Eye behavior analysis method and system based on artificial intelligence

Abstract

The invention relates to an artificial intelligence-based eye behavior analysis method and system, and the method comprises the steps: carrying out the radial segmentation from two horizontal flag bits of two eye images, and dividing the eye images into a plurality of sub-regions; a UX tuple sequence is constructed based on the sub-regions, which is compared to a sample to determine ocular behavior. According to the method, the eye behaviors can be recognized and responded in real time by quickly and continuously judging the eye behaviors, and the method is suitable for application scenes needing quick response.

Description

[Technical field]

[0001] The present invention belongs to the technical field of posture retrieval based on artificial intelligence, and in particular relates to an eye behavior analysis method and system based on artificial intelligence. [Background technology]

[0002] With the rapid development of computer hardware and software technology, multi-modal human-computer interaction (MMHCl) has become a very popular field in human-computer interaction software engineering methods at home and abroad. Because multi-modal human-computer interaction is more in line with people's psychological and physiological habits, it is considered to be a more natural way of human-computer interaction. Artificial intelligence is an important driving force leading a new round of scientific and technological revolution and industrial transformation, and is profoundly changing people's production, life and learning methods. Artificial intelligence technology has moved from theoretical research to practical application, profoundly changing production and lifestyle. In recent years, artificial intelligence large model technology has developed rapidly, relying on "big data + big computing power + strong algorithm" training to improve the ability to handle complex tasks. Artificial intelligence is shifting from single-point application to diversified application, penetrating into industry-specific scenarios, and accelerating towards a new stage of comprehensive application. The breakthrough of large models and the rise of generative artificial intelligence have brought advanced tools for product and process innovation to various industries. The rapid development and widespread application of artificial intelligence technology are driving society into a new era of intelligence and digitalization. With the continuous advancement of technology and strong policy support, artificial intelligence is expected to play a key role in more fields and promote the high-quality development of the economy and society.

[0003] The concept of eye tracking first emerged in the 1930s. The original design purpose was for commercial considerations. However, due to the limitations of sensor technology and information processor computing speed at that time, eye tracking technology encountered many difficulties in its development and developed very slowly. Eye movement behavior accompanies almost all human activities, and eye movement information contains important clues that can reveal different human behaviors. Eye tracking is a scientific application technology. Users can send information and receive feedback without physical contact with interactive devices. In principle, eye tracking mainly studies the acquisition, modeling and simulation of eye movement information, and has a wide range of uses. However, the current analysis methods are mainly based on traditional analysis of eye images or methods that do not distinguish the specific categories of each part of the eye image, lack the expression of eye image information, or lack the understanding of the characteristics of the eye image composition when expressing it. When preprocessing the eye image, there is a lack of utilization of eye characteristics, resulting in high complexity or poor scalability; Based on the above problems, the present invention can identify and respond to eye behavior in real time by quickly and continuously making eye behavior judgments, and is suitable for application scenarios that require rapid response. [Summary of the invention]

[0004] In order to solve the above problems in the prior art, the present invention proposes an eye behavior analysis method and system based on artificial intelligence, the method comprising:

[0005] Step S1: Acquire an eye image and preprocess it; specifically: perform parabola-based shape preprocessing and area-based size preprocessing;

[0006] Step S11: obtaining a horizontal marker, and obtaining a vertical marker based on the horizontal marker; dividing the eye image into four regions based on the vertical marker and the horizontal marker, and performing a parabola-based shape preprocessing and an area-based size preprocessing on each region;

[0007] Step S12: the obtaining of the horizontal markers is specifically as follows: segmenting the eye image to obtain the eye region; obtaining the positions of the two end points of the longest diameter of the eye region as two horizontal markers, constructing an X-axis based on the horizontal markers, and taking the positions of the points with the longest vertical distance from the X-axis as two vertical markers; and constructing a Y-axis based on the two vertical markers;

[0008] Preferably: when the Y axis formed by the two vertical markers is not perpendicular to the X axis, two vertical markers are selected to construct a Y axis perpendicular to the X axis at the center point of the X axis projection, and the intersection of the Y axis and the edge of the eye area is used as the vertical marker; of course, there may be a variety of adjustment and vertical marker setting methods, which will not be described in detail here;

[0009] Step S13: Divide the eye image into four areas based on the X-axis and the Y-axis; connect the horizontal mark and the vertical mark in the same area to form a stretching auxiliary line; determine the first projection height L of the horizontal mark on the Y-axis; determine the second projection height C of the horizontal mark on the X-axis;

