Myopia prevention and control system based on visual environment spatial frequency and contrast polarity regulation

Abstract

The myopia prevention and control system based on visual environment space frequency and contrast polarity regulation adopts standardization processing and feature extraction, and is aimed at problems of abnormal space frequency spectrum distribution and imbalance of contrast polarity in indoor environments such as classrooms, takes a visual environment acquisition module as input, combines space frequency spectrum analysis and a contrast polarity statistical model to construct a quantifiable evaluation index system, realizes extraction, evaluation and grading of key statistical characteristics of the visual environment, and provides replicable, comparable and generalizable parameterized basis for myopia prevention and control; on the basis, the scheme can also combine display / projection / illumination texture and other output modes to assist in parameter adjustment of scene visual characteristics, so that the statistical characteristics of the visual environment are improved, and the prevention and control effect is enhanced.

Description

Technical Field

[0001] This invention relates to the field of myopia prevention and control technology, specifically to a myopia prevention and control system based on the spatial frequency and contrast polarity regulation of the visual environment. Background Technology

[0002] Myopia, a common type of refractive error among children and adolescents, has seen a rapid increase in prevalence in recent years, becoming a significant public health issue. The development of myopia is influenced by both genetic and environmental factors, but the epidemiological trends of recent decades cannot be explained solely by genetics; environmental and behavioral factors are considered the primary driving forces. Among these, visual environmental stimuli, as direct external factors acting on the retina and influencing eye development regulation, have received widespread attention; however, engineerable evaluation methods are still lacking for their key stimulation parameters.

[0003] Studies have found significant differences in visual statistical characteristics between natural outdoor scenes and artificial indoor scenes. Natural scenes typically exhibit stable spatial frequency spectrum characteristics, showing a typical 1 / fα distribution; while artificial indoor environments often show reduced high-frequency visual components and steeper spatial frequency slopes. Furthermore, natural scenes tend to have a higher proportion of negative contrast polarity information across multiple spatial scales, while this type of information is relatively lacking in artificial environments. This may lead to insufficient or biased stimulation of retinal neural coding, thereby affecting normal eye development and regulation and increasing the risk of myopia.

[0004] In indoor environments, existing technologies lack a systematic and quantifiable assessment scheme for visual environment characteristics, such as spatial frequency distribution and contrast polarity, making standardized deployment and continuous intervention difficult. Therefore, there is an urgent need for a myopia control system that can comprehensively evaluate the spatial frequency characteristics and contrast polarity of the visual environment in indoor settings such as classrooms, and assist in myopia risk assessment. This would help construct an artificial visual environment that more closely resembles the statistical characteristics of natural scenes, thereby improving the quality of retinal nerve development stimulation, reducing abnormal visual signal transmission, and ultimately delaying or reducing the onset and progression of myopia. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a myopia prevention and control system based on the spatial frequency and contrast polarity regulation of the visual environment.

[0006] The technical solution adopted in this invention is: a myopia prevention and control system based on spatial frequency and contrast polarity modulation of the visual environment, comprising: Image input and acquisition module: used to acquire visual environment image data to be evaluated and import the images into the calculation module for processing; Color and brightness preprocessing module: converts the input visual environment image data into a standardized signal representation suitable for subsequent contrast polarity and spatial frequency spectrum analysis; The geometry and sampling normalization module performs geometric cropping and sampling standardization on the input visual environment image to eliminate scale differences caused by different cameras or data sources, making different visual scenes more comparable. Spatial frequency spectrum analysis module: Quantitatively evaluates the standardized signals after color and brightness preprocessing to obtain the spatial frequency energy distribution structure of the visual scene; Contrast Polarity Analysis Module: Extracts multi-scale local contrast from the standardized signals after color and brightness preprocessing, and statistically analyzes the ON / OFF polarity distribution characteristics; The quantifiable evaluation index system construction module integrates the spatial frequency energy distribution structure and ON / OFF polarity distribution characteristics of the obtained visual scene to form a quantifiable evaluation index system output.

[0007] The color and brightness preprocessing module performs sRGB inverse gamma correction on the input image to linearize the image, and constructs the CIE relative brightness channel Y based on this for brightness statistics and contrast polarity analysis, and constructs the LMS cone cell space for spatial frequency spectrum analysis.

