Method for relieving juvenile pseudomyopia based on artificial intelligence

Abstract

The invention relates to a method for relieving juvenile pseudomyopia based on artificial intelligence, and relates to the field of vision adjustment, and the method comprises the following steps: obtaining an eye using distance in a sight line direction; when the eye using distance falls into a preset vision interval, acquiring eye using time; when the eye using time exceeds a preset eye using fatigue interval, determining a far-away focal length according to the eye using distance; determining a glasses control parameter according to the distant focal length; and controlling preset zoom glasses to adjust the focal length according to the glasses control parameters. The application has the effects of improving the treatment effect on the pseudomyopia patient and reducing the condition of eye fatigue in daily life.

Description

Technical Field

[0001] The invention relates to the field of vision adjustment, and in particular to an artificial intelligence-based method for alleviating pseudomyopia in adolescents. Background Art

[0002] Pseudomyopia, also known as accommodative myopia or functional myopia, is a non-pathological, temporary myopia state that is mainly caused by abnormal accommodation function of the eye rather than substantial changes in the structure of the eyeball.

[0003] In the prior art, pseudomyopia is usually a temporary blurred vision caused by prolonged close-up use of the eyes, which leads to continuous contraction of the ciliary muscle and convexity of the lens. This condition can be improved through proper rest and treatment, and may even heal itself.

[0004] When patients with pseudomyopia use their eyes too much in daily life, it is easy for the symptoms of pseudomyopia to worsen, and even cause the pseudomyopia to deteriorate into true myopia. Summary of the invention

[0005] In order to improve the treatment effect on patients with pseudomyopia and reduce eye fatigue in daily life, the present invention provides an artificial intelligence-based method for alleviating juvenile pseudomyopia.

[0006] In the first aspect, the present invention provides a method for alleviating pseudomyopia in adolescents based on artificial intelligence, which adopts the following technical solution: A method for alleviating pseudomyopia in adolescents based on artificial intelligence, comprising: Get the eye distance of the sight direction; When the eye use distance falls into the preset visual interval, the eye use time is obtained; When the eye use time exceeds the preset eye fatigue range, the distance away from the focus is determined according to the eye use distance; determining eyewear control parameters based on the away focal distance; The preset varifocal glasses are controlled to adjust the focal length according to the glasses control parameters.

[0007] Optionally, it also includes an eye relaxation method, which includes: When the eye distance falls into the preset relaxation interval, the relaxation time is matched from the preset relaxation time database according to the eye time; Determine the focus change curve according to the relaxation time; Determine relaxation control parameters according to the focal length variation curve; Determine voice prompt parameters according to the focal length change curve; The preset zoom glasses are controlled to adjust the focal length according to the relaxation control parameters, and the prompt device preset on the zoom glasses is controlled to give voice prompts according to the voice prompt parameters.

[0008] Optionally, it also includes a method for generating an eye fatigue interval, and the method for generating an eye fatigue interval includes: Obtaining a visual image of the gaze direction; When the eye distance falls into the preset visual interval, a reference interval is matched from a preset reference interval database according to the eye distance; Determine whether the patient is viewing an electronic screen based on visual images; When the patient is watching an electronic screen, the screen impact factor is determined based on the visual image; Determine if the patient is reading text based on visual images; The font influence factor was determined based on the visual image when the patient was reading text or not; Determine the eye fatigue interval based on the benchmark interval, screen influence factor and font influence factor.

[0009] Optionally, the method for generating an eye fatigue interval further includes: When the eye distance falls into the preset visual interval, the refractive index is determined according to the visual image; Determine the direction of light based on visual images; Determine the refraction influence factor based on the light direction and refractive index; Determine the eye fatigue interval based on the benchmark interval, screen influence factor, font influence factor and refraction influence factor.

[0010] Optionally, an electromagnetic interference prevention method is also included, and the electromagnetic interference prevention method includes: Determine the presence or absence of side-light based on the visual image; When there is side light, the photosensitive number is determined based on the visual image; According to the photosensitive number, a photosensitive sensor preset on one side of the zoom glasses is controlled to obtain the peak light intensity; When the peak value of light intensity exceeds the preset stimulation interval, the valley value of light intensity is determined based on the peak value of light intensity and the visual image; According to the photosensitive number, the photosensitive sensor preset on the other side of the zoom glasses is controlled to obtain the actual light intensity; Determine the light intensity deviation according to the actual light intensity and the light intensity valley value using a preset deviation rate calculation method; When the light intensity deviation does not exceed the preset disturbance interval, the refraction area is determined based on the side-ray; The preset zoom glasses are controlled to adjust the refractive index of the light according to the refractive area.

[0011] By adopting the above technical solution, when strong light is detected on the side of the glasses, the light intensity on the other side of the eye is predicted based on the diffusion characteristics of the light, and compared with the actual detected light intensity to determine whether the detected data is correct, thereby reducing the situation where electromagnetic interference causes the photosensitive sensor to output erroneous data.

[0012] Optionally, the electromagnetic interference prevention method further includes: When the eye distance falls into the preset visual interval, determining whether the eye distance can be directly identified based on the visual image; When the eye distance cannot be directly identified, the reference light group preset on the zoom glasses is controlled to turn on and off, and a reference image illuminated by the reference light group is obtained; Determine a reference distance based on a reference image; When the eye distance can be directly identified, the reference distance is determined based on the visual image; Based on the identification of the reference distance, the difference between the reference distance and the eye distance is calculated and defined as the distance deviation; When the distance deviation does not exceed the preset deviation range, the eye use time is obtained.

