Eyeball movement-based toxic behavior identification method and apparatus, and electronic device

Abstract

The invention provides a toxic behavior identification method and device based on eyeball movement, electronic equipment and a storage medium, and the method comprises the steps: carrying out the eye movement calibration of a target detection person in response to the vision detection and visual field detection of the target detection person meeting preset eye movement detection requirements; collecting eye movement data of the target detection personnel after eye movement calibration; the eye movement data comprises reverse eye movement data and memory eye movement data; and inputting the eye movement data into a preset identification model so as to judge whether the target detection person takes drugs or not based on the preset identification model. Therefore, the efficiency and the accuracy of on-site drug-related detection can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of drug-related crime detection, and in particular to a method, device, electronic device and computer-readable storage medium for identifying drug-related behavior based on eye movement. Background Art

[0002] Drugs are very harmful to society, and drug-related behaviors need to be confirmed through drug testing methods. Common drug testing methods include urine testing, saliva testing, and blood testing, with urine testing being the most common.

[0003] Since urine tests and saliva tests are usually open sampling methods, samples are easily contaminated during the sampling process, resulting in unstable test results and large errors, so they are not suitable for large-scale on-site inspections. Moreover, urine may also cross-react with commonly used legal drugs, which can easily cause false positive results (positive for drug use). In addition, the validity period of urine tests and saliva tests is short, usually about one week or 10 days, and it is easy to fail to detect negatives at the end of the validity period or beyond the validity period (i.e., drug use is not detected). Although blood tests have higher accuracy, drugs diffuse faster in the blood, so the validity period of blood tests is shorter (usually within 48 hours). Moreover, blood tests usually require special testing equipment, which is not suitable for large-scale on-site inspections.

[0004] There is currently a lack of time-saving and efficient on-site testing methods, making it difficult for on-site law enforcement officers to efficiently and accurately screen people who abuse drugs. Summary of the invention

[0005] The purpose of this application is to provide a method, device, electronic device and storage medium for identifying drug-related behavior based on eye movement, which can improve the efficiency and accuracy of on-site drug-related detection.

[0006] To achieve the above objectives:

[0007] In a first aspect, an embodiment of the present invention provides a method for identifying drug-related behaviors based on eye movements, the method comprising:

[0008] In response to the vision test and the visual field test of the target detection person satisfying the preset eye movement detection requirements, performing eye movement calibration on the target detection person;

[0009] Collecting eye movement data of the target detection person after eye movement calibration; the eye movement data includes reverse eye movement data and memory eye movement data;

[0010] The eye movement data is input into a preset identification model to determine whether the target detection person is taking drugs based on the preset identification model.

[0011] In one embodiment, the method further comprises: training the preset identification model; wherein the training process comprises:

[0012] Collect the eye movement data of the experimental test subjects in two tasks: reverse eye movement and memory eye movement;

[0013] The T-test method was used to analyze the significant characteristics of the eye movement parameters and pupil area of ​​the experimental test subjects and normal people, and the significant characteristics were obtained;

[0014] The eye movement data and the corresponding saliency features are input into the preset identification model to train the preset identification model.

[0015] In one embodiment, the use of T-test method to perform significant feature analysis on the eye movement parameters and pupil area of ​​drug addicts and normal people to obtain significant features includes:

[0016] The difference between the eye movement parameters and pupil area of ​​the target person and the normal person is determined according to the following formula:

[0017] t=(x1-x2) / (s*sqrt(1 / n1+1 / n2));

[0018] Among them, x1 and x2 represent the sample mean of the target person and the sample mean of the normal person respectively, s is the pooled standard deviation of the two samples, and n1 and n2 are the sample sizes of the target person and the normal person respectively.

[0019] In one embodiment, the method further comprises:

[0020] Using a Poincare scatter plot to analyze the first Poincare parameter, the second Poincare parameter, and the ratio of the first Poincare parameter to the second Poincare parameter of the pupil area of ​​both eyes on the major axis and the minor axis under the reverse eye movement and the memory eye movement;

[0021] Based on the first Poincare parameter, the second Poincare parameter and the ratio of the first Poincare parameter to the second Poincare parameter, a significant feature analysis is performed to obtain a ROC (Receiver Operating Characteristic Curve) curve of pupil area change characteristics classification in the eye movement task paradigm.

