Method and system for assessing abnormal behavior in children

Abstract

The embodiment of the application provides a kind of child behavior abnormality evaluation method and system, in the case where the child to be measured is located at least one test scene shown in test room, the behavior test data of the child to be measured collected by the data acquisition equipment arranged in test room is acquired;The behavior test data of the child to be measured is analyzed and processed, and the behavior characteristics of the child to be measured are determined;According to behavior characteristics, the probability that the child to be measured exists behavior abnormality is evaluated.The test scene corresponds to the abnormal behavior type to be evaluated of the child to be measured, and the abnormal behavior type includes at least one performance behavior corresponding to autism spectrum disorder.That is, the behavior test data of the child to be measured in real scene is collected by data acquisition equipment, to record the real action and behavior of the child to be measured comprehensively and accurately.In this way, the possibility of abnormal behavior of the child to be measured can be evaluated by analyzing behavior test data, which shortens the test time and improves the evaluation efficiency of abnormal behavior.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and in particular, to a child behavior abnormality evaluation method and system. BACKGROUND

[0002] Behavioral abnormalities, also known as behavioral disorders, refer to abnormalities in actions and behaviors caused by neurological developmental disorders in aspects of cognition, emotion, and will. For example, autism spectrum disorder (ASD) includes a group of neurological developmental disorders, which refers to a broad developmental disorder that occurs in children during the growth process from birth to early childhood. The behavioral abnormalities of patients with ASD mainly manifest in social interaction disorders, communication difficulties, narrow interest range, and repetitive stereotyped behaviors.

[0003] In related technologies, when evaluating whether the behavior of a child under test is abnormal, a professional usually needs to test various behavioral abilities of the child under test in sequence according to the test rules corresponding to each mental developmental disorder, so as to determine whether the actions and behaviors of the child under test are abnormal according to the test results. For example, when evaluating whether a child under test has autism spectrum disorder, a professional doctor communicates with the child under test based on the test items listed in the Autism Diagnostic Observation Schedule (ADOS), Autism Diagnostic Interview Revised (ADIR), Childhood Autism Rating Scale (CARS), and other scales, observes the reactions and actions of the child under test during the communication process, and evaluates the social interaction, language communication, and response behavior of the child under test, so as to determine the risk of the child under test having autism spectrum disorder.

[0004] However, in the above-mentioned behavioral abnormality screening method, the evaluation of behavioral abnormalities needs to rely on professionals, and it takes several hours to complete the evaluation of abnormal behaviors corresponding to each neurological developmental disorder, which leads to low screening efficiency of various neurological developmental disorders and difficulty in wide popularization. SUMMARY

[0005] The present application provides a child behavior abnormality evaluation method and system, which can reduce the professional requirements of evaluators in the process of evaluating behavioral abnormalities, avoid the requirements of the instruction compliance and language ability of the child under test, and is also applicable to young children, shortens the test time required for evaluating behavioral abnormalities, and thus improves the screening efficiency of neurological developmental disorders related to behavioral abnormalities.

[0006] In a first aspect, the present application provides a method for evaluating abnormal behavior of a child, comprising:

[0007] In a case where the child to be tested is located in a test room to display at least one test scene, the data acquisition device arranged in the test room is used to acquire behavior test data of the child to be tested; the test scene corresponds to an abnormal behavior type to be evaluated by the child to be tested, and the abnormal behavior type includes at least one performance behavior corresponding to autism spectrum disorder;

[0008] The behavior test data of the child to be tested is analyzed and processed to determine a behavior feature of the child to be tested.

[0009] According to the behavior feature, a probability of abnormal behavior of the child to be tested is evaluated.

[0010] In a second aspect, the present application further provides a system for evaluating abnormal behavior of a child, comprising a master control room and a test room, the master control room comprising a data processing device, and the test room being arranged with a data acquisition device;

[0011] In a case where the child to be tested is located in a test room to display at least one test scene, the data acquisition device is used to acquire behavior test data of the child to be tested, and send the behavior test data to the data processing device; the test scene corresponds to an abnormal behavior type to be evaluated by the child to be tested, and the abnormal behavior type includes at least one performance behavior corresponding to autism spectrum disorder;

[0012] The data processing device is used to analyze and process the behavior test data of the child to be tested, determine a behavior feature of the child to be tested, and evaluate a probability of abnormal behavior of the child to be tested according to the behavior feature.

[0013] In a third aspect, the present application further provides a device for evaluating abnormal behavior of a child, comprising:

[0014] The data acquisition module is used to acquire behavior test data of the child to be tested collected by the data acquisition device arranged in the test room in a case where the child to be tested is located in a test room to display at least one test scene; the test scene corresponds to an abnormal behavior type to be evaluated by the child to be tested, and the abnormal behavior type includes at least one performance behavior corresponding to autism spectrum disorder;

[0015] The data analysis module is used to analyze and process the behavior test data of the child to be tested, and determine a behavior feature of the child to be tested.

[0016] The behavior evaluation module is used to evaluate a probability of abnormal behavior of the child to be tested according to the behavior feature.

[0017] In a fourth aspect, the present application further provides a computer device comprising a memory and a processor, the memory storing a computer program, and the processor implementing steps of any method in the first aspect when executing the computer program.