[0013] By adopting the above technical solution, it is determined through images whether the eye distance is consistent with the actual situation. When it is impossible to directly determine through images, light is emitted by a reference light group, and the eye distance is determined based on the diffusion of light, thereby reducing the situation where the distance sensor outputs erroneous data due to electromagnetic interference.

[0014] Optionally, the invention further includes a glasses adjustment method, wherein the glasses adjustment method includes: Acquire a wearing image of the glasses; Determine the three-dimensionality of the face based on the worn image; Determine the wearing height based on the three dimensions of the face; Determine the parameters of the bridge of the nose based on the three-dimensional face; Determine nose pad control parameters according to wearing height and nose bridge parameters; The preset zoom glasses are controlled to adjust the nose pads according to the nose pad control parameters.

[0015] By adopting the above technical solution, the appropriate wearing position of the glasses is determined according to the patient's eye position, so that the angle of the nose pads required for the glasses to reach the wearing position is matched according to the patient's nose bridge shape, thereby improving the suitability of the glasses for the patient.

[0016] Optionally, the glasses adjustment method further includes: Determine a set of wearing feature points according to wearing height and three-dimensional face; Determine the facial curve according to the wearing feature point set; Determine the lens angle according to the facial curve; Determine lens control parameters according to lens angle; The preset varifocal glasses are controlled to adjust the lens angle according to the lens control parameters.

[0017] By adopting the above technical solution, the feature points of the glasses wearing position are extracted from the facial data used, so as to fit the concave-convex change curve of the patient's eyes, and then adjust the angle of the lens according to the concave-convex change curve, so as to further improve the suitability of the glasses to the patient.

[0018] Optionally, the glasses adjustment method further includes: Obtaining pressure information of the temples of varifocal glasses; Determine whether the temple angle needs to be adjusted based on the pressure information; When the angle of the temple needs to be adjusted, the wearing position is determined according to the pressure information, and the current angle of the temple is obtained; Determine the adjustment angle according to the wearing position, current angle and pressure information; Determine temple control parameters according to adjustment angle; The temple adjustment angles preset on the varifocal glasses are controlled according to the temple control parameters.

[0019] By adopting the above technical solution, the pressure value on the side of the temple of the glasses is detected to determine whether the angle of the temple of the glasses needs to be adjusted. When the pressure value is too large, it means that the expansion angle of the temple is too small and the glasses are worn too tightly. When the pressure value is too small, it means that the expansion angle of the temple is too large and the glasses are worn too loose. The angle of the temple is adjusted according to the pressure value and the wearing state of the glasses, thereby improving the suitability of the glasses and the patient.

[0020] In summary, the present application includes at least one of the following beneficial technical effects: When the distance to the object is close in the direction of sight, it is judged that the patient is using his eyes. At this time, the time of using the eyes is accumulated, and the focus of the glasses is adjusted when the time of using the eyes is too long, so that the patient cannot see the object clearly, and then the patient is prompted to stop using the eyes and relax the eyes, reduce eye fatigue, and improve the treatment effect of pseudomyopia patients; When the patient stops using their eyes after a long period of time, the focal length of the zoom glasses is controlled to change regularly, thereby assisting the patient to control the ciliary muscles and relax the eyes, thus improving the treatment effect for pseudomyopia patients; The basic fatigue interval is matched according to the patient's visual distance, and then the basic fatigue interval is corrected according to the refresh frequency of the visual target and the font size of the visual target, so as to lower the fatigue interval of the visual target with a higher refresh frequency and smaller font. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a flowchart of a method for alleviating pseudomyopia in adolescents based on artificial intelligence; Figure 2 It is a flowchart of eye relaxation methods; Figure 3 This is the process of using the eye fatigue interval generation method Figure 1 ; Figure 4 This is the process of using the eye fatigue interval generation method Figure 2 ; Figure 5 Is the process of anti-electromagnetic interference method Figure 1 ; Figure 6 Is the process of anti-electromagnetic interference method Figure 2 ; Figure 7 The process of glasses adjustment method Figure 1 ; Figure 8 The process of glasses adjustment method Figure 2 ; Fig. 9 The process of glasses adjustment method Figure 3 ; Fig.10 The process of glasses adjustment method Figure 4 . DETAILED DESCRIPTION

[0022] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0023] The embodiment of the present application discloses an artificial intelligence-based method for alleviating pseudomyopia in adolescents, which reduces the patient's eye fatigue by controlling the focal length of zoom glasses, and assists the patient in relaxing the eyes by changing the focal length, thereby improving the treatment effect of pseudomyopia.

[0024] Reference Figure 1 , an artificial intelligence-based method for alleviating pseudomyopia in adolescents, including: Step 100: Obtain the eye distance in the sight direction.

[0025] Varifocal glasses refer to glasses that adjust the focal length by changing the voltage. Generally, varifocal glasses are used to alleviate pseudomyopia. The varifocal glasses are selected by the staff according to the actual situation, and will not be elaborated here. The eye distance refers to the distance between the object closest to the varifocal glasses in the line of sight of the varifocal glasses and the varifocal glasses. The eye distance can be obtained through the distance sensor preset on the varifocal glasses. The method of obtaining the eye distance is selected by the staff according to the actual situation, and will not be elaborated here.