[0008] The geometry and sampling normalization module performs a central square crop on the image; downsamples the image to a uniform pixel scale and enables anti-aliasing; estimates the horizontal and vertical field of view of the image based on the EXIF ​​information of the image, and takes the smaller value as the effective field of view; when the EXIF ​​information is missing or abnormal, a preset backsliding field of view is used, and the angular resolution is calculated based on the pixel size and field of view of the cropped image.

[0009] The spatial frequency spectrum analysis module performs a two-dimensional FFT in the LMS cone cell space to construct a composite amplitude spectrum and obtain a spectral profile by radial averaging; it performs linear regression fitting on the logarithmic-logarithmic relationship of the spectral curve and outputs the spatial frequency spectrum slope α.

[0010] The contrast polarity analysis module uses a multi-scale differential Gaussian (DoG) filter to perform DoG filter convolution on each preset spatial frequency scale in the luminance channel Y, calculates the response of the center and the surrounding area, calculates the Contrast according to the formula, completes ON / OFF classification based on the Contrast sign, statistically analyzes the proportion of OFF pixels and ON pixels at each scale, and outputs a contrast polarity statistical feature vector.

[0011] The visual environment image data in the image input and acquisition module includes real-shot image data of indoor learning environments and publicly available datasets of natural or artificial environment images.

[0012] The quantifiable evaluation index system construction module is based on a large sample scene database to build a natural scene friendliness score, which is divided into three levels: low, medium and high to reflect the degree to which the scene is "close to the natural visual environment". It also collects clinical data related to myopia in children and gradually optimizes the evaluation criteria and classification.

[0013] The output of the quantifiable evaluation index system construction module includes evaluation vectors, evaluation scores, and risk classification results.

[0014] The beneficial effects of this invention are as follows: A myopia prevention and control system based on spatial frequency and contrast polarity regulation of the visual environment employs standardized processing and feature extraction. Through color linearization, geometric clipping, and contrast calculation, it ensures comparability across different environments and devices, providing unified basic data for subsequent analysis. Relying on contrast polarity analysis and spatial frequency spectrum analysis modules, it quantitatively assesses the characteristics of the visual environment through multi-scale processing and spectrum analysis. By combining visual environment characteristics with myopia data, a comprehensive risk assessment system is constructed, providing accurate predictive basis for personalized myopia intervention. This invention addresses issues such as abnormal spatial frequency spectrum distribution and contrast polarity imbalance in indoor environments such as classrooms. Using a visual environment acquisition module as input, it constructs a quantifiable evaluation index system by combining spatial frequency spectrum analysis and contrast polarity statistical models. This enables the extraction, evaluation, and classification of key statistical features of the visual environment, providing replicable, comparable, and scalable parameterized basis for myopia prevention and control. Furthermore, this solution can also combine display / projection / lighting texture and other output methods to adjust auxiliary parameters of scene visual features, thereby improving the statistical characteristics of the visual environment and enhancing the prevention and control effect. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating the workflow of the myopia prevention and control system based on the spatial frequency and contrast polarity regulation of the visual environment according to the present invention.

[0016] Figure 2 ON / OFF classification of the original image at a low-frequency environmental scale.

[0017] Figure 3 ON / OFF classification of the original image at the mid-frequency environmental scale.

[0018] Figure 4 ON / OFF classification of the original image at high-frequency environmental scales.

[0019] Figure 5 This is a schematic diagram of spatial frequency distribution analysis. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0021] System / parameter composition of the present invention The myopia prevention and control evaluation system and method based on visual environment spatial frequency spectrum analysis and contrast polarity statistical model provided by this invention includes at least the following functional modules and parameter system: Image input and acquisition module This module is used to acquire visual environment image data to be evaluated. Image sources include, but are not limited to: Real photos of indoor learning environments (such as classrooms, home study desks, bedrooms, etc.); Publicly available datasets of images of natural scenes or artificial environments.

[0022] Color and brightness preprocessing parameter system The method for converting the input image into a standardized signal representation suitable for subsequent contrast polarity and spatial frequency spectrum analysis includes at least: Linearization is achieved by performing sRGB inverse gamma correction on the input image; Construct a CIE relative luminance channel Y for luminance statistics and contrast polarity analysis; LMS cone cell space was constructed for spatial frequency spectrum analysis.

[0023] Geometric and Sampling Standardization Parameter System To eliminate scale differences caused by different cameras or data sources, and to make different visual scenes more comparable, at least the following should be included: Perform a center square crop on the input image; Downsample the image to a uniform size and enable anti-aliasing; Estimate the horizontal and vertical field of view of the image based on the EXIF ​​information, and take the smaller value as the effective field of view. When EXIF ​​information is missing or abnormal, a preset back-off field of view is used, and the angular resolution is calculated based on the pixel size and field of view of the cropped image.