[0018] In a fifth aspect, the present application also provides a computer readable storage medium, having stored thereon a computer program, which when executed by a processor implements the steps of any method of the first aspect.

[0019] In a sixth aspect, the present application also provides a computer program product comprising a computer program which when executed by a processor implements the steps of any method of the first aspect.

[0020] The technical solutions provided by the present application can achieve at least the following beneficial effects:

[0021] The child behavior abnormality evaluation method and system provided by the present application can obtain behavior test data of the child to be tested collected by a data collection device arranged in a test room, in the case that the child to be tested is located in at least one test scene displayed in the test room; analyze and process the behavior test data of the child to be tested to determine behavior characteristics of the child to be tested; and evaluate a probability of the child to be tested having a behavior abnormality according to the behavior characteristics. The test scene corresponds to an abnormal behavior type to be evaluated of the child to be tested, and the abnormal behavior type includes at least one performance behavior corresponding to autism spectrum disorder. As can be seen, the behavior test data of the child to be tested in a real scene is collected by the data collection device, and the behavior test data can comprehensively and accurately record real actions and behaviors of the child to be tested. Thus, the behavior characteristics of the child to be tested can be determined by analyzing the behavior test data of the child to be tested, so as to evaluate the possibility of the child to be tested having an abnormal behavior. Moreover, in the process of evaluating the behavior abnormality, only a real test scene corresponding to the abnormal behavior type needs to be set in advance, and the data collection device needs to be deployed in the test scene. The test process does not need a professional physician to guide the child to be tested to perform a plurality of test items according to test rules, and does not depend on the experience of the professional physician to determine a test result, thereby shortening the test time, improving the evaluation efficiency of the abnormal behavior, and thus improving the screening efficiency of the neurodevelopmental disorder related to the behavior abnormality. BRIEF DESCRIPTION OF DRAWINGS

[0022] Figure 1 FIG. 1 is a schematic diagram of a child behavior abnormality evaluation system according to an example embodiment of the present application;

[0023] Figure 2 FIG. 2 is a flowchart of a child behavior abnormality evaluation method according to an example embodiment of the present application;

[0024] Figure 3 FIG. 3 is a schematic diagram of a preference test scene according to an example embodiment of the present application;

[0025] Figure 4 FIG. 4 is a schematic diagram of a joint attention test scene according to an example embodiment of the present application;

[0026] Figure 5 is a schematic diagram of a sound response test scene according to an example embodiment of the present application;

[0027] Figure 6 is a structural schematic diagram of a behavior abnormality evaluation device according to an example embodiment of the present application;

[0028] Figure 7 is a structural schematic diagram of a computer device according to an example embodiment of the present application. DETAILED DESCRIPTION

[0029] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application is further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application.

[0030] At present, in the early screening of neurodevelopmental disorders in children, subjective evaluation is mainly used, that is, a professional doctor communicates with the tested child, observes the behavior of the tested child, and combines the description of the daily behavior of the tested child by the guardian to evaluate whether the tested child has behavior abnormalities, and then analyzes the possibility of the tested child having neurodevelopmental disorders.

[0031] In the above process of evaluating the behavior abnormalities of children, due to the lack of objective and quantitative behavior indicators, the risk of misdiagnosis of neurodevelopmental disorders is increased; and the evaluation result of behavior abnormalities depends on the experience of the doctor, which requires a higher professional degree of the doctor.

[0032] Based on this, the present application provides a method and system for evaluating behavior abnormalities of children, which does not require invasive examination of the tested child, does not require intelligence and high-level cognitive ability of the tested child, and does not require the tested child to have certain language ability. By obtaining the real behavior test data of the tested child and the tester in the simple interaction process in the test scene, the behavior characteristics of the tested child can be analyzed, and the probability of the existence of behavior abnormalities can be evaluated. In addition, for the tester, he / she only needs to master the related operation of the child behavior abnormality evaluation system to complete the behavior abnormality evaluation work, which greatly improves the efficiency of behavior abnormality evaluation. In this way, by evaluating the behavior abnormalities of the tested child, auxiliary information is provided for diagnosing / screening neurodevelopmental disorders related to behavior abnormalities, thereby improving the diagnosis efficiency and accuracy of neurodevelopmental disorders. In this way, early detection, early intervention and treatment can be realized, and the adverse effects of neurodevelopmental disorders on children can be reduced.

[0033] In an example embodiment, as shown in FIG. 1, a sound response test scene 100 is provided, which includes a test personnel 110 and a tested child 120. Figure 1As shown, the application provides a child behavior abnormality evaluation system, which comprises a master control room and a test room, the master control room comprises a data processing device 110, and the test room is provided with a data acquisition device 120.

[0034] Specifically, in the case that the child to be tested is located in the test room to show at least one test scene, the data acquisition device 120 is configured to collect behavior test data of the child to be tested and send the behavior test data to the data processing device 110; the data processing device 110 is configured to analyze and process the behavior test data of the child to be tested, determine the behavior characteristics of the child to be tested, and evaluate the probability of the child to be tested having behavior abnormality according to the behavior characteristics.

