VR imaging control method based on image recognition

Abstract

The invention relates to the technical field of VR imaging, and particularly discloses a VR imaging control method based on image recognition, and the method comprises the following steps: S1, taking pupil monitoring data set by a normal vision user as a standard, setting standard monitoring data, and obtaining a standard index graph according to the standard monitoring data; s2, monitoring the light reflection speed of pupils of the myopic user, obtaining pupil image data of the myopic user, and obtaining pupil monitoring data of the myopic user according to the pupil image data; s3, presetting a training scheme, selecting sample myopia users to carry out grouping training, recording change values of all indexes after training, and obtaining a monitoring surface on the standard index graph; and S4, calculating an area proportion, comparing the area proportions of the lifting surface and the monitoring surface obtained by different training schemes, and selecting the training scheme with the maximum area proportion to train the myopia user.

Description

Technical Field

[0001] The present invention relates to the field of VR imaging technology, and in particular to a VR imaging control method based on image recognition. Background Art

[0002] In recent years, with the rapid development of science and technology and the changes in people's lifestyles, myopia has become more and more common worldwide, especially among teenagers and young people. This trend has prompted myopia training devices to gradually become a highly-regarded solution. Myopia training devices combine advanced technology and scientific methods to provide users with a non-invasive, convenient and effective way to correct vision.

[0003] Myopia training devices usually use virtual reality (VR) technology, which can simulate different visual environments to help users perform targeted vision training; for example, some devices create dynamic three-dimensional images to guide the user's eyes to perform specific movements to exercise eye muscles and improve the eye's ability to adjust.

[0004] In the prior art, most myopia training devices follow a set of preset, standardized training programs to exercise the user's eye muscles; although this method can provide basic training for vision improvement to a certain extent, its limitations are also quite obvious; since the causes, degree and physiological conditions of myopia vary from person to person, a unified training model is often difficult to accurately meet individual needs, thereby limiting the maximization of the training effect; for example, users of different age groups have different eye adjustment abilities and recovery speeds; similarly, the intervention strategies required for pseudomyopia caused by long-term close-range eye use and hereditary high myopia should also be different; in addition, some users may have specific eye diseases or surgical histories, and these special situations need to be specially considered and adjusted in the training program; however, fixed training programs cannot flexibly adapt to these complex and changing situations, resulting in some users experiencing ineffective or even counterproductive training processes. Summary of the invention

[0005] The purpose of the present invention is to provide a VR imaging control method based on image recognition to solve the above technical problems.

[0006] The purpose of the present invention can be achieved through the following technical solutions: A VR imaging control method based on image recognition comprises the following steps: Step S1: Set pupil monitoring data of a user with normal vision as standard monitoring data, wherein the pupil monitoring data includes several indicators; use the value of each indicator in the standard monitoring data as a standard value; obtain the number n of indicators, establish a regular n-gon with equal sides, obtain the center point of the regular n-gon, and connect each vertex with the center point respectively to obtain n line segments, and sequentially associate the n line segments with each indicator; and record the value of the indicator corresponding to the center point on the regular n-gon as 0, and the value of the indicator corresponding to the vertex as the standard value, to obtain a standard indicator graph; Step S2: Step S2: Monitor the pupil light reflex speed of the myopic user and obtain pupil image data of the myopic user in real time; obtain pupil monitoring data of the myopic user according to the pupil image data; According to the value of each indicator in the pupil monitoring data, marking is performed at the corresponding position of each line segment of the standard indicator graph to obtain the numerical point of each indicator on the corresponding line segment; Step S3: Preset a number of training schemes, wherein the training schemes are composed of different VR imaging; select a number of sample myopic users, group the sample myopic users, train each group using a training scheme, and obtain the change value of each indicator of the sample myopic users of each group after training; Step S4: on the standard indicator diagram, each numerical point is moved according to the change value to obtain a new numerical point, and each new numerical point is connected to obtain a monitoring surface; and the area R occupied by the monitoring surface is obtained to obtain an area ratio P=R / S, where S is the area of ​​the standard indicator diagram; The area ratio of the monitoring surface of each training program is obtained, and the training program with the largest area ratio is selected to train the myopic user.

[0007] As a further solution of the present invention: the indicators include maximum pupil diameter, minimum pupil diameter, pupil change rate, maximum contraction speed and average expansion speed.

[0008] As a further solution of the present invention: the process of setting the standard monitoring data includes: All indicators of several users with normal vision are collected, and the average value of each user with normal vision on the same indicator is obtained, which is recorded as the indicator mean; the pupil monitoring data whose value of each indicator is the indicator mean is recorded as the standard monitoring data.

[0009] As a further solution of the present invention: when among the indicators of the myopic user, there is an indicator whose value is greater than the indicator mean, the value of the indicator is directly recorded as being equal to the indicator mean.