[0010] Step S14: performing parabola-based shape preprocessing on each region; specifically, stretching the region along the first direction or the second direction and performing pixel interpolation until the stretching condition is met to complete the shape preprocessing;

[0011] Step S15: stretching the area along the X-axis toward the outside of the eye area so that the first projection height L reaches a first projection height preset value, thereby completing the size preprocessing based on the area;

[0012] Step S2: starting from the left horizontal mark AL and the right horizontal mark AR of the two horizontal marks A as the division center, U1 left rays and U1 right rays are set in the part between the upper tangent formed by the left mark AL and the lower tangent formed by the lower part of the eye area; the U1 left ray and U1 right ray divide the eye image into UX sub-areas;

[0013] Step S3: Ternary each sub-region to obtain a triplet corresponding to each sub-region, and arrange all the triplet according to the distribution position of the sub-region to obtain a UX tuple corresponding to the eye region ((I i，0 , I i，1 , I i，2 )); where: i is the sub-area number; I i，0 is the number of pixels belonging to the sclera region in sub-region i; i，1 is the number of pixels belonging to the iris region in sub-region i; i，2 is the number of pixels belonging to the pupil area in sub-area i; i = 1 ~ UX;

[0014] Step S4: Use the N-length current UX tuple sequence composed of the UX tuples corresponding to the most recent N consecutive human eye areas to compare with the UX tuple sample sequence in the eye behavior sample library to determine the similarity between the current UX tuple sequence and the UX tuple sample sequence; select the eye behavior corresponding to the UX tuple sample sequence corresponding to the one with the smallest similarity as the determined eye behavior.

[0015] Furthermore, the stretching condition is that any point in the edge of the area coincides with a preset parabola passing through the horizontal mark position and the vertical mark position, or the second projection height C reaches a preset value of the second projection height after stretching.

[0016] Furthermore, the first direction is the direction of the Y axis toward the outside of the eye area.

[0017] Furthermore, the second direction is perpendicular to the stretching auxiliary line and is directed toward the outside of the eye area.

[0018] Further, the first direction or the second direction is similarly selected for stretching according to the positions of the four regions.

[0019] Furthermore, the U left rays and the U right rays are evenly arranged.

[0020] Furthermore, an artificial intelligence model is used to classify each pixel in the eye image so that the pixel belongs to the sclera, iris or pupil.

[0021] An eye behavior analysis system based on artificial intelligence, comprising: the system is used to implement the eye behavior analysis method based on artificial intelligence.

[0022] An eye behavior analysis platform based on artificial intelligence includes a processor, the processor is coupled to a memory, the memory stores program instructions, and when the program instructions stored in the memory are executed by the processor, the eye behavior analysis method based on artificial intelligence is implemented.

[0023] A computer-readable storage medium includes a program, which, when running on a computer, enables the computer to execute the eye behavior analysis method based on artificial intelligence.

[0024] The beneficial effects of the present invention include:

[0025] (1) Through parabola-based shape preprocessing and area-based size preprocessing, the shape and size of the eye image can be adjusted more accurately to adapt to eye images of different shapes, states and sizes, so that each segmented sub-region can concentrate more information in the horizontal direction, and differentiate the amount of information that can be brought by different eye positions, so that the information in the eye image can be distributed to different sub-regions as much as possible; thus, an effective balance is achieved between complexity and not losing important information;

[0026] (2) Using the characteristics of eye images, we construct triple-based UX tuples and their tuple sequences. Based on the existing image segmentation and pixel classification counting, we can use the understanding of the characteristics of eye image composition when expressing, thereby converting the representation of this characteristic into a basis for quantifiable calculation, providing fast and continuous eye behavior judgment, and can recognize and respond to eye behavior in real time, which is suitable for application scenarios that require fast response;

[0027] (3) The unidirectional discrete sequence based on spatial information can sensitively detect the changes in real-time eye behavior in a real-time environment. When it is impossible to obtain effective eye behavior judgments using extensive information, the bidirectional discrete sequence based on temporal information can reduce the fixed information factors caused by environmental factors. By being sensitive to possible complex factors, it can increase the possibility of matching with samples and improve the stability and effectiveness of eye behavior judgments.

Brief Description of the Drawings

[0028] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present application, but do not constitute an improper limitation of the present invention. In the drawings:

[0029] Figure 1 Schematic diagram of the artificial intelligence-based eye behavior analysis method provided by the present invention.