[0035] The test scene corresponds to an abnormal behavior type to be evaluated by the child to be tested, and the abnormal behavior type includes at least one performance behavior corresponding to autism spectrum disorder.

[0036] Optionally, in order to reduce other interference factors in the test room and enable the child to be tested to move freely in the test room and show real behavior and action, the master control room and the test room can be independently arranged and separated to form independent space areas. In this way, when the child to be tested is evaluated for behavior abnormality, the evaluator controls the test process in the master control room, and the child to be tested and the test personnel interact in the test room, so as to avoid the influence of the evaluator on the actual test behavior of the child to be tested.

[0037] The test room can be one or multiple, and the number of test rooms is not limited in the embodiments of the application. In a specific implementation, multiple test rooms can be arranged, one test scene corresponding to one abnormal behavior type is arranged in each test room; one test room can be arranged, and multiple test scenes corresponding to multiple abnormal behavior types can be arranged in the test room according to test requirements; or different test scenes can be arranged in different areas of one test room when the space of the test room is large enough.

[0038] The evaluator in the master control room can be a doctor or other staff who can control the data processing device. According to test requirements, one test personnel or multiple test personnel can enter the test room to interact with the child to be tested and complete the test process.

[0039] When the data acquisition device is deployed in the test room, it can be pre-deployed in a fixed position area, or the deployment position of the device can be adjusted in real time according to test requirements. In order to be able to comprehensively record all behavior test data of the child to be tested in the test room, the number of data acquisition devices and the deployment position of the data acquisition devices in the test room are not limited in the embodiments.

[0040] In some embodiments, the data acquisition device in this application includes an image acquisition device and a sound acquisition device. The image acquisition device may be, but is not limited to, an RGB camera, an RGB-D camera, a depth camera, a dual-mode camera (e.g., an infrared camera + a visible light camera), an eye tracker, etc., and the sound acquisition device may include a microphone array deployed in a testing room, and / or wearable microphones worn by the child being tested and the testing personnel.

[0041] In some embodiments, the data processing device includes a controller, a processor, and a display. The controller controls the activation, operation, and deactivation of the data acquisition equipment in the testing room, as well as the status of the testing equipment within the testing room. The processor receives behavioral test data sent by the data acquisition equipment, analyzes and processes the behavioral test data, determines the behavioral characteristics of the tested child, assesses the probability of the tested child's behavioral abnormalities, and assesses the risk of the tested child having neurodevelopmental disorders related to behavioral abnormalities. The display shows information such as behavioral test data and assessment results.

[0042] As an example, the data processing device can be any computer device, such as a terminal or a server. A terminal can be, but is not limited to, various personal computers, laptops, smartphones, tablets, portable wearable devices, etc.; a server can be, but is not limited to, at least one standalone server, distributed server, cloud server, server cluster, etc.

[0043] The method for assessing abnormal behavior in children provided in this application can be executed by any computer device or a behavior assessment device. The device can be implemented by software, hardware or a combination of software and hardware as part or all of the processor in the computer device.

[0044] Based on the aforementioned child behavior abnormality assessment system, the behavioral abnormalities of patients with autism spectrum disorder are mainly manifested as social interaction impairment, communication difficulties, narrow range of interests, and repetitive stereotyped behaviors. Therefore, when using behavioral abnormality assessment to assist in screening whether the tested child has autism spectrum disorder, at least one test scenario can be set up in the testing room according to the abnormal behavior type corresponding to autism spectrum disorder, so as to analyze the behavioral characteristics of the tested child in the test scenario.

[0045] In some embodiments, to facilitate autism spectrum disorder screening, the testing scenarios arranged in the testing room may include at least one of the following: preference testing scenario, shared attention testing scenario, and vocal response testing scenario.

[0046] The preference scenario includes a tester and at least one first test prop to attract the child's attention; the joint attention test scenario includes a tester and multiple second test props placed in front of the tester to guide the child to look at the same object; the sound response test scenario includes at least one controllable sound-emitting object.

[0047] It should be noted that the first and second test props can be arranged according to the actual preferences of the children being tested, and the types of test props are not limited in this application embodiment.

[0048] As an example, the first test prop could be a balloon, the second test prop could be a doll, and the controllable sound-producing object could be a bell. It should be understood that the test props and the controllable sound-producing object in the sound response test can also be other objects.

[0049] Furthermore, different testing scenarios are used to test different types of abnormal behaviors in the children being tested. Therefore, the behavioral test data of the children being tested needs to be obtained in different testing scenarios. The corresponding data acquisition devices and behavioral test data for the three testing scenarios mentioned above are as follows:

[0050] (1) The test scenario is a preference test scenario;

[0051] In this test scenario, the data acquisition equipment includes a first image acquisition device, a second image acquisition device, and a first sound acquisition device, all set up in the test scenario.

[0052] Specifically, during the process of the child being tested being positioned opposite the tester and the tester engaging in face-to-face communication with the child, the first image acquisition device is used to acquire the first frontal video of the child being tested, the second image acquisition device is used to acquire the first interactive video of the tester and the child being tested, and the first audio acquisition device is used to acquire the first interactive audio of the tester and the child being tested.