[0024] Contrast Polarity Analysis Parameter System Used to extract multi-scale local contrast and statistically analyze ON / OFF polarity distribution features, including at least: Multi-scale differential Gaussian (DoG) filter for contrast calculation at different spatial frequency scales; Calculate image contrast and classify it; Calculate the ratio of ON / OFF pixels; Output multi-scale OFF pixel proportion vector.

[0025] Spatial frequency spectrum analysis parameter system For quantitative evaluation of the spatial frequency energy distribution structure of a visual scene, at least the following are included: Perform a two-dimensional fast Fourier transform (FFT) in the LMS space; Construct a composite amplitude spectrum and obtain a one-dimensional spectral curve by radial averaging; Log-log linear regression was performed on the spectral curve to calculate the spatial frequency spectrum slope α.

[0026] Module for constructing a quantifiable evaluation index system This module aims to construct a quantifiable myopia risk assessment system by combining image analysis results with myopia-related data (such as refractive error, axial length, and myopia progression rate) to guide visual optimization in the living environment. It is expected to cover the following data and analysis directions: Comprehensive Evaluation and Risk Scoring: Based on a large-sample scene database (20,000 natural environment images and 15,600 indoor environment images), this study constructed a Protective Exposure Score (PES) based on image statistical features to reflect the degree to which a scene "closes to the natural visual environment" (a higher score indicates greater closeness to natural scene features). Images were graded (low / medium / high) based on the ternary rank of the PES. The evaluation criteria and grading methods in this part need to be validated and improved with more clinical data. Examples are attached. Figure 2-5 As shown.

[0027] The correlation between image analysis results and myopia data: Explore the potential correlation between image analysis results (such as contrast polarity, spatial frequency spectrum, etc.) and myopia-related data, further assess the impact of environmental factors on the occurrence and progression of myopia, and optimize assessment criteria and classification.

[0028] A comprehensive evaluation method for a myopia prevention and control system based on visual environment spatial frequency and contrast polarity modulation, using the present invention, includes the following steps: S1: Image Acquisition and Import Acquire image data of the target scene and import the images into the computing module for processing.

[0029] S2: Color linearization and construction of luminance / cone representation Perform sRGB inverse gamma correction on the input image to linearize it, and build upon this: The luminance channel Y is used for contrast polarity analysis; LMS cone cell space is used for spatial frequency spectrum analysis.

[0030] S3: Geometric Cutting and Sampling Standardization Perform a center square crop on the image; downsample the image to a uniform pixel scale and enable anti-aliasing; estimate the horizontal and vertical field of view of the image based on the EXIF ​​information of the image, and take the smaller value as the effective field of view; when the EXIF ​​information is missing or abnormal, use a preset backsliding field of view, and calculate the angular resolution based on the pixel size and field of view of the cropped image.

[0031] S4: Spatial Frequency Spectrum Calculation and Slope Fitting Perform a two-dimensional FFT in LMS space to construct a composite amplitude spectrum and obtain a spectral profile by radial averaging; perform linear regression fitting on the logarithmic-logarithmic relationship of the spectral curve to output the spatial frequency spectrum slope α.

[0032] S5: Multi-scale DoG filtering and contrast calculation On the luminance channel Y, a DoG filter convolution is performed on each preset spatial frequency scale to calculate the response at the center and the surrounding area, and Contrast is calculated according to the formula.

[0033] S6: Contrast Polarity Classification and Statistical Output Based on the Contrast symbol, complete the ON / OFF classification and perform statistics at each scale: OFF pixel ratio; ON pixel ratio; Output the contrast polarity statistical feature vector.

[0034] S7: Indicator Integration and Quantitative Evaluation Output (Key Step) The obtained features are fused to form a quantifiable evaluation index system, including but not limited to: Evaluation vectors (α, OFF percentage, etc.); Evaluation and rating; Risk classification results (e.g., low risk / medium risk / high risk level).

[0035] Features of this invention: Standardization and feature extraction: By linearizing color, clipping geometrically, and calculating contrast, we ensure comparability across different environments and devices, providing a unified foundation of data for subsequent analysis.

[0036] Contrast polarity analysis and spatial frequency spectrum analysis module: Quantitatively assess the characteristics of the visual environment through multi-scale processing and spectral analysis.