[0030] Figure 2 A schematic diagram of a method for segmenting eye images provided by the present invention. [Specific implementation method]

[0031] The present invention will be described in detail below with reference to the accompanying drawings and specific embodiments, wherein the exemplary embodiments and descriptions are only used to explain the present invention, but are not intended to limit the present invention.

[0032] The present invention proposes an eye behavior analysis method and system based on artificial intelligence, as shown in the attached Figure 1 As shown, the method comprises the following steps:

[0033] Step S1: Acquire an eye image and preprocess it; specifically: perform parabola-based shape preprocessing and area-based size preprocessing;

[0034] As attached Figure 2 As shown, the steps specifically include the following steps:

[0035] Step S11: obtaining a horizontal marker, and obtaining a vertical marker based on the horizontal marker; dividing the eye image into four regions based on the vertical marker and the horizontal marker, and performing a parabola-based shape preprocessing and an area-based size preprocessing on each region;

[0036] Step S12: the obtaining of the horizontal markers is specifically as follows: segmenting the eye image to obtain the eye region; obtaining the positions of the two end points of the longest diameter of the eye region as two horizontal markers (A), constructing an X-axis based on the horizontal markers, and taking the positions of the points with the longest vertical distance from the X-axis as two vertical markers (B); and constructing a Y-axis based on the two vertical markers; that is, the axis of the coordinate axis created here does not necessarily pass through the center of the eye, but is related to the eye behavior at that time; Figure 2 The lower right area of ​​the eye region is shown;

[0037] Step S13: Divide the eye image into four areas based on the X-axis and the Y-axis; connect the horizontal mark and the vertical mark in the same area to form a stretching auxiliary line (TT′); determine the first projection height L of the horizontal mark on the Y-axis; determine the second projection height C of the horizontal mark on the X-axis;

[0038] Step S14: performing parabola-based shape preprocessing on each region; specifically, stretching the region along the first direction or the second direction and performing pixel interpolation until the stretching condition is met to complete the shape preprocessing;

[0039] Preferably: the stretching condition is that any point on the edge of the region coincides with a preset parabola passing through the horizontal mark position and the vertical mark position, or the second projection height C reaches a preset second projection height value after stretching;

[0040] Preferably, the first direction is the direction of the Y axis toward the outside of the eye area. Figure 2 The first direction is the direction of the middle arrow; the second direction is perpendicular to the stretching auxiliary line and toward the outside of the eye area;

[0041] Preferably: the first direction or the second direction is selected for stretching according to the similar orientations of the four regions;

[0042] Step S15: stretching the area along the X-axis toward the outside of the eye area so that the first projection height L reaches a first projection height preset value, thereby completing the size preprocessing based on the area;

[0043] Step S2: starting from the left horizontal mark AL and the right horizontal mark AR of the two horizontal marks A as the division center, U1 left rays and U1 right rays are set in the part between the upper tangent formed by the upper part of the eye area and the lower tangent formed by the lower part of the eye area; U1 left rays and U1 right rays divide the eye image into UX sub-areas; when U1 left rays or U1 right rays act independently, the eye area is divided into U1+1 sub-areas; after the two form an intersection, these areas will be further subdivided to form UX irregular areas; and the number of the irregular areas is related to the shape of the eye area, the preset value of the projection height, and the division method; the irregular division method can enable each sub-area after segmentation to concentrate more information in the horizontal direction, and differentiate and consider the amount of information that can be brought by different eye positions, so that the information in the eye image can be distributed to different sub-areas as much as possible;

[0044] Preferably: the U-bar left-side rays and the U-bar right-side rays are evenly arranged, and the angles between adjacent rays are the same;

[0045] Preferably: the U-bar left-side rays and the U-bar right-side rays are unevenly arranged, and the angles between adjacent rays are different;

[0046] Preferably: the step also includes, starting from the upper horizontal mark BU and the lower horizontal mark BB of the two vertical marks B, respectively, setting U2 upper rays and U2 lower rays in the part between the upper horizontal mark BU and the left tangent of the left part of the eye area and the right tangent of the right part of the eye area; the U2 upper rays and the U2 lower rays divide the eye image into UY sub-areas; of course, it is also possible to start from the upper horizontal mark BU and the lower horizontal mark BB of the two vertical marks B at the same time, and start from the left horizontal mark AL and the right horizontal mark AR of the two horizontal marks A as the division center, so as to form a more complex division method, which is not described in detail; considering that for the eye image, not every area is of the same importance, therefore, uniform grid division cannot obtain an effective balance between complexity and not losing importance, the present invention divides the grid by a radial segmentation method based on a uniform angle, and finally forms UX or UY sub-areas; the following takes UX sub-areas as an example;