[0026] Step 101: When the eye use distance falls into a preset visual interval, the eye use time is obtained.

[0027] The visual interval refers to the eye-use distance interval that is easy to cause eye fatigue. The visual interval is selected by the staff according to the actual situation and will not be elaborated here. When the eye-use distance falls into the visual interval, it means that the eyes are being used at this time. The eye-use time refers to the accumulated time length from the time when the eye-use distance falls into the visual interval. The eye-use time can be obtained through the timer. The method of obtaining the eye-use time is selected by the staff according to the actual situation and will not be elaborated here.

[0028] Step 102: When the eye use time exceeds a preset eye fatigue interval, the away focus distance is determined according to the eye use distance.

[0029] The eye fatigue interval refers to the eye use time interval that is likely to cause the worsening of pseudomyopia. The eye fatigue interval is selected by the staff according to the actual situation and will not be elaborated here. Eye use time exceeding the eye fatigue interval means that the patient is already eye fatigued at this time. The away focus can be obtained through the away focus matching model. The away focus matching model refers to a neural network model obtained through sample training in advance. The away focus matching model can first calculate the difference between the upper limit value of the visual interval and the eye distance, and define it as the correction distance. After correcting the original focal length of the zoom glasses according to the correction distance, the away focus is the focal length value obtained after correcting the original focal length of the zoom glasses according to the correction distance.

[0030] Step 103: Matching model glasses control parameters according to the away focus distance.

[0031] Glasses control parameters refer to parameters used to control the refractive index of varifocal glasses to adjust the focal length, and the glasses control parameters can be obtained through glasses control parameter matching models. Glasses control parameter matching models refer to neural network models obtained through sample training in advance, and glasses control parameter matching models can match glasses control parameters that control the focal length of varifocal glasses to reach the farthest focal length.

[0032] Step 104: Control the preset varifocal glasses to adjust the focal length according to the glasses control parameters.

[0033] When the patient suffers from eye fatigue, the focal length can be adjusted so that the patient needs to adjust the eye distance to beyond the visual range to see the visual target clearly, thereby reducing the patient's eye fatigue.

[0034] Reference Figure 2 , eye relaxation methods include: Step 200: When the eye use distance falls into the preset relaxation interval, the relaxation time is matched from the preset relaxation time database according to the eye use time.

[0035] The relaxation interval refers to the eye distance interval that can relax the eyes. The relaxation interval is selected by the staff according to the actual situation and will not be described here. The eye distance falling into the relaxation interval means that the patient is relaxing the eyes at this time. The relaxation time database refers to a database that records the eye use time and its corresponding minimum eye relaxation time in advance. The relaxation time is the minimum eye relaxation time corresponding to the eye use time matched from the relaxation time database.

[0036] Step 201: Determine a focal length variation curve according to the relaxation time.

[0037] The focal length change curve can be obtained through the focal length change curve matching model. The focal length change curve matching model refers to a neural network model obtained through sample training in advance. The focal length change curve matching model can match the focal length change of the patient's visual focus from near to far, and from far to near after staying at a distance for a period of time, and repeat the process during the relaxation time to obtain the focal length change curve.

[0038] Step 202: Determine relaxation control parameters according to the focal length variation curve.

[0039] The relaxation control parameters can be obtained through the relaxation control parameter matching model. The relaxation control parameter matching model refers to a neural network model obtained through sample training in advance. The relaxation control parameter matching model can match the relaxation control parameters that control the zoom glasses to adjust the focal length according to the focal length change curve.

[0040] Step 203: Determine voice prompt parameters according to the focal length variation curve.

[0041] Voice prompt parameters refer to parameters used to control the varifocal glasses to issue voice prompts. The voice prompt parameters can be obtained through a voice prompt parameter matching model. The voice prompt parameter matching model refers to a neural network model obtained in advance through sample training. The voice prompt parameter matching model can match voice prompt parameters that control the varifocal glasses to adjust the focal length according to the focal length change curve to assist the patient in relaxing the eyes through voice. The content of the voice assistance may include music for relaxing the mind, and may also include prompt sounds that prompt the patient to change the focus of his or her vision.

[0042] Step 204: controlling the preset varifocal glasses to adjust the focal length according to the relaxation control parameter, and controlling the prompting device preset on the varifocal glasses to give voice prompts according to the voice prompting parameter.

[0043] When patients relax their eyes after using their eyes for a long time, the focus can be adjusted through varifocal glasses so that patients can continuously adjust the focus of their eyes without the help of external objects, thereby relaxing the ciliary muscles and relieving eye fatigue.

[0044] Reference Figure 3 , the eye fatigue interval generation method includes: Step 300: Acquire a visual image in the sight direction.

[0045] The visual image refers to the picture in the line of sight of the zoom glasses. The visual image can be obtained through a camera preset on the zoom glasses. The method of obtaining the visual image is selected by the staff according to the actual situation and will not be elaborated here.

[0046] Step 301: When the eye distance falls into a preset visual interval, a reference interval is matched from a preset reference interval database according to the eye distance.