[0010] As a further solution of the present invention: the process of monitoring pupil light reflex speed includes: The monitoring time is set, and pupil images of both sides of the myopic user are obtained in real time during the monitoring time; and an infrared light source is used to aim at one pupil of the myopic user during the monitoring time to perform light stimulation for a time of t, where t is generally 100ms, to obtain pupil image data.

[0011] As a further solution of the present invention: the setting process of the training program includes: Preset m types of VR imaging, set a training cycle, and divide the training cycle into num sub-cycles; group the m types of VR imaging to obtain a number of imaging combinations, where there are N types of VR imaging, 1≤N≤m; for any imaging combination, obtain the number of VR imaging types N in the combination, then divide the num sub-cycles into N parts in the ratio of num1:num2:...:num N , where num1 represents the number of sub-periods in the first part, and num1+num2+...+num N =num; then each sub-period is divided into M different proportions, and each proportion is associated with an imaging combination, and each associated proportion and imaging combination is recorded as a training scheme.

[0012] As a further solution of the present invention: the process of obtaining the change value includes: Get the values ​​of the indicators of each sample myopic user before training {Z1, Z2, ..., Z i}, where Z1 represents the value of the index of the first sample myopia user before training, and i is the total number of sample myopia users; and obtain the value of the index of each sample myopia user after training {Z1´, Z2´, ..., Z i ´}, where Z1´ represents the value of the indicator after the first sample myopia user is trained; then the change value of the indicator is obtained , where s∈[1,i] and s is a positive integer.

[0013] As a further solution of the present invention: when there are several training schemes with equal area proportions, any one training scheme is randomly selected to train the myopic user.

[0014] Beneficial effects of the present invention: The present invention monitors the pupil image data of myopic users in real time and adjusts the training program according to these data, so that the training is more in line with the individual's physiological conditions and needs. This personalized method can intervene more effectively in the specific problems of users, rather than adopting a "one-size-fits-all" standard training program. Using image recognition-based technology to optimize the selection of VR imaging can provide more accurate training parameters for each user, thereby improving the training effect. This method can reduce ineffective training or even counter-effects caused by inappropriate training intensity or methods. The present invention can also dynamically adjust the training content according to the monitored data, so it can adapt to users of different ages, myopia degrees and different eye conditions. Whether it is pseudomyopia or hereditary high myopia, or even users with special eye disease history, a more suitable training program can be obtained. Providing a better user experience will help promote the application and development of VR technology in education, entertainment, medical and other fields. At the same time, the large amount of data generated can also be used for scientific research, helping to better understand the human visual system and develop new vision training methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] The present invention will be further described below in conjunction with the accompanying drawings.