[0047] Preferably: U1, U2 are preset values;

[0048] Preferably, the upper rays or the lower rays are spaced at a preset uniform angle; the uniform angle is 0.1-1°;

[0049] Step S3: Ternary each sub-region to obtain a triplet corresponding to each sub-region, and arrange all the triplet according to the distribution position of the sub-region to obtain a UX tuple corresponding to the eye region ((I i，0 , I i，1 , I i，2 )); where: i is the sub-area number; I i，0 is the number of pixels belonging to the sclera region in sub-region i; i，1 is the number of pixels belonging to the iris region in sub-region i; i，2 is the number of pixels belonging to the pupil area in sub-area i; i = 1 ~ UX;

[0050] Preferably, the distribution positions are arranged in order from top to bottom and from left to right;

[0051] Preferably: the distribution position is a fixed distribution position;

[0052] Preferably: an artificial intelligence model is used to segment the eye image into sclera, iris and pupil area categories; in the segmentation process, a classification of each pixel is formed, so that the pixel belongs to the sclera (the white part of the eye, that is, the white area outside the eyeball), the iris (the colored part located around the pupil, which is responsible for controlling the size of the pupil to adjust the amount of light entering the eye) or the pupil (the black circular hole in the center of the iris, through which light enters the eyeball);

[0053] Preferably: the artificial intelligence model is mainly composed of an encoder and a decoder, such as a U-Net network; the eye image is converted into a single-channel grayscale image through a 1×1 convolution and passed to the encoder;

[0054] Alternatively, the step S3 is specifically as follows: binarizing each sub-region to obtain a binary group corresponding to each sub-region, and arranging all the binary groups according to the distribution positions of the sub-regions to obtain a UX group corresponding to the eye image ((I i，0 , I i，1 )); where: i is the sub-area number; I i，0 is the number of pixels belonging to the iris region in sub-region i; i，1 is the number of pixels belonging to the pupil area in sub-region i; converting ternary value into binary value can reduce the computational complexity;

[0055] Step S4: compare the current UX tuple sequence of length N, which is composed of the UX tuples corresponding to the most recent N consecutive human eye regions, with the UX tuple sample sequence in the eye behavior sample library to determine the similarity between the current UX tuple sequence and the UX tuple sample sequence; select the eye behavior corresponding to the UX tuple sample sequence corresponding to the one with the smallest similarity as the determined eye behavior;

[0056] Preferably, a continuous human eye image is obtained from the human eye image video, and a human eye region including only the eye region is obtained based on the human eye image;

[0057] Preferably: pre-constructing UX tuple sample sequences corresponding to different eye behaviors under different viewing angles, environments, and user eye categories to construct an eye behavior sample library;

[0058] Preferably: use Euclidean distance to calculate the similarity between the current UX tuple sequence and the UX tuple sample sequence;

[0059] The step S4 specifically includes the following steps:

[0060] Step S41: The current UX tuple sequence is expressed as The sample sequence of UX tuples is represented as

[0061] Step S42: Calculate the intra-sequence discreteness of the current UX tuple sequence and the UX tuple sample sequence respectively to obtain the current unidirectional discreteness sequence And the sample one-way discreteness sequence of the UX tuple sample sequence Specifically, the current one-way discreteness sequence and the sample one-way discreteness sequence are calculated using the following formulas (1) and (2); where: i, j = 3~UX;

[0062]

[0063]

[0064] Step S43: Entropy normalization is performed on the current one-way discreteness sequence and the sample one-way discreteness sequence; specifically, the current one-way discreteness sequence after entropy normalization is obtained by using the following formula (3) (4):

[0065] and the sample one-way discreteness series in: They are respectively the means of the first, second, and third elements in the triplet in the current one-way discreteness sequence; are the means of the first, second, and third elements of the triplet in the sample one-way discreteness sequence;

[0066]

[0067]

[0068] Step S44: Calculate the current one-way discreteness sequence after entropy value normalization and the sample one-way discreteness series The similarity between them is taken as the similarity between the current UX tuple sequence and the UX tuple sample sequence; Specifically: the similarity sim is calculated using the following formula (5): I，J ;

[0069]

[0070] Alternatively: Use the following formula (6) to calculate the similarity sim I，J ; Where: ε is a minimum value, for example: ε = 0.001;

[0071]