[0047] The benchmark interval database refers to a database that records the useful eye distance and its corresponding minimum eye fatigue duration in advance. The benchmark interval is a duration range matched from the benchmark interval database with the minimum eye fatigue duration corresponding to the eye distance as the lower limit of the interval and infinity as the upper limit of the interval.

[0048] Step 302: Determine whether the patient is watching an electronic screen based on the visual image.

[0049] The electronic screen recognition model can be used to determine whether the patient is watching an electronic screen. The electronic screen recognition model refers to a neural network model obtained through sample training in advance. The electronic screen recognition model can determine whether the object closest to the patient in the direction of sight in the visual image is an electronic screen.

[0050] Step 303: When the patient is watching an electronic screen, determine a screen influence factor according to the visual image.

[0051] The screen influence factor refers to a parameter used to correct the fatigue interval. The screen influence factor can be obtained through a screen influence factor matching model. The screen influence factor matching model refers to a neural network model obtained in advance through sample training. The screen influence factor matching model can match the screen influence factor of the corrected baseline interval according to the refresh rate of the electronic screen that the patient is watching. The refresh rate of the electronic screen can be obtained by acquiring multiple visual images in a short period of time and estimating the refresh rate of the electronic screen based on the changes in the picture on the electronic screen in the image.

[0052] Step 304: Determine whether the patient is reading text based on the visual image.

[0053] The text recognition model can be used to determine whether the patient is reading text. The text recognition model refers to a neural network model obtained through sample training in advance. The text recognition model can determine whether there is text on the object closest to the patient in the direction of sight in the visual image.

[0054] Step 305: When the patient is reading text, determine the font influence factor according to the visual image.

[0055] The font influence factor refers to a parameter used to correct the fatigue interval. The font influence factor can be obtained through a font influence factor matching model. The font influence factor matching model refers to a neural network model obtained in advance through sample training. The font influence factor matching model can match the font influence factor of the corrected benchmark interval according to the font size, font color and background color of the text that the patient is reading.

[0056] Step 306: Determine the eye fatigue interval according to the reference interval, the screen influence factor and the font influence factor.

[0057] The eye fatigue interval can be obtained by calculating the eye fatigue interval matching model. The eye fatigue interval matching model refers to a neural network model obtained through sample training in advance. The eye fatigue interval matching model can calculate the product of the screen influence factor, the font influence factor and the benchmark interval, and define it as the eye fatigue interval.

[0058] Reference Figure 4 , the eye fatigue interval generation method also includes: Step 307: When the eye distance falls within a preset visual interval, the refractive index is determined according to the visual image.

[0059] The refractive index can be obtained through refractive index matching model identification. The refractive index matching model refers to a neural network model obtained through sample training in advance. The refractive index matching model can determine the material of the object closest to the patient in the direction of sight in the visual image, and then match the corresponding refractive index according to the material.

[0060] Step 308: Determine the light direction according to the visual image.

[0061] The light direction can be obtained through the light direction matching model. The light direction matching model refers to a neural network model obtained through sample training in advance. The light direction matching model can identify the main direction of the light with the highest light intensity in the visual image.

[0062] Step 309: Determine the refraction influence factor according to the light direction and the refractive index.

[0063] The refraction influence factor refers to a parameter used to correct the fatigue interval. The refraction influence factor can be obtained through a refraction influence factor matching model. The refraction influence factor matching model refers to a neural network model obtained through sample training in advance. The refraction influence factor matching model can match the refraction influence factor of the reference interval corrected according to the light direction and the refractive index of the visual target.

[0064] Step 310: Determine the eye fatigue interval according to the reference interval, the screen influence factor, the font influence factor and the refraction influence factor.

[0065] The eye fatigue interval is the product of the reference interval calculated by the eye fatigue interval matching model, the screen influence factor, the font influence factor and the refraction influence factor.

[0066] Among them, the baseline interval, screen influence factor, font influence factor and refraction influence factor can be analyzed to generate an eye use report, and the eye use report can be sent to the terminal, so that personnel can download the eye use report from the terminal, and then it is convenient for doctors to formulate further treatment plans according to the eye use conditions reflected in the eye use report, and it is convenient for individuals to check their eye use conditions.

[0067] Reference Figure 5 , anti-electromagnetic interference methods include: Step 400: Determine whether there is side light according to the visual image.

[0068] The presence of side light can be determined through a side light recognition model. The side light recognition model refers to a neural network model obtained through sample training in advance. The side light recognition model can determine whether there is light coming from the side within the visual range of the glasses, wherein the side refers to an area within the visual range that exceeds a pre-set main visual area.

[0069] Step 401: When there is side light, determine the photosensitivity number based on the visual image.

[0070] The photosensitive sensor refers to a sensor installed on the zoom glasses to obtain the optical fiber strength. Each zoom pair of glasses is provided with photosensitive sensors on both sides. The photosensitive sensors are selected by the staff according to the actual situation and will not be elaborated here.

[0071] The photosensitive number refers to a number set in advance to distinguish the photosensitive sensors. The photosensitive number corresponds to the photosensitive sensor one-to-one. The setting method of the photosensitive number is selected by the staff according to the actual situation and will not be described here. The photosensitive number can be obtained through the photosensitive sensor number matching model. The photosensitive sensor number matching model refers to a neural network model obtained through sample training in advance. The photosensitive sensor number matching model can match the area in the visual image where the side light is directed to the zoom glasses lens, and match the photosensitive number of the photosensitive sensor close to the area.