[0053] Correspondingly, the behavioral test data includes a first positive video, a first interactive video, and a first interactive audio.

[0054] (2) The test scenario is a common interest test scenario;

[0055] In this test scenario, the data acquisition devices include a third image acquisition device, a fourth image acquisition device, and a second sound acquisition device, all set up in the test scenario.

[0056] Specifically, while the child being tested is positioned opposite the tester, and the tester guides the child to look at the test prop, a third image acquisition device is used to acquire a second frontal video of the child being tested, a fourth image acquisition device is used to acquire a third frontal video of the tester, and a second sound acquisition device is used to acquire a second interactive audio between the tester and the child being tested.

[0057] Correspondingly, the behavioral test data includes a second frontal video, a third frontal video, and a second interactive audio.

[0058] (3) The test scenario is a sound response test scenario;

[0059] In this test scenario, the data acquisition devices include a fifth image acquisition device and a third sound acquisition device set up in the test scenario.

[0060] Specifically, after the child being tested is within a preset range of a controllable sound-emitting object, and the controllable sound-emitting object emits a sound, the fifth image acquisition device is used to acquire the video of the child's reaction, and the third sound acquisition device is used to acquire the audio of the child's reaction.

[0061] Correspondingly, the behavioral test data includes reaction videos and reaction audios.

[0062] It should be understood that the terms "first" and "second" mentioned above are only used to distinguish data acquisition devices in different testing scenarios, and are not intended to restrict the type, location, or order of data acquisition devices.

[0063] In this embodiment, the child behavior abnormality assessment system includes a control room and a testing room. Firstly, the control room and testing room are independent of each other, reducing interference from other personnel during the testing process. Secondly, for screening different neurodevelopmental disorders, at least one testing scenario can be set up in the testing room to assess the child's behavior abnormality, based on the corresponding abnormal behavior type. This realistic testing scenario allows the child to behave more naturally within it; moreover, the testing scenario can be adjusted in real-time according to the type of abnormal behavior being tested, enabling the assessment of more abnormal behaviors. Furthermore, when assessing the child's behavior abnormality in the testing scenario, behavioral test data can be acquired through data acquisition devices installed in the testing room. This data is more complete, effective, and accurate, and is more helpful in analyzing the child's behavioral characteristics.

[0064] Based on the above-described child behavioral abnormality assessment system, in one exemplary embodiment, such as Figure 2 As shown, this application also provides a method for assessing abnormal behavior in children, which is applied to... Figure 1 The data processing device 110 shown includes the following steps:

[0065] Step 210: With the child being tested in at least one test scenario displayed in the test room, acquire the behavioral test data of the child being tested collected by the data acquisition device set up in the test room.

[0066] The test scenarios correspond to the types of abnormal behaviors that the children being tested are to be assessed for, and these abnormal behavior types include at least one type of behavior associated with autism spectrum disorder.

[0067] In some embodiments, the behavioral test data of the tested child includes all the child's action and behavior data in the test scenario. For example, the child's speech expression information, facial feature information, generalized eye movement information, limb deviation information, and walking position information can be recorded in the form of video, audio, images, etc.

[0068] As an example, the assessment of behavioral abnormalities in children with autism spectrum disorder may include at least one of the following test scenarios: preference test scenario, shared attention test scenario, and vocal response test scenario.

[0069] It should be noted that the specific details regarding the test scenarios, the data acquisition devices deployed in each test scenario, and the behavioral test data of the children to be tested to be acquired in each test scenario can be found in the descriptions in the above system embodiments, and will not be repeated here.

[0070] It should be understood that, depending on the testing requirements, test scenarios corresponding to the abnormal behavior types exhibited in other neurodevelopmental disorders can also be arranged in the testing room. This application embodiment does not limit the number of test scenarios or the specific arrangement area of ​​the test scenarios in the testing room, and can be adjusted according to actual testing needs.

[0071] Step 220: Analyze and process the behavioral test data of the tested children to determine their behavioral characteristics.

[0072] As machine learning technology is increasingly used in the diagnosis and risk prediction of systemic diseases, machine learning algorithms can be used to analyze and process large amounts of behavioral test data of children being tested when analyzing their behavioral characteristics. This enables the machine to have a certain degree of inductive and summarizing ability to assist in the assessment of behavioral abnormalities, thereby shortening the screening time for neurodevelopmental disorders related to behavioral abnormalities and reducing the misdiagnosis rate.

[0073] Specifically, the trained neural network model is used to process the behavioral test data of the tested children through feature extraction, data fusion, and action prediction to determine the behavioral characteristics of the tested children.

[0074] For autism spectrum disorder, the behavioral characteristics to be assessed may include attentional preference characteristics, shared attention characteristics, and vocal response characteristics. These different behavioral characteristics are determined based on behavioral test data from different testing scenarios.

[0075] Next, the process of analyzing the behavioral characteristics of the children under test in different testing scenarios will be explained.