[0016] Figure 1 It is a flow chart of a VR imaging control method based on image recognition according to the present invention. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] See also Figure 1 As shown, the present invention is a VR imaging control method based on image recognition, comprising the following steps: Step S1: Set pupil monitoring data of a user with normal vision as standard monitoring data, wherein the pupil monitoring data includes several indicators; use the value of each indicator in the standard monitoring data as a standard value; obtain the number n of indicators, establish a regular n-gon with equal sides, obtain the center point of the regular n-gon, and connect each vertex with the center point respectively to obtain n line segments, and sequentially associate the n line segments with each indicator; and record the value of the indicator corresponding to the center point on the regular n-gon as 0, and the value of the indicator corresponding to the vertex as the standard value, to obtain a standard indicator graph; It is worth noting that since pupil monitoring data varies with the age of the user, due to differences in physiological structure, degree of development of the nervous system and visual needs, key indicators such as pupil reaction speed to light changes, maximum and minimum pupil diameters will vary; therefore, when setting standard monitoring data, it is necessary to divide age groups, and different age groups correspond to different standard monitoring data; the age groups include childhood, adolescence to adulthood, middle-aged and elderly stages; As a preferred embodiment of the present invention, the indicators include maximum pupil diameter, minimum pupil diameter, pupil change rate, maximum contraction speed and average expansion speed; It is understandable that the maximum pupil diameter of an adult user with normal vision is usually [4, 8] mm, and the minimum pupil diameter is usually [2, 3] mm; As a preferred embodiment of the present invention, the process of setting the standard monitoring data includes: Collect all indicators of several users with normal vision, and obtain the average value of the same indicator of each user with normal vision, which is recorded as the indicator mean; then record the pupil monitoring data whose value of each indicator is the indicator mean as the standard monitoring data; It is understandable that to set the pupil monitoring data standard for users with normal vision, it is first necessary to collect all the indicators of several users with normal vision; these indicators include but are not limited to the maximum pupil diameter, the minimum pupil diameter, the pupil change rate, the maximum contraction speed and the average expansion speed, etc.; by obtaining the average value of the same indicator for each user with normal vision, recorded as the indicator mean, the pupil monitoring data with the values ​​of these indicators as the indicator mean can be used as the standard monitoring data; ensure that the standard monitoring data is representative and accurate, and provide a reliable basis for subsequent vision correction plans; Step S2: monitoring the pupil light reflex speed of the myopic user and acquiring pupil image data of the myopic user in real time; and obtaining pupil monitoring data of the myopic user according to the pupil image data; According to the value of each indicator in the pupil monitoring data, marking is performed at the corresponding position of each line segment of the standard indicator graph to obtain the numerical point of each indicator on the corresponding line segment; Understandably, pupillary light reflex speed monitoring is one of the important methods for assessing eye health and visual function; this monitoring usually involves measuring the pupil's reaction speed to light changes, that is, the speed at which the pupil shrinks from a larger state (dark adaptation) to a smaller state (light adaptation); the pupillary light reflex is a natural protective mechanism of the eye, designed to regulate the amount of light entering the eye to protect the retina from excessive light damage; when the light increases, the pupil shrinks rapidly; when the light decreases, the pupil dilates; this process is controlled by the autonomic nervous system in the brain; pupillary light reflex speed monitoring focuses on the time required for the pupil to shrink from its maximum diameter to its minimum diameter; this time reflects the pupil's sensitivity and reaction speed to light changes; As a preferred embodiment of the present invention, when among the indicators of the myopic user, there is an indicator whose value is greater than the indicator mean, the value of the indicator is directly recorded as being equal to the indicator mean; It is understandable that when evaluating pupil monitoring data of myopic users, if it is found that the value of a certain indicator exceeds the average value of users with normal vision (i.e., the indicator mean), then in order to simplify processing or avoid the influence of outliers, this excessively high indicator value will be adjusted to the same level as that of users with normal vision, i.e., the indicator mean; As a preferred embodiment of the present invention, the process of monitoring pupil light reflex speed includes: Setting a monitoring time, obtaining pupil images of both sides of the myopic user in real time during the monitoring time; and using an infrared light source to aim at one pupil of the myopic user during the monitoring time, performing light stimulation for a time of t, where t is generally set to 100ms, to obtain pupil image data; It can be understood that the pupil image data is acquired based on an image acquisition device; Step S3: Preset a number of training schemes, wherein the training schemes are composed of different VR imaging; select a number of sample myopic users, group the sample myopic users, train each group using a training scheme, and obtain the change value of each indicator of the sample myopic users of each group after training; It can be understood that the change value represents the change range of each pupil monitoring index after the myopic user uses a specific training program; As a preferred embodiment of the present invention, the process of setting the training program includes: Preset m types of VR imaging, set a training cycle, and divide the training cycle into num sub-cycles; group the m types of VR imaging to obtain a number of imaging combinations, where there are N types of VR imaging, 1≤N≤m; for any imaging combination, obtain the number of VR imaging types N in the combination, then divide the num sub-cycles into N parts in the ratio of num1:num2:...:num N, where num1 represents the number of sub-periods in the first part, and num1+num2+...+num N =num; then each sub-period is divided into M different proportions, and each proportion is associated with an imaging combination, and each associated proportion and imaging combination is recorded as a training scheme; It should be noted that different VR imaging has different effects on the vision training of myopic users, so different VR combinations have different training effects, and the number of sub-cycles occupied by each VR imaging in different imaging combinations is different, and the focus of the training program is different; the number of sub-cycles reflects the time allocation of different vision training; As a preferred embodiment of the present invention, the process of obtaining the change value includes: Get the values ​​of the indicators of each sample myopic user before training {Z1, Z2, ..., Z i}, where Z1 represents the value of the index of the first sample myopia user before training, and i is the total number of sample myopia users; and obtain the value of the index of each sample myopia user after training {Z1´, Z2´, ..., Z i ´}, where Z1´ represents the value of the indicator after the first sample myopia user is trained; then the change value of the indicator is obtained , where s∈[1,i] and s is a positive integer; It can be understood that the change value can more intuitively understand the overall improvement of the myopic user group after training. If the change value is a positive value, it means that the index value of the user has improved after training; Step S4: on the standard indicator diagram, each numerical point is moved according to the change value to obtain a new numerical point, and each new numerical point is connected to obtain a monitoring surface; and the area R occupied by the monitoring surface is obtained to obtain an area ratio P=R / S, where S is the area of ​​the standard indicator diagram; Obtaining the area ratio of the monitoring surface of each training program, selecting the training program with the largest area ratio, and performing training on the myopic user; It should be noted that the process of moving each value point according to the change value includes: In the standard Standard In the figure, the direction from the center point to the vertex is recorded as the outside, and the direction from the vertex to the center point is recorded as the inside; if the change value is a positive value, the value z of the change value is obtained, and the value point is moved outward by a distance z; if the change value is a negative value, the value z of the change value is obtained, and the value point is moved inward by a distance z; As a preferred embodiment of the present invention, when there are several training schemes with equal area proportions, any one training scheme is randomly selected to train the myopic user.