[0037] Quantifiable evaluation index system: By combining visual environment characteristics with myopia data, a comprehensive risk assessment system is constructed to provide accurate predictive basis for personalized myopia intervention.

[0038] The above description is merely a preferred embodiment of the present invention, and the scope of protection of the present invention is not limited to the above embodiments and experimental examples. Any technical solution following the concept of the present invention is included within the scope of protection of the present invention. It should be emphasized that for those skilled in the art, any modifications or equivalent substitutions without departing from the spirit and scope of the present invention should be considered as part of the scope of protection of the present invention.

Claims

1. A myopia control system based on spatial frequency and contrast polarity modulation of the visual environment, characterized in that, include: Image input and acquisition module: used to acquire visual environment image data to be evaluated and import the images into the calculation module for processing; Color and brightness preprocessing module: converts the input visual environment image data into a standardized signal representation suitable for subsequent contrast polarity and spatial frequency spectrum analysis; The geometry and sampling standardization module performs geometric cropping and sampling standardization on the input visual environment image to eliminate scale differences caused by different cameras or data sources, making different visual scenes comparable. Spatial frequency spectrum analysis module: Quantitatively evaluates the standardized signals after color and brightness preprocessing to obtain the spatial frequency energy distribution structure of the visual scene; Contrast Polarity Analysis Module: Extracts multi-scale local contrast from the standardized signals after color and brightness preprocessing, and statistically analyzes the ON / OFF polarity distribution characteristics; The quantifiable evaluation index system construction module integrates the spatial frequency energy distribution structure and ON / OFF polarity distribution characteristics of the obtained visual scene to form a quantifiable evaluation index system output.

2. The myopia prevention and control system based on spatial frequency and contrast polarity modulation of the visual environment according to claim 1, characterized in that, The color and brightness preprocessing module performs sRGB inverse gamma correction on the input image to linearize the image, and constructs the CIE relative brightness channel Y based on this for brightness statistics and contrast polarity analysis, and constructs the LMS cone cell space for spatial frequency spectrum analysis.

3. The myopia prevention and control system based on spatial frequency and contrast polarity modulation of the visual environment according to claim 1, characterized in that, The geometry and sampling normalization module performs a central square crop on the image; downsamples the image to a uniform pixel scale and enables anti-aliasing; estimates the horizontal and vertical field of view of the image based on the EXIF ​​information of the image, and takes the smaller value as the effective field of view; when the EXIF ​​information is missing or abnormal, a preset backsliding field of view is used, and the angular resolution is calculated based on the pixel size and field of view of the cropped image.

4. The myopia prevention and control system based on spatial frequency and contrast polarity modulation of the visual environment according to claim 2, characterized in that, The spatial frequency spectrum analysis module performs a two-dimensional FFT in the LMS cone cell space to construct a composite amplitude spectrum and obtain a spectral profile by radial averaging; it performs linear regression fitting on the logarithmic-logarithmic relationship of the spectral curve and outputs the spatial frequency spectrum slope α.

5. The myopia control system based on spatial frequency and contrast polarity modulation of the visual environment according to claim 2, characterized in that, The contrast polarity analysis module uses a multi-scale differential Gaussian (DoG) filter to perform DoG filter convolution on each preset spatial frequency scale in the luminance channel Y, calculates the response of the center and the surrounding area, calculates the Contrast according to the formula, completes ON / OFF classification based on the Contrast sign, statistically analyzes the proportion of OFF pixels and ON pixels at each scale, and outputs a contrast polarity statistical feature vector.

6. The myopia prevention and control system based on spatial frequency and contrast polarity modulation of the visual environment according to claim 1, characterized in that, The visual environment image data in the image input and acquisition module includes real-shot image data of indoor learning environments and publicly available datasets of natural or artificial environment images.

7. The myopia prevention and control system based on spatial frequency and contrast polarity modulation of the visual environment according to claim 1, characterized in that, The quantifiable evaluation index system construction module is based on a large sample scene database to build a natural scene-friendly score, which is divided into three levels: low, medium and high to reflect the degree to which the scene is "close to the natural visual environment". It also collects clinical data related to myopia in children and gradually optimizes the evaluation criteria and classification.

8. The myopia prevention and control system based on spatial frequency and contrast polarity modulation of the visual environment according to claim 1, characterized in that, The output of the quantifiable evaluation index system construction module includes evaluation vectors, evaluation scores, and risk classification results.

Images