[0076] (1) If the test scenario is a preference test scenario, then the behavioral characteristics are attention preference characteristics;

[0077] In one possible implementation, step 220 can be performed as follows: frame parsing of the first interactive audio and / or the first interactive video; after detecting the first test start command in the first interactive audio, and / or after detecting the tester's first test start action in the first interactive video, acquiring the first human key point data of the child under test contained in each frame of the first frontal video; analyzing the limb deviation movement of the child under test and the duration of the movement corresponding to the limb deviation movement based on the first human key point data corresponding to each frame; and evaluating the attentional preference characteristics of the child under test based on the limb deviation movement and the duration of the movement.

[0078] The first test start instruction can be the first sentence spoken by the tester in an active verbal exchange with the child being tested, or it can be other fixed starting statements. The first test start action can be any action by the tester, such as raising their hand, waving their hand, or turning around; this embodiment of the application does not impose any restrictions on this.

[0079] In some embodiments, after acquiring the first interactive audio, the first interactive audio is detected frame by frame to identify the speech text corresponding to each frame of audio. Then, based on the sampling timestamp of the audio frame corresponding to the first test start command as the starting point, the starting frame image corresponding to the sampling timestamp is determined in the first frontal video. Then, the first human key point data of the tested child is extracted from each frame image after the starting frame image in the first frontal video to analyze the attentional preference characteristics of the tested child.

[0080] In some embodiments, after acquiring the first interactive video, the first interactive video is detected frame by frame to identify the behavior of the tester in each interactive image. Then, based on the sampling timestamp of the interactive image corresponding to the first test start action as the starting point, the starting frame image corresponding to the sampling timestamp is determined in the first frontal video. Then, the first human key point data of the tested child is extracted from each frame image after the starting frame image in the first frontal video to analyze the attentional preference characteristics of the tested child.

[0081] Specifically, the first frontal video is analyzed frame by frame, and the human body sub-image of the child being tested is determined in each frame. Then, a pre-trained skeletal point detection model is used to identify human skeletal points in multiple human body sub-images of the child being tested, and the first human body key point data of the child being tested corresponding to each human body sub-image is obtained.

[0082] The first set of human body key point data includes the coordinate information of multiple limb key points and facial key points of the child being tested. Among them, facial key points include, but are not limited to, the left and right corners of the eyes, the tip of the nose, and the corners of the mouth; limb key points include, but are not limited to, the shoulder joint, elbow joint, wrist joint, chest, abdomen, hip joint, knee joint, and ankle joint.

[0083] In some embodiments, based on the coordinate information of facial key points, the facial orientation of the tested child is analyzed using algorithms such as face detection / face alignment, and the distribution of the tested child's attention points is determined through spatial transformation.

[0084] In some embodiments, the limb deviation movements of the tested child can be analyzed based on the key point coordinates of each key point on the human body and the correlation between the key points. For example, the arm swing posture of the tested child can be analyzed based on the key point coordinates of the shoulder joint, the elbow joint, and the wrist joint.

[0085] Since the first frontal video consists of multiple consecutive frames, for a limb deviation action, the duration of the action can be calculated based on the number of image frames corresponding to the limb deviation action in the first frontal video and the frame interval duration.

[0086] Furthermore, based on the limb deviation movements, a focus distribution map of the tested children is determined. The focus distribution map is used to record the location of at least one primary region of interest that the tested children are looking at. Based on the duration of the movements, the focus duration of the tested children looking at each primary region of interest is determined. Based on the focus distribution map and focus duration, the attentional preference characteristics of the tested children are assessed.

[0087] It should be understood that the first region of interest can be the area where the tester is located in the preference test scenario, the area where the first test prop is located, or other areas in the preference test scenario.

[0088] Specifically, based on the attention distribution map, we can analyze the changes in the first area of ​​interest that the tested children prefer to focus on, and then determine whether the tested children's behavior is attracted by the first test prop, or by the tester, or whether the tested children are immersed in their own world and their actions are not affected by the tester and the first test prop.

[0089] Regarding attention span, the duration of each action of the child being tested can be analyzed to determine the duration of the child's gaze at the first test prop, or the duration and level of attention of the child's interaction with the tester.

[0090] Among them, attentional preference characteristics can include preference information, that is, whether the tested child prefers to communicate / spend time with people or with objects.

[0091] As an example, such as Figure 3 As shown, the first test prop is a balloon. During the preference test, the child is led into the preference test scenario set up in the testing room by a guardian / staff member, or the child enters the testing room on their own. As soon as the child enters the edge of the preference test scenario, the tester begins to greet them, making exaggeratedly positive expressions.

[0092] During the test, the tester faced the child and attracted the child's attention through a series of actions and facial expressions. A balloon was placed next to the tester to further attract the child's attention. The first image acquisition device was positioned behind the tester to clearly capture a frontal image of the child; the second image acquisition device recorded the interaction between the tester and the child; and the first sound acquisition device recorded the sound data during the test.

[0093] After the test, the first frontal video captured by the first image acquisition device was used to determine the child's reaction when viewing the tester and the first test prop. Specifically, the first image acquisition device recorded the child's head and limb data. Based on the head data, the child's facial orientation was determined, and changes in the child's focus were analyzed accordingly. Based on the limb data, information such as the child's limb orientation, walking behavior, and playtime duration in the preference test scenario was determined to assess the child's attentional preference characteristics.