[0072] Step 402: Control a photosensitive sensor preset on one side of the varifocal glasses to obtain a light intensity peak value according to the photosensitive number.

[0073] The peak light intensity refers to the light intensity signal obtained by the photosensor near the area where the side light is directed to the lens of the zoom glasses. The method for obtaining the peak light intensity is selected by the staff according to the actual situation and will not be elaborated here.

[0074] Step 403: When the light intensity peak value exceeds the preset stimulation interval, the light intensity valley value is determined according to the light intensity peak value and the visual image.

[0075] The stimulation range refers to the range of light intensity in which the light can easily irritate the eyes and cause worsening of pseudomyopia. The stimulation range is selected by the staff based on actual conditions and will not be elaborated here.

[0076] The light intensity valley value can be obtained through the light intensity valley value matching model. The light intensity valley value matching model refers to a neural network model obtained in advance through sample training. The light intensity valley value matching model can predict the light intensity signal obtained by the photosensor in the area away from the side light rays directed toward the lens of the varifocal glasses based on the light diffusion of the side light rays in the visual image, and define the signal as the light intensity valley value.

[0077] Step 404: Control the photosensitive sensor preset at the other side of the zoom glasses to obtain the actual light intensity according to the photosensitive number.

[0078] The actual light intensity refers to the light intensity signal obtained by the photosensor in the area away from the side light rays incident on the lens of the zoom glasses. The method for obtaining the actual light intensity is selected by the staff according to the actual situation and will not be elaborated here.

[0079] Step 405: Determine the light intensity deviation according to the actual light intensity and the light intensity valley value using a preset deviation rate calculation method.

[0080] The deviation rate calculation method refers to a method used to calculate the degree of deviation between the actual light intensity and the light intensity valley value. Generally, the difference between the light intensity valley value and the actual light intensity is calculated first, and the difference is divided by the actual light intensity to obtain the deviation value, and the deviation value is defined as the light intensity deviation.

[0081] Step 406: When the light intensity deviation does not exceed the preset disturbance interval, determine the refraction area based on the side-illuminated light.

[0082] The disturbance interval refers to the range of deviation values ​​allowed between the actual light intensity and the light intensity valley value. The disturbance interval is selected by the staff according to the actual situation and will not be elaborated here. The light intensity deviation does not exceed the disturbance interval, which means that the detection result of the photosensor is reliable. The refraction area can be obtained by identifying the refraction area matching model. The refraction area matching model refers to a neural network model obtained through sample training in advance. The refraction area matching model can match the area where the side light is incident on the zoom glasses lens and define the area as the refraction area.

[0083] Step 407: Control the preset variable focus glasses to adjust the refractive index of the light according to the refractive area.

[0084] When there is strong side light, the refractive index of the refractive area of ​​the zoom glasses is increased to reduce the stimulation of the side light to the eyes, thereby reducing eye fatigue.

[0085] When the light intensity deviation exceeds the preset disturbance range, it means that the detection result of the photosensor is unreliable. At this time, the refractive index of the zoom glasses is not adjusted, thereby reducing the situation where the patient interferes with the detection result of the photosensor when using portable devices such as mobile phones.

[0086] Reference Figure 6 , the anti-electromagnetic interference method also includes: Step 408: When the eye distance falls within the preset visual interval, determine whether the eye distance can be directly identified based on the visual image.

[0087] Whether the eye distance can be directly identified can be determined through an eye distance determination model. The eye distance determination model refers to a neural network model obtained through sample training in advance. The eye distance determination model can identify reference objects in visual images, and determine that the eye distance can be directly identified when the reference object is identified, and determine that the eye distance cannot be directly identified when the reference object is not identified, wherein the reference object refers to a pre-set object for determining the eye distance.

[0088] Step 409: When the eye distance cannot be directly identified, the reference light set preset on the zoom glasses is controlled to be turned on and off, and a reference image illuminated by the reference light set is obtained.

[0089] The reference light group refers to a light group consisting of at least two lamps set on the zoom glasses. The illumination directions of the lamps in the reference light group are different. The setting method of the reference light group is selected by the staff according to actual conditions and will not be elaborated here.

[0090] The reference image refers to a picture obtained when the reference light group is started. The reference image can be obtained through the camera on the zoom glasses. The method for obtaining the reference image is selected by the staff according to the actual situation and will not be elaborated here.

[0091] Step 410: Determine a reference distance according to a reference image.

[0092] The reference distance can be obtained through a reference distance recognition model. The reference distance recognition model refers to a neural network model obtained in advance through sample training. The reference distance recognition model can judge the reference distance between the object closest to the zoom glasses in the line of sight of the zoom glasses and the zoom glasses from the diffusion of light emitted by the reference light group in the reference image.

[0093] Step 411: When the eye distance can be directly identified, the reference distance is determined based on the visual image.

[0094] The reference distance can be obtained through an eye distance recognition model. The eye distance recognition model refers to a neural network model obtained through sample training in advance. The eye distance recognition model can recognize the reference distance according to the size of the reference object in the visual image.

[0095] Step 412: Based on the identified reference distance, the difference between the reference distance and the eye distance is calculated and defined as the distance deviation.