[0022] In a second aspect, an embodiment of the present invention provides a drug-related behavior identification device based on eye movement, comprising:

[0023] A detection module, used to perform vision detection and visual field detection on the target detection person to ensure that the target detection person meets the eye movement detection requirements;

[0024] A calibration module, used for performing eye movement calibration on the target detection person so that the eye tracker can track the eye movement of the target detection person;

[0025] A collection module, used for collecting eye movement data of the target detection personnel; the eye movement data includes reverse eye movement data and memory eye movement data;

[0026] The judgment module is used to input the eye movement data into a preset identification model to judge whether the target detection person is taking drugs based on the preset identification model.

[0027] In a third aspect, an embodiment of the present invention provides an electronic device, including:

[0028] processor;

[0029] a memory for storing instructions executable by the processor;

[0030] Wherein, the processor is configured to execute the instructions to implement the drug-related behavior identification method based on eye movement as described in any one of the first aspects.

[0031] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium. When the instructions in the computer-readable storage medium are executed by a processor, the steps of the method for identifying drug-related behaviors based on eye movements as described in any one of the first aspects are implemented.

[0032] The drug-related behavior identification method, device, electronic device and storage medium provided by the embodiment of the present invention based on eye movement include: in response to the target detection person's vision detection and visual field detection both meeting the preset eye movement detection requirements, the target detection person is calibrated for eye movement; the eye movement data of the target detection person after eye movement calibration is collected; the eye movement data includes reverse eye movement data and memory eye movement data; the eye movement data is input into a preset identification model to determine whether the target detection person is taking drugs based on the preset identification model. In this way, the efficiency and accuracy of on-site drug-related detection can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0034] Figure 1 A schematic diagram of a flow chart of a method for identifying drug-related behaviors based on eye movements provided by an embodiment of the present invention;

[0035] Figure 2 A schematic diagram of the reverse eye movement test results of target detection personnel extracted from a drug rehabilitation center;

[0036] Figure 3 A schematic diagram of the results of the memory eye movement test of the target detection personnel extracted from the drug rehabilitation center;

[0037] Figure 4 A schematic diagram of a module of a drug-related behavior identification device based on eye movement provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0038] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0039] It should be noted that, in this article, the terms "include", "comprises" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the existence of other identical elements in the process, method, article or device including the element. In addition, components, features, and elements with the same name in different embodiments of the present application may have the same meaning or different meanings, and their specific meanings need to be determined by their explanation in the specific embodiment or further combined with the context of the specific embodiment.

[0040] See also Figure 1 , which is a flowchart of a method for identifying drug-related behaviors based on eye movements provided in an embodiment of the present application. The method for identifying drug-related behaviors based on eye movements can be executed by a device for identifying drug-related behaviors based on eye movements provided in an embodiment of the present application. The device can be implemented in software and / or hardware, such as a server, a computer, a processor, or other electronic device. In this embodiment, the execution subject of the method for identifying drug-related behaviors based on eye movements is the processor of a vehicle-mounted terminal as an example. The method for identifying drug-related behaviors based on eye movements provided in this embodiment includes:

[0041] Step S1, in response to the vision detection and visual field detection of the target detection person satisfying the preset eye movement detection requirements, performing eye movement calibration on the target detection person.

[0042] Understandably, if the eyesight and visual field of the target detection personnel do not meet the standards for drug identification, it will affect the identification results. Therefore, the first step in drug identification should be to ensure that the eyesight and visual field of the target detection personnel meet the standards, otherwise the identification results may be incorrect.

[0043] Step S2, collecting the eye movement data of the target detection person after eye movement calibration; the eye movement data includes reverse eye movement data and memory eye movement data.

[0044] Understandably, drugs affect the central nervous system and visual system. Drug use may interfere with the memory process, leading to difficulties in performing memory eye movements, such as errors in eye movement trajectories or prolonged reaction times.

[0045] The embodiment of the present invention compares the eye movement characteristics of drug users and non-drug users, thereby determining whether the detected person has taken drugs through eye movement.

[0046] Specifically, target persons were extracted from compulsory isolation drug rehabilitation centers and detention centers to collect data on two eye movement tasks: reverse eye movement and memory eye movement. The subsequent research in the embodiments of the present invention can be understood as being conducted based on the collected data.

