart glasses;

[0086] Set up a spatial coordinate system according to the multi-axial acceleration sensor, then identify the emission point of the light detection device, connect the emission points to obtain the starting line segment, take the midpoint position of the starting line segment, and mark the midpoint position as the line of sight center;

[0087] Based on the rays emitted by the light detection device, a visual center line is also set at the line of sight center position. Among them, the visual center line and the two rays emitted by the light detection device are in the same plane and parallel to each other;

[0088] Collect the distance length of the visual center line and mark it as the center distance. Then, based on the spatial coordinate system in the multi-axial acceleration sensor, obtain the angle between the visual center line and the xoy plane in the spatial coordinate system, and mark it as the pitch angle θ;

[0089] After that, there is a one-way communication connection between the posture recognition end and the posture analysis end;

[0090] The posture analysis end is used to determine the habitual eye use range of the wearer according to the visual center line and the internal visual space. Among them, the internal visual space is transmitted from the scene analysis end to the posture analysis end. The specific method for determining the habitual eye use range includes:

[0091] SS1: Obtain the internal visual space, and collect the point distances in the internal visual space based on the matrix optical distance sensor. Among them, the point distances are collected corresponding to the positions of the matrix elements set in the matrix optical distance sensor;

[0092] SS2: Compare the point distances in the internal visual space with the center distance. If all the point distances in the internal visual space are greater than the center distance, identify the minimum value among the point distances and mark it as the true eye use distance d;

[0093] If there is a point distance less than or equal to the center distance, at this time, extract the pitch angle θ and the spatial coordinate system, and first determine the components θ xz and θ yz of the pitch angle θ in the spatial coordinate system. Among them, θ xz is the component of the pitch angle θ on the xoz plane of the spatial coordinate system, and θ yz is the component of the pitch angle θ on the yoz plane of the spatial coordinate system;

[0094] Identify the xoz plane, and set a threshold angle θ xz-th in the fourth quadrant of the xoz plane. The specific value of the threshold angle θ xz-th is obtained by those skilled in the art through big data operations. Compare the component θ xz with the threshold angle θ xz-th . If θ xz <θ xz-thWhen , the visual center line is rotated downward by an angle α1, with the visual center fixed at the visual center. If θ xz-th ≤θ xz When α2=k2×(-θ xz ), when θ xz ≥0, the visual center line is rotated downward by an angle α3, α3 = k3 × θ xz , where k1, k2 and k3 are all reduced ratios, and the specific values ​​of k1, k2 and k3 are obtained by technicians in this field after big data calculation;

[0095] Reference Figure 2 , based on the center distance, the offset angle α and the pitch angle θ, the actual eye distance d is calculated using the trigonometric function formula, wherein the offset angle α includes α1, α2 and α3, and the specific value of the offset angle α is determined by the actual situation;

[0096] SS3: Compare the actual eye distance d with the standard sight interval Db. If d∈Db, a normal eye signal is generated. When the actual eye distance d is less than the minimum value of the standard sight interval, the generated reminder signal is a close-distance eye signal; if the actual eye distance d is greater than the maximum value of the standard sight interval, the generated reminder signal is a long-distance eye signal;

[0097] The intelligent reminder terminal generates corresponding sound and light reminder information based on the received reminder signal, and uses the terminal device to provide real-time reminders to the wearer.

[0098] Embodiment 3: This embodiment is based on Embodiment 1 and Embodiment 2 and is used to combine and implement Embodiment 1 and Embodiment 2.

[0099] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can make equivalent replacements or changes according to the technical scheme and inventive concept of the present invention within the technical scope disclosed by the present invention, which should be covered by the protection scope of the present invention.

Claims

1. A myopia protection smart glasses based on scene recognition to judge eye health, characterized in that: include: The information collection terminal is used to collect the basic information of the wearer and transmit it to the light analysis terminal and the posture analysis terminal; The light recognition end, based on the two rays emitted by the light detection device, marks the point where each ray falls on the object as an endpoint, and then sets a first edge plane, a second edge plane and a third edge plane based on the individual rays and the endpoints, and then divides the visual space of the smart glasses into an internal visual space and an edge visual space based on the first edge plane, the second edge plane and the third edge plane; The maximum value of the light brightness value in the internal visual space is detected and marked as the central light feature value, and then the light brightness value in the edge visual space is detected. At the same time, based on the central light feature value, the light brightness value in the edge visual space is processed, and then the processed light brightness value is averaged to obtain the edge light feature value. After that, the light recognition end and the light analysis end are connected in a one-way communication manner; The scene recognition end is used to recognize the real-time scene and transmit it to the light analysis end; A light analysis end is used to receive a real-time scene, obtain a standard brightness value and a comfortable contrast value of the real-time scene, and perform a difference calculation between a central light characteristic value and an edge light characteristic value to obtain an absolute difference. Based on the absolute difference, a light consistency signal and a light difference adjustment signal are determined. When a light consistency signal is generated, the absolute difference is compared with the standard brightness value to determine a light compensation value. When a light difference adjustment signal is generated, a light compensation value is determined based on the central light characteristic value, the edge light characteristic value and the comfortable contrast value. Then, the light analysis end transmits the light compensation value to the parameter adjustment end. The parameter adjustment end is used to adjust the light of the smart glasses according to the light compensation value.