[0094] (2) If the test scenario is a common interest test scenario, then the behavioral feature is a common interest feature;

[0095] In one possible implementation, step 220 can be performed as follows: frame parsing of the second interactive audio and / or the third frontal video; after detecting a second test start command in the second interactive audio and / or detecting the tester's second test start action in the third frontal video, obtaining the common attention area guided by the tester to the child under test from the third frontal video; after detecting the second test start command and / or the second test start action, obtaining the face images contained in each frame of the second frontal video; obtaining the head feature data of the child under test based on the face images; determining the second region of interest and the region attention duration based on the head feature data of the child under test; and evaluating the common attention characteristics of the child under test based on the common attention area, the second region of interest, and the region attention duration.

[0096] The common area of ​​interest is the area where any of the second test props is located; head feature data includes the child's head orientation, turning angle, and gaze direction.

[0097] Similarly, the second test start instruction can be the first sentence of the active verbal communication between the tester and the child being tested, or it can be other fixed starting sentences. The second test start action can be any action by the tester, such as raising their hand, waving their hand, or pointing to the second test prop; this application embodiment does not limit this.

[0098] In some embodiments, after acquiring the second interactive audio, the second interactive audio is detected frame by frame to identify the speech text corresponding to each frame of audio. Then, based on the sampling timestamp of the audio corresponding to the second test start instruction as the starting point, the starting frame image corresponding to the sampling timestamp is determined in the third frontal video. Then, multiple frames in the third frontal video located after the starting frame image are analyzed to determine the common attention area that the tester guides the tested child to focus on.

[0099] In some embodiments, after acquiring the third frontal video, the third frontal video is subjected to frame-by-frame detection to identify the behavior and actions of the tester in each frame. Then, based on the sampling timestamp of the image corresponding to the second test start action as the starting point, the starting frame image corresponding to the sampling timestamp is determined in the third frontal video. Then, multiple frames in the third frontal video located after the starting frame image are analyzed to determine the common attention area that the tester guides the child being tested to focus on.

[0100] Specifically, the third frontal video is analyzed frame by frame. First, the human body sub-image of the tester is determined in each frame. Then, a pre-trained skeletal point detection model is used to identify human skeletal points in multiple human body sub-images of the tester, obtaining the human body key point data of the tester in each sub-image. Finally, based on the human body key point data of the tester, the common attention area that the tester guides the children to focus on is determined.

[0101] The human body key point data of the tester includes the key point coordinates of multiple key points on the tester's body. For example, based on the key point coordinates of the elbow joint, the key point coordinates of the wrist joint, and the coordinates of other key points on the hand, the direction pointed by the tester's fingers can be analyzed, and then a common area of ​​interest can be determined based on that direction.

[0102] In addition to determining the common area of ​​focus for the tester, it is also necessary to determine whether the child being tested understands the tester's guidance and whether the child can follow the tester's guidance and focus on the common area of ​​focus.

[0103] Similarly, frame analysis is performed on the second frontal video of the tested child. First, the child's human body sub-image is determined in each frame. Then, a pre-trained skeletal point detection model is used to identify human skeletal points in multiple human body sub-images of the tested child, obtaining the human body key point data of the tested child in each human body sub-image. This human body key point data includes the coordinate information of facial key points and limb key points. Therefore, based on the coordinate information of the facial key points of the tested child, the head orientation, yaw angle, and gaze direction (i.e., eye gaze direction) of the tested child can be determined, obtaining head feature data. Based on the head feature data, the second region of interest of the tested child's gaze in the test scene can be determined.

[0104] Since the second frontal video consists of multiple consecutive images, for the same head feature data, the number of image frames corresponding to that head feature data in the second frontal video can be determined based on the head orientation and the same gaze direction. Then, based on the number of image frames and the frame interval duration, the regional attention duration of the tested child focusing on the second region of interest can be calculated.

[0105] Furthermore, the regional deviation between the common area of ​​interest and the second area of ​​interest is calculated to determine the attention deviation information of the tested children; the action response duration of the tested children is determined based on the second area of ​​interest and the duration of regional attention; and the common attention characteristics of the tested children are assessed based on the attention deviation information and the action response duration.

[0106] The second region of interest can be the area where the testers are located in the shared test scenario, the area where the second test prop is located, or other areas in the shared test scenario.

[0107] It should be understood that if the shared area of ​​attention and the second area of ​​interest are the same, it means that the second area of ​​interest that the child is looking at is the area that the tester is guiding and hoping the child will focus on. Therefore, when the second area of ​​interest that the child is looking at is the shared area of ​​attention, it indicates that the child can understand the tester's interactive intentions and can give an accurate response.

[0108] Action response time is the time it takes for the child being tested to understand the meaning of the actions guided by the tester and to respond accordingly during the guided attention process. This is used to determine whether the child being tested can correctly understand the meaning of the interaction and give the correct response action during the communication process, in order to assess the child's comprehension and responsiveness.

[0109] Commonly concerned characteristics may include information comprehension, responsiveness, and interaction skills, reflecting whether the tested child can communicate normally with the tester and generate correct responses to external interactive actions.