[0047] It should be noted that when collecting eye movement data, the target person’s vision and visual field are tested to ensure that the drug user meets the eye movement test requirements, and eye movement calibration is performed to ensure that the eye tracker accurately tracks eye movements.

[0048] Step S3, inputting the eye movement data into a preset identification model to determine whether the target detection person is taking drugs based on the preset identification model.

[0049] In one embodiment, the preset identification model is trained; wherein the training process includes:

[0050] The eye movement data of the target detection personnel in the two tasks of reverse eye movement and memory eye movement are collected; the eye movement parameters and pupil area of ​​the target personnel and normal people are analyzed for significant features using the T-test method to obtain significant features; the eye movement data and the corresponding significant features are input into the preset identification model to train the preset identification model.

[0051] Understandably, by recording different types of eye movement tasks, a variety of eye movement features can be captured, which can reflect the visual processing and cognitive functions of individuals. Statistical analysis methods are used to determine which eye movement parameters and pupil areas have statistically significant differences between target persons (e.g., drug addicts) and normal persons. T-tests can help identify differences in eye movement features between drug addicts and normal persons, which are considered "significant features" and are essential for distinguishing different groups. Through classification analysis methods (e.g., machine learning algorithms), previously collected eye movement data are used to distinguish different groups (e.g., drug addicts and normal persons). Classification analysis can reveal which eye movement features are most effective in distinguishing different groups and generate a classification model that can classify individuals based on eye movement data. The information obtained from the significant feature analysis and classification analysis is used to train the preset identification model. In this way, the model learns how to identify and distinguish different groups based on eye movement data. During the training process, the model adjusts internal parameters to better perform classification tasks, thereby improving accuracy and reliability in subsequent identification work.

[0052] In one embodiment, the use of the T-test method to perform a significant feature analysis on the eye movement parameters and pupil area of ​​drug addicts and normal people to obtain significant features includes: determining the difference between the eye movement parameters and pupil area of ​​the target person and the normal person according to the following formula:

[0053] t=(x1-x2) / (s*sqrt(1 / n1+1 / n2));

[0054] Among them, x1 and x2 represent the sample mean of the target person and the sample mean of the normal person respectively, s is the pooled standard deviation of the two samples, and n1 and n2 are the sample sizes of the target person and the normal person respectively.

[0055] Furthermore, the pooling standard deviation is expressed as:

[0056] s=sqrt(((n1-1)*s1 2 +(n2-1)*s2^2) / (n1+n2-2)); where s1 and s2 are the standard deviations of the two samples.

[0057] In this embodiment, it is found through analysis that in the results of significance analysis using the T-test method, there are significant differences in the accuracy, spatial error (horizontal), first-step saccade amplitude, first-step saccade amplitude standard deviation, average speed, blink rate, landing point error (horizontal), reflex saccade, reflex saccade amplitude gain, average number of reflex saccade steps, corrected saccade amplitude gain, corrected saccade spatial error, corrected saccade latency, and corrected saccade landing point error of the eye movement data of the target person and the normal person.

[0058] In one embodiment, the method further includes: using a Poincare scatter plot to analyze the first Poincare parameter, the second Poincare parameter, and the ratio of the first Poincare parameter to the second Poincare parameter of the pupil area of ​​both eyes in the major axis and the minor axis under the reverse eye movement and memory eye movement; performing a significant feature analysis based on the first Poincare parameter, the second Poincare parameter, and the ratio of the first Poincare parameter to the second Poincare parameter to obtain a ROC curve for the change feature classification of the pupil area in the eye movement task paradigm.

[0059] Understandably, the Poincare scatter plot can reveal the nonlinear dynamic characteristics of pupil area changes, and display the structure of time series data through a two-dimensional graph, which helps to identify dynamic behaviors such as periodicity and chaos in the data. The first Poincare parameter (SD1) and the second Poincare parameter (SD2) of the pupil area of ​​both eyes under the long axis and the short axis are analyzed. SD1 and SD2 quantify the rapid and slow fluctuations of pupil area changes, respectively, and provide quantitative information on pupil dynamic characteristics. The SD1 / SD2 ratio provides frequency distribution information of pupil dynamic changes, reflecting the complexity of pupil reactions and the balance state of autonomic nervous system activity. The ROC curve is a tool for evaluating the performance of classification models, which quantifies the distinguishing ability of the model through the area under the curve (AUC). It shows the relationship between the true positive rate (TPR) and the false positive rate (FPR) of the model under different threshold settings, helps determine the optimal classification threshold and provides an evaluation of the overall accuracy of the model. Through the above steps, a classification system based on eye movement data can be developed to classify different eye movement tasks and improve the accuracy of drug-related behavior recognition.