[0019] The above is a detailed description of an embodiment of the present invention, but the content is only a preferred embodiment of the present invention and cannot be considered to limit the scope of implementation of the present invention. All equivalent changes and improvements made within the scope of the present invention should still fall within the scope of the patent coverage of the present invention.

Claims

1. A VR imaging control method based on image recognition, characterized in that: The following steps are involved: Step S1: Set pupil monitoring data of a user with normal vision as standard monitoring data, wherein the pupil monitoring data includes several indicators; use the value of each indicator in the standard monitoring data as a standard value; obtain the number n of indicators, establish a regular n-gon with equal sides, obtain the center point of the regular n-gon, and connect each vertex with the center point respectively to obtain n line segments, and sequentially associate the n line segments with each indicator; and record the value of the indicator corresponding to the center point on the regular n-gon as 0, and the value of the indicator corresponding to the vertex as the standard value, to obtain a standard indicator graph; Step S2: monitoring the pupil light reflex speed of the myopic user and obtaining pupil image data of the myopic user in real time; Obtaining pupil monitoring data of a myopic user according to the pupil image data; According to the value of each indicator in the pupil monitoring data, marking is performed at the corresponding position of each line segment of the standard indicator graph to obtain the numerical point of each indicator on the corresponding line segment; Step S3: Preset a number of training schemes, wherein the training schemes are composed of different VR imaging; select a number of sample myopic users, group the sample myopic users, train each group using a training scheme, and obtain the change value of each indicator of the sample myopic users of each group after training; Step S4: on the standard indicator diagram, each numerical point is moved according to the change value to obtain a new numerical point, and each new numerical point is connected to obtain a monitoring surface; and the area R occupied by the monitoring surface is obtained to obtain an area ratio P=R / S, where S is the area of ​​the standard indicator diagram; The area ratio of the monitoring surface of each training program is obtained, and the training program with the largest area ratio is selected to train the myopic user.

2. The VR imaging control method based on image recognition according to claim 1, characterized in that: In step S1, the indicators include maximum pupil diameter, minimum pupil diameter, pupil change rate, maximum contraction speed and average dilation speed.

3. The VR imaging control method based on image recognition according to claim 1, characterized in that: In step S1, the process of setting the standard monitoring data includes: All indicators of several users with normal vision are collected, and the average value of each user with normal vision on the same indicator is obtained, which is recorded as the indicator mean; the pupil monitoring data whose value of each indicator is the indicator mean is recorded as the standard monitoring data.

4. The VR imaging control method based on image recognition according to claim 3 is characterized in that: In step S2, when there is an indicator whose value is greater than the indicator mean among the indicators of the myopic user, the value of the indicator is directly recorded as being equal to the indicator mean.

5. The VR imaging control method based on image recognition according to claim 1, characterized in that: In step S2, the process of monitoring pupil light reflex speed includes: The monitoring time is set, and pupil images of both sides of the myopic user are obtained in real time during the monitoring time; and an infrared light source is used to aim at one pupil of the myopic user during the monitoring time to perform light stimulation for a time of t, where t is generally 100ms, to obtain pupil image data.

6. The VR imaging control method based on image recognition according to claim 1, characterized in that: In step S3, the training program setting process includes: Preset m types of VR imaging, set a training cycle, and divide the training cycle into num sub-cycles; group the m types of VR imaging to obtain a number of imaging combinations, where there are N types of VR imaging, 1≤N≤m; for any imaging combination, obtain the number of VR imaging types N in the combination, then divide the num sub-cycles into N parts in the ratio of num1:num2:...:num N , where num1 represents the number of sub-periods in the first part, and num1+num2+...+num N =num; then each sub-period is divided into M different proportions, and each proportion is associated with an imaging combination, and each associated proportion and imaging combination is recorded as a training scheme.

7. The VR imaging control method based on image recognition according to claim 1, characterized in that: In step S3, the process of obtaining the change value includes: Get the values ​​of the indicators of each sample myopic user before training {Z1, Z2, ..., Z i }, where Z1 represents the value of the index of the first sample myopia user before training, and i is the total number of sample myopia users; and obtain the value of the index of each sample myopia user after training {Z1´, Z2´, ..., Z i ´}, where Z1´ represents the value of the indicator after the first sample myopia user is trained; then the change value of the indicator is obtained , where s∈[1,i] and s is a positive integer.

8. The VR imaging control method based on image recognition according to claim 1, characterized in that: In step S4, when there are several training schemes with equal area proportions, any one training scheme is randomly selected to train the myopic user.

Images