[0072] Preferably, the step S4 further comprises a step S45: selecting the eye behavior corresponding to the UX tuple sample sequence corresponding to the one with the smallest similarity, and repeatedly executing step S4 to make continuous eye behavior judgments; when all similarities are greater than the similarity threshold, proceeding to step S5; otherwise, repeatedly executing step S4 to make continuous eye behavior judgments;

[0073] Alternatively: when all similarities are greater than the similarity threshold and the latest eye behavior judgment moment is reached, step S5 is entered; at this time, when the eye behavior judgment moment is reached and a corresponding judgment must be made, the next step is entered to perform a bidirectional discrete sequence based on time information, reduce the fixed information amount factor caused by environmental factors, and increase the possibility of matching with the sample by reducing the sensitivity to possible environmental complex factors;

[0074] Step S5: Use the N-length current UX tuple sequence and UX tuple sample sequence to construct the current bidirectional discreteness sequence based on time information and the sample bidirectional dispersion sequence The similarity between the two is taken as the similarity between the current UX tuple sequence and the UX tuple sample sequence; the eye behavior corresponding to the UX tuple sample sequence corresponding to the one with the smallest similarity is selected as the determined eye behavior; unlike spatial information, when considering time information, the environmental factor information contained in the time information before and after the recognition of eye behavior is the most complete, so here determining the discreteness sequence based on bidirectional factors can most effectively improve stability;

[0075] The step S5 specifically includes the following steps:

[0076] Step S51: The current UX tuple sequence of length N is expressed as The N-length UX tuple sample sequence is expressed as

[0077] Step S52: Calculate the time discreteness of the current UX tuple sequence and the UX tuple sample sequence respectively to obtain the current bidirectional discreteness sequence and the sample bidirectional discreteness sequence of the UX tuple sample sequence Specifically, the current two-way discreteness sequence and the sample two-way discreteness sequence are calculated using the following formulas (7) and (8); where: i, j = 1 to UX, n = 2 to N-1;

[0078]

[0079]

[0080] Step S53: Entropy normalization is performed on the current two-way discreteness sequence and the sample two-way discreteness sequence; specifically, the current two-way discreteness sequence after entropy normalization is obtained by using the following formula (9) (10): and the sample bidirectional dispersion sequence in:

[0081] They are respectively the means of the first, second, and third elements in the triplet in the current bidirectional discreteness sequence; are the means of the first, second, and third elements in the triplet in the sample bidirectional discreteness sequence;

[0082]

[0083]

[0084]

[0085] Step S54: Calculate the current bidirectional discreteness sequence after entropy value normalization and the sample bidirectional dispersion sequence The similarity between them is taken as the similarity between the current UX tuple sequence and the UX tuple sample sequence; Specifically: the similarity sim is calculated using the following formula (11): I，J ;

[0086]

[0087] Step S55: selecting the eye behavior corresponding to the UX tuple sample sequence corresponding to the one with the smallest similarity as the determined eye behavior; and repeatedly executing steps S4 and / or S5 to perform continuous eye behavior judgments;

[0088] Based on the same inventive concept, the present invention also provides an eye behavior analysis system based on artificial intelligence, and the system is used to implement the above-mentioned eye behavior analysis method based on artificial intelligence;

[0089] Based on the same inventive concept, the present invention also provides an eye behavior analysis server based on artificial intelligence, and the server is used to implement the above-mentioned eye behavior analysis method based on artificial intelligence;

[0090] Based on the same inventive concept, the present invention also provides an eye behavior analysis device based on artificial intelligence, and the device is used to implement the above-mentioned eye behavior analysis method based on artificial intelligence;

[0091] Based on the same inventive concept, the present invention also provides an eye behavior analysis platform based on artificial intelligence, and the platform is used to implement the above-mentioned eye behavior analysis method based on artificial intelligence;

[0092] A computer program (also referred to as a program, software, software application, script, or code) can be written in any form of a programming language, including an assembled or interpreted language, a declarative or procedural language, and it can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, object, or other unit suitable for use in a computing environment. A computer program may, but need not, correspond to a file in a file system. A program can be stored in a portion of a file that holds other programs or data (e.g., one or more scripts stored in a markup language document), in a single file dedicated to the program, or in multiple collaborative files (e.g., files storing one or more modules, subroutines, or code portions). A computer program can be deployed to execute on one computer or on multiple computers located at one site or distributed across multiple sites and interconnected by a communication network.