[0060] In one embodiment, after calculating the first Poincare parameter and the second Poincare parameter, the method further includes: calculating the average value of the first Poincare parameter and the second Poincare parameter, comparing the average value with a preset standard mean value to obtain a comparison result; and determining the difference between the pupil area of ​​the target detection person and the pupil area of ​​a normal person based on the comparison result.

[0061] Understandably, there is no clear range for the difference in accuracy and spatial error between drug addicts and normal people in eye movement data, but the first and second Poincare parameters can be averaged to obtain the specific average difference. The pupil area is represented by the first and second Poincare parameters above, and the difference between the two can be obtained based on the meaning of the first and second Poincare parameters.

[0062] In addition, the drug-related behavior identification method based on eye movement provided in the present application can also obtain screening results by using the Relief feature selection method to filter the eye movement speed and landing point error contained in the eye movement data of the target detection personnel respectively; and classify and analyze the screening results through a decision tree classifier.

[0063] Furthermore, the relief feature selection method is used to perform feature analysis on the eye movement speed and landing point error of the target detection personnel, including:

[0064] Filter according to the following formula:

[0065] Among them, diff(A,R1,R2) represents the difference between sample R1 and sample R2 on feature A, M j (C) represents the jth nearest neighbor sample in class C.

[0066]

[0067] Specifically, see Figure 2 and Figure 3 , which are respectively schematic diagrams of the test results of reverse eye movement and memory eye movement of the target detection personnel extracted from the drug rehabilitation center in this application. Figure 2 As shown in the figure, by extracting 163 sets of eye movement features and using the T-test method for significance analysis, it was found that the accuracy, latency, average speed, maximum speed, multi-step saccade incidence, average number of saccade steps, reflex saccade rate, reflex saccade amplitude gain, average number of reflex saccade steps, and reflex saccade latency were significantly different. Figure 3 As shown in the figure, 119 groups of eye movement features were extracted and analyzed by T-test method. It was found that there were significant differences in spatial error, average speed, maximum speed, multi-step saccade incidence, final saccade landing point error, Break, and average number of saccade steps.

[0068] From the analysis of the two main eye movement task paradigms, it was found that the average speed and maximum speed under the two task modes had significant differences, and the drug addicts were greater than the normal control group. The reflex saccade parameters (incidence, amplitude gain, average number of steps, latency) in the reverse eye movement task had outstanding characteristics, which may play a key role in distinguishing drug addicts from healthy people. The break characteristics in memory saccades were the most significantly different, and the drug addicts were significantly higher than the normal control group. At the same time, the significant features under the two eye movement task paradigms had certain differences, indicating that the two task paradigms can reveal the eye movement control characteristics of drug addicts in different eye movement feature dimensions. The fusion analysis of the eye movement features under the two eye movement task paradigms may be helpful to improve the detection success rate of drug addicts.

[0069] The drug-related behavior identification method based on eye movement provided in this embodiment can be specifically summarized as follows: 1. Perform vision and visual field tests on the target person (drug addict) to ensure that the target person (drug addict) meets the eye movement detection requirements; 2. Perform eye movement calibration to ensure that the eye tracker accurately tracks eye movements; 3. Perform reverse eye movement and memory eye movement task paradigm tests in sequence to collect eye movement data; 4. Input the collected data into the diagnostic model to identify whether the person is a drug addict.

[0070] Based on the same inventive concept as the above embodiments, the present application provides a drug-related behavior identification device based on eye movement, such as Figure 4 As shown, the drug-related behavior identification device based on eye movement includes:

[0071] A calibration module, used for performing eye movement calibration on the target detection person when the vision detection and visual field detection of the target detection person both meet the preset eye movement detection requirements;

[0072] A collection module, used for collecting eye movement data of the target detection personnel after eye movement calibration; the eye movement data includes reverse eye movement data and memory eye movement data;

[0073] The judgment module is used to input the eye movement data into a preset identification model to judge whether the target detection person is taking drugs based on the preset identification model.