[0096] Distance deviation refers to the difference between the reference distance and the eye distance. The distance deviation is used to determine whether the eye distance is reliable.

[0097] Step 413: When the distance deviation does not exceed the preset deviation interval, the eye use time is obtained.

[0098] The deviation range refers to the error range between the reference distance and the eye distance that can exist. The deviation range is selected by the staff according to the actual situation and will not be elaborated here. If the distance deviation does not exceed the deviation range, it means that the eye distance is reliable at this time and subsequent operations can be carried out.

[0099] When the distance deviation exceeds the preset deviation range, it means that the eye distance is unreliable and the sensor for obtaining the eye distance may be subject to electromagnetic interference. At this time, no subsequent operation is performed to reduce the situation where the varifocal glasses cause abnormal operation due to electromagnetic interference.

[0100] Reference Figure 7 , glasses adjustment methods include: Step 500: Acquire a wearing image of the glasses.

[0101] The wearing image refers to the patient's facial picture, which can be obtained through the camera on the zoom glasses. The method of obtaining the wearing image is selected by the staff according to the actual situation and will not be elaborated here.

[0102] Step 501: Determine the three-dimensional face according to the wearing image.

[0103] The facial three-dimensional can be established through a facial three-dimensional matching model. The facial three-dimensional matching model refers to a neural network model obtained through sample training in advance. The facial three-dimensional matching model can establish a three-dimensional model of the patient's face from the wearing image, wherein the facial three-dimensional is the three-dimensional model of the patient's face.

[0104] Step 501: Determine the wearing height based on the three dimensions of the face.

[0105] The wearing height can be obtained through a glasses wearing height matching model. The glasses wearing height matching model refers to a neural network model obtained through sample training in advance. The glasses wearing height matching model can recognize the eye features in three dimensions of the face and extract the distance from the eye features to the midpoint of the chin as the wearing height.

[0106] Step 503: Determine nose bridge parameters based on the three-dimensional face.

[0107] The nose bridge parameters can be extracted through the nose bridge parameter matching model. The nose bridge parameter matching model refers to a neural network model obtained through sample training in advance. The nose bridge parameter matching model can extract the three-dimensional parameters of the part of the nose bridge that is in contact with the glasses after wearing the glasses from the three-dimensional face. The nose bridge parameters are the three-dimensional parameter results obtained by matching.

[0108] Step 504: Determine nose pad control parameters according to the wearing height and nose bridge parameters.

[0109] Nose pad control parameters refer to parameters used to control the adjustment of nose pad angles of varifocal glasses. The nose pad control parameters can be obtained by converting a nose pad angle control model. The nose pad angle control model refers to a neural network model obtained in advance through sample training. The nose pad angle control model can match the nose pad control parameters that control the nose pad to adapt to the direction of the nose bridge so that the glasses are worn at the same height as the eyes.

[0110] Step 505: Control the preset zoom glasses to adjust the nose pads according to the nose pad control parameters.

[0111] The nose pads of the zoom glasses are adjusted according to the shape of the patient's nose bridge, thereby improving the wearing comfort of the zoom glasses and allowing the zoom glasses to be stably worn at a wearing height.

[0112] Reference Figure 8 , the glasses adjustment method also includes: Step 506: Determine a wearing feature point set based on the wearing height and the three-dimensional face.

[0113] The wearing feature point set can be obtained through the wearing feature point extraction model. The wearing feature point extraction model refers to a neural network model obtained through sample training in advance. The wearing feature point extraction model can extract the coordinates of data points located at the wearing height in the three-dimensional face and define them as the wearing feature point set.

[0114] Step 507: Determine the facial curve according to the wearing feature point set.

[0115] The facial curve fitting model refers to a neural network model obtained through sample training in advance. The facial curve fitting model can fit the facial curve according to the wearing feature point set.

[0116] Step 508: Determine the lens angle according to the facial curve.

[0117] The lens angle can be obtained by matching the lens angle fitting model. The lens angle fitting model refers to a neural network model obtained in advance through sample training. The lens angle fitting model can match the lens angle according to the direction of the eye in the facial curve.

[0118] Step 509: Determine lens control parameters according to the lens angle.

[0119] The lens control parameters refer to the glasses angle control model, which refers to the neural network model obtained through sample training in advance. The glasses angle control model can match the glasses control parameters that control the lenses of the varifocal glasses to adjust according to the lens angle.

[0120] Step 510: Control the preset varifocal glasses to adjust the lens angle according to the lens control parameters.

[0121] Adjust the angle of the lens according to the direction of the eye so that the lens is parallel to the eye, thereby improving the comfort of wearing the zoom glasses and reducing the situation where the vision correction effect of the glasses is reduced due to the lens being not parallel to the eye.

[0122] Reference Fig. 9 , the glasses adjustment method also includes: Step 511: Determine the adjustment angle according to the lens control parameters.

[0123] The adjustment angle can be extracted from the lens control parameters through a parameter analysis model. The parameter analysis model refers to a neural network model obtained through sample training in advance. The parameter analysis model can match the adjustment angle that the lens needs to be adjusted from the lens control parameters.

[0124] Step 512: Determine a positive adjustment area and a reverse adjustment area according to the adjustment angle.