2. The myopia protection smart glasses for judging eye health based on scene recognition according to claim 1, characterized in that: Methods for determining internal visual space and peripheral visual space include: A light detection device is arranged on the smart glasses, wherein the light detection device is arranged as a matrix optical distance sensor; The light detection device emits light to the outside, identifies the object to which the light is directed, and marks the point where the light falls on the object as an endpoint. The two endpoints are then connected to obtain a local line segment, wherein the ray emitted by the light detection device is perpendicular to the surface of the smart glasses; Acquire the positions of two rays at the same time, construct two planes based on the individual rays, and mark them as the first edge plane and the second edge plane respectively, and then construct another plane based on the local line segment, and mark it as the third edge plane, wherein the first edge plane and the second edge plane are parallel to each other, and the third edge plane is parallel to the smart glasses, and further, the third edge plane is between the first edge plane and the second edge plane, and the third edge plane is perpendicular to the first edge plane and the second edge plane; Select a first edge plane, identify the position of the ray on the first edge plane, take the ray position as the midpoint of the height threshold X1, and set a height edge point based on the height threshold X1, wherein the height edge point includes an upper edge point and a lower edge point, and then set an upper edge line and a lower edge line at the positions of the upper edge point and the lower edge point, respectively, wherein the upper edge line and the lower edge line are both on the first edge plane and are parallel to the ray on the first edge plane. Similarly, the upper edge line and the lower edge line are also set on the second plane according to the above method; Connect the upper edge lines between the two planes to form an upper edge plane, and connect the lower edge lines between the two planes to form a lower edge plane; The space enclosed by the first edge plane, the second edge plane, the third edge plane, the upper edge plane, the lower edge plane and the smart glasses is marked as the internal visual space, and the space outside the internal vision is marked as the edge visual space.

3. The myopia protection smart glasses for judging eye health based on scene recognition according to claim 2, characterized in that: When the two endpoints are not on the same plane, that is, the lengths of the two rays are inconsistent, the ray with the shorter length will be extended along the original ray path to make the lengths of the two rays consistent. At this time, the original ray with the shorter length will reset its endpoints, and then reconnect the two endpoints to obtain a local line segment, where the local line segment is parallel to the smart glasses.

4. The myopia protection smart glasses for judging eye health based on scene recognition according to claim 2, characterized in that: The method for determining the central light characteristic value and the edge light characteristic value includes: Detecting the light brightness value in the internal visual space, identifying the maximum value of the light brightness in the internal visual space, and marking it as the central light characteristic value of the internal visual space; Acquire the edge visual space, and detect the light brightness value in the edge visual space; In the process of detecting the light brightness value in the edge visual space, the edge visual space is converted into a plane edge area according to the wearer's visual direction. Furthermore, each point in the plane edge area has a corresponding light brightness value. Obtain the central light characteristic value of the internal visual space, compare the central light characteristic value with the light brightness value in the plane edge area, and if there is a light brightness value in the plane edge area that is consistent with the central light characteristic value, set a blank area to replace the area corresponding to the light brightness value in the plane edge area; After the comparison of the light brightness values ​​in the plane edge area is completed, the light brightness values ​​in the plane edge area are detected again, and the average value of the light brightness values ​​in the plane edge area is calculated, and the obtained result is marked as the edge light characteristic value; When calculating the edge light characteristic value, the light brightness value corresponding to the blank area in the plane edge area does not participate in the average value calculation process.

5. The myopia protection smart glasses for judging eye health based on scene recognition according to claim 1, characterized in that: The method for determining the light consistency signal and the light difference adjustment signal includes: According to the real-time scene, a standard brightness value and a comfort contrast value corresponding to the real-time scene are obtained, wherein the comfort contrast value indicates a comfortable ratio between dark and bright light, and the comfort contrast values ​​in different activity scenes are different; Obtain the wearer's light environment, first subtract the edge light characteristic value from the central light characteristic value in the light environment to obtain the light difference, take the absolute value of the light difference, and mark it as the absolute difference; The absolute difference is compared with the difference threshold value X2. If the absolute difference is less than or equal to the difference threshold value X2, a light consistency signal is generated. Otherwise, if the absolute difference is greater than the difference threshold value X2, a light difference adjustment signal is generated.