[0110] As an example, such as Figure 4 As shown, the second test prop is a set of dolls on a toy shelf. During the joint attention test, the child being tested is brought into the joint attention test scenario set up in the testing room by a guardian / staff member, or the child enters the joint attention test scenario in the testing room on their own. The tester sits / stands behind a children's table, and the child sits / stands in front of the table, with the tester and child interacting face-to-face. After the tester raises their hand to signal the start, the tester points to any doll on the toy shelf at the same level as the table, while simultaneously issuing a verbal command (e.g., "Look at this!"), and waits for the child to respond for 3-5 seconds. Multiple dolls on the shelf are pointed to in turn to guide the child's attention to the indicated doll.

[0111] The third and fourth image acquisition devices can be placed on a table to capture a second frontal video of the child being tested and a third frontal video of the tester.

[0112] After the test, based on the third frontal video of the tester, the tester's skeleton data was extracted. Using hand skeleton data and head turning, the common region of interest (ROI) was calculated. Simultaneously, the second frontal video of the child was split into frames. Face detection was performed in each frame, and face alignment was used to identify faces within the video. These face images were then input into a pre-trained neural network model to filter out irrelevant individuals and extract sub-images of the child to calculate head and limb feature data. Based on the head feature data, a direct linear transformation (DLP) problem was used to solve the Point of Interest (PnP) problem, obtaining the child's head tilt angle and pose relative to the camera, thus determining the second ROI of the child.

[0113] Finally, by comparing the shared area of ​​interest guided by the tester with the second area of ​​interest of the tested children, the degree of shared attention between the tested children and the tester was calculated to assess the children's ability to share attention.

[0114] (3) If the test scenario is a sound response test scenario, then the behavioral characteristics are sound response characteristics;

[0115] In one possible implementation, step 220 can be performed as follows: frame parsing of the reaction video to obtain the second human key point data of the child being tested contained in each frame of the reaction video; determining the pose change information and reaction time of the child being tested based on the second human key point data corresponding to each frame; and evaluating the vocal response characteristics of the child being tested based on the pose change information and reaction time.

[0116] Among them, human body key point data extraction and pose calculation can be achieved through pre-trained neural network models, which improves feature extraction efficiency and calculation accuracy.

[0117] In some embodiments, sound response characteristics may include reaction actions and reaction durations to sound stimuli, used to describe the tested child's ability to respond to sound stimuli.

[0118] As an example, such as Figure 5 As shown, the controllable sound-producing object is a bell. During the sound response test, the child is either led into the sound response test scenario set up in the testing room by a guardian / staff member, or the child enters the sound response test scenario on their own. When the child, guided by the tester or walking independently, passes the bell hanging on the wall, the evaluator in the control room controls the bell to emit a sound via data processing equipment, awaiting the child's response.

[0119] The fifth image acquisition device can be placed to the side and rear of the child being tested to capture a video of the child's reaction after hearing the bell sound.

[0120] After the test, the skeletal data of the tested child is extracted from the reaction video, the head position of the tested child is located and the head image is extracted, and / or the limb position of the tested child is located and the full-body image is extracted; the head image and / or the full-body image are input into the pre-trained motion detection model, and the amplitude, angle and speed of the child's head / body deflection movement are calculated according to the optical flow method to determine the head rotation and reaction time of the tested child after hearing the bell sound, thereby evaluating the child's sound response ability.

[0121] Step 230: Based on the behavioral characteristics, assess the probability of the tested child having abnormal behavior.

[0122] It should be noted that when assessing the probability of abnormal behavior in the tested child, the assessment can be based on behavioral characteristics in one test scenario or by combining multiple behavioral characteristics obtained from multiple test scenarios. This application embodiment does not impose any restrictions on this.

[0123] If the test scenarios include preference testing, shared attention testing, and vocal response testing, then after testing them sequentially, the attention preference characteristics, shared attention characteristics, and vocal response characteristics of the tested child can be analyzed. In one possible implementation, step 230 involves: obtaining the influence weights of each behavioral characteristic, and combining each behavioral characteristic with its corresponding influence weight to assess the probability of behavioral abnormalities in the tested child.

[0124] Alternatively, deep learning can be performed on massive amounts of data to train a behavioral abnormality detection model. Then, after obtaining the behavioral test data of the children being tested, the model can be used to extract features and output the probability of behavioral abnormalities.

[0125] Furthermore, after assessing the presence of abnormal behavior in the tested child, it is also possible to assess whether the tested child has a neurodevelopmental disorder related to such abnormal behavior.

[0126] For example, based on the child's attentional preferences, shared attention patterns, and vocal responses, it can be determined whether the child exhibits abnormal behaviors such as social interaction difficulties, communication problems, narrow interests, or repetitive stereotyped behaviors. If such abnormal behaviors are present, the risk of the child having autism spectrum disorder can be further assessed.