[0093] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0094] The present invention is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0095] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0096] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0097] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents. Any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. An eye behavior analysis method based on artificial intelligence, It is characterized in that The method comprises: Step S1: Acquire an eye image and preprocess it; specifically: perform parabola-based shape preprocessing and area-based size preprocessing; Step S11: Obtain a horizontal flag, and obtain a vertical flag based on the horizontal flag; The eye image is divided into four regions based on the vertical markers and the horizontal markers, and a parabola-based shape preprocessing and an area-based size preprocessing are performed on each region; Step S12: the obtaining of the horizontal markers is specifically as follows: segmenting the eye image to obtain the eye region; obtaining the positions of the two end points of the longest diameter of the eye region as two horizontal markers, constructing an X-axis based on the horizontal markers, and taking the positions of the points with the longest vertical distance from the X-axis as two vertical markers; and constructing a Y-axis based on the two vertical markers; Step S13: dividing the eye image into four areas based on the X-axis and the Y-axis; connecting the horizontal markers and the vertical markers in the same area to form a stretching auxiliary line; Determine a first projection height L of the horizontal mark on the Y axis; determine a second projection height C of the horizontal mark on the X axis; Step S14: performing parabola-based shape preprocessing on each region; specifically, stretching the region along the first direction or the second direction and performing pixel interpolation until the stretching condition is met to complete the shape preprocessing; Step S15: stretching the area along the X-axis toward the outside of the eye area so that the first projection height L reaches a first projection height preset value, thereby completing the size preprocessing based on the area; Step S2: starting from the left horizontal mark AL and the right horizontal mark AR of the two horizontal marks A as the division center, U1 left rays and U1 right rays are set in the part between the upper tangent formed by the left mark AL and the lower tangent formed by the lower part of the eye area; the U1 left ray and U1 right ray divide the eye image into UX sub-areas; Step S3: Ternary each sub-region to obtain a triplet corresponding to each sub-region, and arrange all the triplet according to the distribution position of the sub-region to obtain a UX tuple corresponding to the eye region ((I i，0 , I i，1 , I i，2 )); where: i is the sub-area number; I i，0 is the number of pixels belonging to the sclera region in sub-region i; i，1 is the number of pixels belonging to the iris region in sub-region i; i，2 is the number of pixels belonging to the pupil area in sub-area i; i = 1 ~ UX; Step S4: Use the N-length current UX tuple sequence composed of the UX tuples corresponding to the most recent N consecutive human eye areas to compare with the UX tuple sample sequence in the eye behavior sample library to determine the similarity between the current UX tuple sequence and the UX tuple sample sequence; select the eye behavior corresponding to the UX tuple sample sequence corresponding to the one with the smallest similarity as the determined eye behavior.

2. The eye behavior analysis method based on artificial intelligence according to claim 1, It is characterized in that The stretching condition is that any point in the edge of the area coincides with a preset parabola passing through the horizontal mark position and the vertical mark position, or the second projection height C reaches a preset value of the second projection height after stretching.

3. The eye behavior analysis method based on artificial intelligence according to claim 2, It is characterized in that The first direction is the direction of the Y axis toward the outside of the eye area.

4. The method for analyzing eye behavior based on artificial intelligence according to claim 3, It is characterized in that The second direction is perpendicular to the stretching auxiliary line and is toward the outside of the eye area.

5. The method for analyzing eye behavior based on artificial intelligence according to claim 4, It is characterized in that The first direction or the second direction is selected for stretching similarly according to the positions of the four regions.

6. The method for analyzing eye behavior based on artificial intelligence according to claim 5, It is characterized in that The U-bar left-side rays and the U-bar right-side rays are evenly arranged.

7. The artificial intelligence-based eye behavior analysis system according to claim 6, It is characterized in that An artificial intelligence model is used to classify each pixel in the eye image so that the pixel belongs to the sclera, iris or pupil.

8. An eye behavior analysis system based on artificial intelligence, It is characterized in that include: The system is used to implement the artificial intelligence-based eye behavior analysis method described in any one of claims 1 to 5.

9. An artificial intelligence-based eye behavior analysis platform, It is characterized in that It includes a processor, the processor is coupled to a memory, the memory stores program instructions, and when the program instructions stored in the memory are executed by the processor, the eye behavior analysis method based on artificial intelligence described in any one of claims 1 to 5 is implemented.

10. A computer-readable storage medium, It is characterized in that It comprises a program which, when running on a computer, enables the computer to execute the eye behavior analysis method based on artificial intelligence as described in any one of claims 1 to 5.

Images