6. The myopia protection smart glasses for judging eye health based on scene recognition according to claim 5, characterized in that: Methods for determining the compensation light value include: When a light consistency signal is detected, the central light characteristic value is extracted, and the central light characteristic value is subtracted from the standard brightness value, and then the result is divided by the central light characteristic value to obtain the compensation light value; When the light difference adjustment signal is detected, the central light characteristic value and the edge light characteristic value are obtained again, and the central light characteristic value is divided by the edge light characteristic value to obtain the light ratio value, and then the light ratio value is subtracted from the comfort contrast value to obtain the compensation light value.

7. The myopia protection smart glasses for judging eye health based on scene recognition according to claim 1, characterized in that: When the parameter adjustment end adjusts the light of the smart glasses according to the compensation light value, the light compensation value includes positive and negative signs. When the sign in the light compensation value is positive, it means that the light transmittance effect of the smart glasses needs to be enhanced. When the sign in the light compensation value is negative, it means that the light transmittance effect of the smart glasses needs to be reduced.

8. The myopia protection smart glasses for judging eye health based on scene recognition according to claim 1, characterized in that: It also includes a posture recognition terminal, a posture analysis terminal, and an intelligent reminder terminal; The posture recognition end is used to set the visual center line according to the emission point of the light detection device, and obtain the center distance, then set the spatial coordinate system, collect the pitch angle between the visual center line and the spatial coordinate system, and transmit it to the posture analysis end; The posture analysis end is used to collect the point distance in the internal visual space, compare the point distance with the center distance, and analyze the comparison result to determine the real eye distance, then generate a reminder signal according to the real eye distance, and transmit the reminder signal to the intelligent reminder end; The intelligent reminder terminal generates corresponding sound and light reminder information based on the received reminder signal, and uses the terminal device to provide real-time reminders to the wearer.

9. The myopia protection smart glasses for judging eye health based on scene recognition according to claim 8, characterized in that: Methods for determining the center distance and pitch angle include: A multi-axial acceleration sensor is also provided on the smart glasses. The multi-axial acceleration sensor is used to detect the motion state of the wearer, and the multi-axial acceleration sensor is composed of at least a three-axial linear acceleration sensor. The multi-axial acceleration sensor is provided in the middle position of the smart glasses. Set the spatial coordinate system according to the multi-axial acceleration sensor, then identify the emission point of the light detection device, connect the emission points to obtain the starting line segment, take the midpoint position of the starting line segment, and mark the midpoint position as the line of sight center; Based on the rays emitted by the light detection device, a visual center line is also set at the center of the line of sight, wherein the visual center line and the two rays emitted by the light detection device are simultaneously on the same plane and are parallel to each other; The distance length of the visual center line is collected and marked as the center distance. Based on the spatial coordinate system in the multi-axial acceleration sensor, the angle between the visual center line and the xoy plane in the spatial coordinate system is obtained and marked as the pitch angle θ.

10. The myopia protection smart glasses for judging eye health based on scene recognition according to claim 8, characterized in that: Methods for determining the actual eye distance include: SS1: Acquire the internal visual space and collect the point distances in the internal visual space based on the matrix optical distance sensor; SS2: Compare the point distances in the internal visual space with the center distance. If all point distances in the internal visual space are greater than the center distance, identify the minimum value in the point distances and mark it as the true eye distance d. If there is a point distance less than or equal to the center distance, extract the pitch angle θ and the spatial coordinate system, and determine the component θ of the pitch angle θ in the spatial coordinate system xz , where θ xz is the pitch angle θ and the component on the xoz plane of the spatial coordinate system; Identify the xoz plane and set the threshold angle θ in the fourth quadrant of the xoz plane xz-th , the component θ xz With the threshold angle θ xz-th For comparison, if θ xz <θ xz-th When , the visual center line is rotated downward by an angle α1, with the visual center fixed at the visual center. If θ xz-th ≤θ xz When α2=k2×(-θ xz ), when θ xz ≥0, the visual center line is rotated downward by an angle α3, α3 = k3 × θ xz , where k1, k2 and k3 are all reduction ratios; Based on the center distance, the offset angle α and the pitch angle θ, the actual eye distance d is calculated using the trigonometric formula, and the offset angle α refers to α1, α2 and α3; SS3: Compare the actual eye distance d with the standard sight interval Db. If d∈Db, a normal eye signal is generated. When the actual eye distance d is less than the minimum value of the standard visual range, the generated reminder signal is a close-distance eye signal; if the actual eye distance d is greater than the maximum value of the standard visual range, the generated reminder signal is a long-distance eye signal.

Images