[0125] The positive accommodation area and the reverse accommodation area can be segmented by an accommodation area segmentation model. The accommodation area segmentation model refers to a neural network model obtained in advance through sample training. The accommodation area segmentation model can identify that in the process of adjusting the lens angle according to the adjustment angle, the part on which force needs to be applied to the lens is defined as the positive accommodation area, and the remaining part is defined as the reverse accommodation area.

[0126] Step 513: Match the airbag number and airbag position from a preset airbag number database according to the positive adjustment area.

[0127] An airbag refers to a device that is pre-set on the varifocal glasses and is used to fix the lens to the frame of the varifocal glasses. There are multiple airbags arranged around the lens, and the airbags are symmetrically arranged with respect to the lens. The airbags are selected by the staff according to actual conditions and will not be elaborated here.

[0128] The airbag number refers to a number set in advance to distinguish airbags. Each airbag corresponds to an airbag number one by one. The setting method of the airbag number is selected by the staff according to the actual situation and will not be described here. The airbag number database refers to a database that records the airbag numbers and their corresponding airbag positions in advance. The airbag number and airbag position are the airbag numbers and airbag positions of the airbags matched from the airbag numbers and falling into the positive adjustment area.

[0129] Step 514: Determine the adjustment distance according to the airbag position and the adjustment angle.

[0130] The adjustment distance can be obtained through an adjustment distance matching model. The adjustment distance matching model refers to a neural network model obtained in advance through sample training. The adjustment distance matching model can identify the adjustment distance adjusted by the airbag position of the lens when the lens is adjusted according to the adjustment angle.

[0131] Step 515: Determine the forward adjustment gas volume according to the adjustment distance.

[0132] The positive adjustment gas volume can be calculated through the positive adjustment gas volume matching model. The positive adjustment gas volume matching model refers to a neural network model obtained through sample training in advance. The positive adjustment gas volume matching model can match the positive adjustment gas volume required to fill the air bag to move the air bag position of the lens according to the adjustment distance.

[0133] Step 516: Determine the reverse airbag number according to the airbag position.

[0134] The reverse airbag number can be obtained by matching the reverse airbag number matching model. The reverse airbag number matching model refers to a neural network model obtained in advance through sample training. The reverse airbag number matching model can match the reverse airbag number of the airbag with the same airbag position and different airbag number.

[0135] Step 517: Control the preset airbag to inflate according to the airbag number according to the forward regulated air volume, and control the preset airbag to exhaust according to the forward regulated air volume according to the reverse airbag number.

[0136] The angle of the lens is adjusted by inflating and deflating the airbag, while reducing the misalignment between the lens and the eye caused by the center deviation of the lens during angle adjustment, thereby improving the convenience of using zoom glasses.

[0137] Reference Fig.10 , the glasses adjustment method also includes: Step 518: Obtain pressure information of the temples of the variable focus glasses.

[0138] Pressure information refers to the pressure detection information on the horizontal direction of the temples of the zoom glasses. The pressure information can be obtained through the pressure sensors on the temples. The method for obtaining the pressure information is selected by the staff according to the actual situation and will not be elaborated here.

[0139] Step 519: Determine whether the temple angle needs to be adjusted based on the pressure information.

[0140] The temple adjustment recognition model can be used to determine whether the temple angle needs to be adjusted. The temple adjustment recognition model refers to a neural network model obtained through sample training in advance. The temple adjustment recognition model can convert pressure information into a pressure value, and compare the pressure value with a preset adjustment range. When the pressure value exceeds the preset adjustment range, it is determined that the temple angle needs to be adjusted. The adjustment range is selected by the staff based on actual conditions and will not be elaborated here.

[0141] Step 520: When the temple angle needs to be adjusted, the wearing position is determined according to the pressure information, and the current angle of the temple is obtained.

[0142] The wearing position can be obtained by analyzing the pressure information through a wearing position recognition model. The wearing position recognition model refers to a neural network model obtained in advance through sample training. The wearing position recognition model can identify the position of the pressure sensor that obtains the pressure information from the pressure information and use the position as the wearing position.

[0143] The current angle refers to the current angle of the temple. The current angle can be obtained by an angle sensor arranged between the temple and the frame. The method for obtaining the current angle is selected by the staff according to the actual situation and will not be elaborated here.

[0144] Step 521: Determine the adjustment angle according to the wearing position, current angle and pressure information.

[0145] The adjustment angle can be calculated through an adjustment angle matching model. The adjustment angle matching model refers to a neural network model obtained in advance through sample training. The adjustment angle matching model can match the adjustment angle by rotating the temples so that the pressure value of the wearing position falls into the adjustment range.

[0146] Step 522: Determine temple control parameters according to the adjustment angle.

[0147] The temple control parameters refer to the parameters used to control the adjustment of the temple angles of the varifocal glasses. The temple control parameters can be obtained by matching the temple control model. The temple control model refers to a neural network model obtained in advance through sample training. The temple control model can match the temple control parameters that control the temples to rotate according to the adjustment angle.

[0148] Step 523: Control the temple adjustment angle preset on the varifocal glasses according to the temple control parameters.

[0149] When the temples are too tight or too loose, the temples are controlled to rotate, thereby improving the wearing comfort of the varifocal glasses.

[0150] The training methods of the neural network models described above are common knowledge among those skilled in the art and will not be elaborated here.