[0074] For the specific limitations of the drug-related behavior identification device based on eye movements, please refer to the limitations of the drug-related behavior identification method based on eye movements above, which will not be repeated here. Each module in the above-mentioned drug-related behavior identification device based on eye movements can be implemented in whole or in part through software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor of the electronic device in the form of hardware, or can be stored in the memory of the electronic device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0075] Based on the same inventive concept as the aforementioned embodiments, an embodiment of the present invention further provides an electronic device, comprising: a processor; a memory for storing instructions executable by the processor; wherein the processor is configured to execute the instructions to implement a method for identifying drug-related behaviors based on eye movements as described in any one of the aforementioned embodiments.

[0076] Based on the same inventive concept as the aforementioned embodiments, an embodiment of the present invention further provides a computer-readable storage medium. When the instructions in the computer-readable storage medium are executed by a processor, the steps of the method for identifying drug-related behaviors based on eye movements as described in any of the aforementioned embodiments are implemented.

[0077] It should be noted that the various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0078] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the attached claims.

Claims

1. A method for identifying drug-related behavior based on eye movement, characterized in that: include: In response to the vision test and the visual field test of the target detection person satisfying the preset eye movement detection requirements, performing eye movement calibration on the target detection person; Collecting eye movement data of the target detection person after eye movement calibration; the eye movement data includes reverse eye movement data and memory eye movement data; The eye movement data is input into a preset identification model to determine whether the target detection person is taking drugs based on the preset identification model.

2. The method according to claim 1, characterized in that Also includes: The preset identification model is trained; wherein the training process includes: Collect the eye movement data of the experimental test subjects in two tasks: reverse eye movement and memory eye movement; The T-test method was used to analyze the significant characteristics of the eye movement parameters and pupil area of ​​the experimental test subjects and normal people, and the significant characteristics were obtained; The eye movement data and the corresponding saliency features are input into the preset identification model to train the preset identification model.

3. The method according to claim 2, characterized in that The T-test method is used to perform significant feature analysis on the eye movement parameters and pupil area of ​​drug addicts and normal people to obtain significant features, including: The difference between the eye movement parameters and pupil area of ​​the target person and the normal person is determined according to the following formula: t=(x1-x2) / (s*sqrt(1 / n1+1 / n2)); Among them, x1 and x2 represent the sample mean of the target person and the sample mean of the normal person respectively, s is the pooled standard deviation of the two samples, and n1 and n2 are the sample sizes of the target person and the normal person respectively.

4. The method according to claim 2, characterized in that: The method further comprises: Using a Poincare scatter plot to analyze the first Poincare parameter, the second Poincare parameter, and the ratio of the first Poincare parameter to the second Poincare parameter of the pupil area of ​​both eyes on the major axis and the minor axis under the reverse eye movement and the memory eye movement; Based on the first Poincare parameter, the second Poincare parameter and the ratio of the first Poincare parameter to the second Poincare parameter, a significant feature analysis is performed to obtain a ROC curve of the pupil area change feature classification in the eye movement task paradigm.

5. The method according to claim 4, characterized in that After calculating the first Poincare parameter and the second Poincare parameter, the method further includes: Calculating an average value of the first Poincare parameter and the second Poincare parameter, and comparing the average value with a preset standard mean value to obtain a comparison result; The difference between the pupil area of ​​the target person and the pupil area of ​​a normal person is determined based on the comparison result.

6. A drug-related behavior identification device based on eye movement, characterized in that: include: A calibration module, used for performing eye movement calibration on the target detection person when the vision detection and visual field detection of the target detection person both meet the preset eye movement detection requirements; A collection module, used for collecting eye movement data of the target detection personnel after eye movement calibration; the eye movement data includes reverse eye movement data and memory eye movement data; The judgment module is used to input the eye movement data into a preset identification model to judge whether the target detection person is taking drugs based on the preset identification model.

7. An electronic device, characterized in that: include: processor; a memory for storing instructions executable by the processor; Wherein, the processor is configured to execute the instructions to implement the drug-related behavior identification method based on eye movement as described in any one of claims 1-5.

8. A computer-readable storage medium, characterized in that: When the instructions in the computer-readable storage medium are executed by a processor, the steps of the drug-related behavior identification method based on eye movement as described in any one of claims 1 to 5 are implemented.

Images