Claims

1. A method for alleviating pseudomyopia in adolescents based on artificial intelligence, characterized in that: include: Get the eye distance of the sight direction; When the eye use distance falls into the preset visual interval, the eye use time is obtained; When the eye use time exceeds the preset eye fatigue range, the distance away from the focus is determined according to the eye use distance; determining eyewear control parameters based on the away focal distance; The preset varifocal glasses are controlled to adjust the focal length according to the glasses control parameters.

2. The method for alleviating pseudomyopia in adolescents based on artificial intelligence according to claim 1, characterized in that: Also included is an eye relaxation method, the eye relaxation method comprising: When the eye distance falls into the preset relaxation interval, the relaxation time is matched from the preset relaxation time database according to the eye time; Determine the focus change curve according to the relaxation time; Determine relaxation control parameters according to the focal length variation curve; Determine voice prompt parameters according to the focal length change curve; The preset zoom glasses are controlled to adjust the focal length according to the relaxation control parameters, and the prompt device preset on the zoom glasses is controlled to give voice prompts according to the voice prompt parameters.

3. The method for alleviating pseudomyopia in adolescents based on artificial intelligence according to claim 2, characterized in that: The invention also includes a method for generating an eye fatigue interval, wherein the method comprises: Obtaining a visual image of the gaze direction; When the eye distance falls into the preset visual interval, a reference interval is matched from a preset reference interval database according to the eye distance; Determine whether the patient is viewing an electronic screen based on visual images; When the patient is watching an electronic screen, the screen impact factor is determined based on the visual image; Determine if the patient is reading text based on visual images; The font influence factor was determined based on the visual image when the patient was reading text or not; Determine the eye fatigue interval based on the benchmark interval, screen influence factor and font influence factor.

4. The method for alleviating pseudomyopia in adolescents based on artificial intelligence according to claim 3, characterized in that: The method for generating an eye fatigue interval further includes: When the eye distance falls into the preset visual interval, the refractive index is determined according to the visual image; Determine the direction of light based on visual images; Determine the refraction influence factor based on the light direction and refractive index; Determine the eye fatigue interval based on the benchmark interval, screen influence factor, font influence factor and refraction influence factor.

5. The method for alleviating pseudomyopia in adolescents based on artificial intelligence according to claim 3, characterized in that: Also included is an electromagnetic interference prevention method, the electromagnetic interference prevention method comprising: Determine the presence or absence of side-light based on the visual image; When there is side light, the photosensitive number is determined based on the visual image; According to the photosensitive number, a photosensitive sensor preset on one side of the zoom glasses is controlled to obtain the peak light intensity; When the peak value of light intensity exceeds the preset stimulation interval, the valley value of light intensity is determined based on the peak value of light intensity and the visual image; According to the photosensitive number, the photosensitive sensor preset on the other side of the zoom glasses is controlled to obtain the actual light intensity; Determine the light intensity deviation according to the actual light intensity and the light intensity valley value using a preset deviation rate calculation method; When the light intensity deviation does not exceed the preset disturbance interval, the refraction area is determined based on the side-ray; The preset zoom glasses are controlled to adjust the refractive index of the light according to the refractive area.

6. The method for alleviating pseudomyopia in adolescents based on artificial intelligence according to claim 5, characterized in that: The anti-electromagnetic interference method also includes: When the eye distance falls into the preset visual interval, determining whether the eye distance can be directly identified based on the visual image; When the eye distance cannot be directly identified, the reference light group preset on the zoom glasses is controlled to turn on and off, and a reference image illuminated by the reference light group is obtained; Determine a reference distance based on a reference image; When the eye distance can be directly identified, the reference distance is determined based on the visual image; Based on the identification of the reference distance, the difference between the reference distance and the eye distance is calculated and defined as the distance deviation; When the distance deviation does not exceed the preset deviation range, the eye use time is obtained.

7. The method for alleviating pseudomyopia in adolescents based on artificial intelligence according to claim 1, characterized in that: Also included is a glasses adjustment method, the glasses adjustment method comprising: Acquire a wearing image of the glasses; Determine the three-dimensionality of the face based on the worn image; Determine the wearing height based on the three dimensions of the face; Determine the parameters of the bridge of the nose based on the three-dimensional face; Determine nose pad control parameters according to wearing height and nose bridge parameters; The preset zoom glasses are controlled to adjust the nose pads according to the nose pad control parameters.

8. The method for alleviating pseudomyopia in adolescents based on artificial intelligence according to claim 7, characterized in that: The glasses adjustment method further comprises: Determine a set of wearing feature points according to wearing height and three-dimensional face; Determine the facial curve according to the wearing feature point set; Determine the lens angle according to the facial curve; Determine lens control parameters according to lens angle; The preset varifocal glasses are controlled to adjust the lens angle according to the lens control parameters.

9. The method for alleviating pseudomyopia in adolescents based on artificial intelligence according to claim 8, characterized in that: The glasses adjustment method further comprises: Obtaining pressure information of the temples of varifocal glasses; Determine whether the temple angle needs to be adjusted based on the pressure information; When the angle of the temple needs to be adjusted, the wearing position is determined according to the pressure information, and the current angle of the temple is obtained; Determine the adjustment angle according to the wearing position, current angle and pressure information; Determine temple control parameters according to adjustment angle; The temple adjustment angles preset on the varifocal glasses are controlled according to the temple control parameters.

Images