[0127] In this embodiment, when the child being tested is located in at least one test scenario displayed in a testing room, behavioral test data of the child is acquired by a data acquisition device set up in the testing room. The behavioral test data is analyzed and processed to determine the child's behavioral characteristics. Based on these characteristics, the probability of the child exhibiting behavioral abnormalities is assessed. The test scenario corresponds to the type of abnormal behavior to be assessed in the child, and the abnormal behavior type includes at least one manifestation of autism spectrum disorder. Therefore, this application collects behavioral test data of the child in a real-world scenario using a data acquisition device. This behavioral test data can comprehensively and accurately record the child's actual actions and behaviors. Thus, by analyzing the child's behavioral test data, the behavioral characteristics of the child can be determined to assess the likelihood of abnormal behavior. Furthermore, in the process of assessing behavioral abnormalities, it is only necessary to pre-set a real test scenario corresponding to the type of abnormal behavior and deploy the data acquisition device within that scenario. The testing process does not require a professional physician to guide the child through multiple test items according to the testing rules, nor does it rely on the medical experience of a professional physician to determine the test results. This shortens the testing time, improves the efficiency of assessing abnormal behavior, and thus improves the screening efficiency of neurodevelopmental disorders related to behavioral abnormalities.

[0128] Based on the same technical concept, this application also provides a device for assessing abnormal child behavior corresponding to the above-described method for assessing abnormal child behavior. The implementation scheme provided by this device in solving the technical problem is similar to the implementation scheme described in the above-described method embodiments. Therefore, the specific functional limitations of one or more embodiments of the abnormal behavior assessment device provided below can be found in the limitations of the relevant steps in the above-described method for assessing abnormal child behavior, and will not be repeated here.

[0129] In one exemplary embodiment, such as Figure 6 As shown in the illustration, this application also provides a behavioral anomaly assessment device, the device 600 comprising:

[0130] The data acquisition module 610 is used to acquire behavioral test data of the child being tested collected by the data acquisition device set in the test room when the child being tested is in at least one test scenario displayed in the test room; the test scenario corresponds to the abnormal behavior type to be evaluated of the child being tested, and the abnormal behavior type includes at least one manifestation behavior corresponding to autism spectrum disorder.

[0131] Data analysis module 620 is used to analyze and process the behavioral test data of the tested children to determine their behavioral characteristics;

[0132] The behavioral assessment module 630 is used to assess the probability of behavioral abnormalities in the tested child based on behavioral characteristics.

[0133] In one possible implementation, the test scenario includes a preference test scenario, which includes a tester and at least one first test prop for attracting the attention of the child being tested.

[0134] The data acquisition equipment includes a first image acquisition device, a second image acquisition device, and a first sound acquisition device set in the test scenario. During the process of the child being tested being opposite the tester and the tester and the child being tested communicating face-to-face, the first image acquisition device is used to acquire the first frontal video of the child being tested, the second image acquisition device is used to acquire the first interactive video of the tester and the child being tested, and the first sound acquisition device is used to acquire the first interactive audio of the tester and the child being tested.

[0135] Correspondingly, the behavioral test data includes a first positive video, a first interactive video, and a first interactive audio.

[0136] In one possible implementation, the behavioral characteristics include attentional preference characteristics; the data analysis module 620 includes:

[0137] The first audio and video parsing unit is used to perform frame parsing on the first interactive audio and / or the first interactive video.

[0138] The first key point detection unit is used to detect the first test start command in the first interactive audio, and / or, after detecting the tester's first test start action in the first interactive video, acquire the first human key point data of the child being tested contained in each frame of the first frontal video.

[0139] The first motion analysis unit is used to analyze the limb deviation movements and the duration of the corresponding movements of the tested child based on the first human key point data corresponding to each frame of the image.

[0140] The first feature assessment unit is used to assess the attentional preference characteristics of the tested child based on limb bias and duration of movement.

[0141] In one possible implementation, the feature evaluation unit is specifically used for:

[0142] Based on limb deviation movements, determine the distribution map of the child's attention points; the distribution map of attention points is used to record the location of at least one primary region of interest that the child's gaze is fixed on;

[0143] Based on the duration of the action, determine the attention span of the child being tested when focusing on each primary area of ​​interest;

[0144] Based on the attention distribution map and attention duration, the attentional preference characteristics of the tested children were assessed.

[0145] In one possible implementation, the test scenario includes a joint attention test scenario, which includes a tester and multiple second test props set in front of the tester to guide the children being tested to jointly look at each other.

[0146] The data acquisition equipment includes a third image acquisition device, a fourth image acquisition device, and a second sound acquisition device set in the test scenario; while the child being tested is facing the tester and the tester is guiding the child to look at the test prop, the third image acquisition device is used to acquire a second frontal video of the child being tested, the fourth image acquisition device is used to acquire a third frontal video of the tester, and the second sound acquisition device is used to acquire a second interactive audio between the tester and the child being tested.

[0147] Correspondingly, the behavioral test data includes a second frontal video, a third frontal video, and a second interactive audio.

[0148] In one possible implementation, the behavioral characteristics include shared interests; the data analysis module 620 includes:

[0149] The second audio and video parsing unit is used to perform frame parsing on the second interactive audio and / or the third frontal video.

[0150] The area determination unit is used to detect the second test start instruction in the second interactive audio, and / or, after detecting the tester's second test start action in the third frontal video, to obtain the common attention area that the tester guides the child being tested to focus on from the third frontal video; the common attention area is the area where any of the second test props is located;

[0151] The face detection unit is used to acquire face images contained in each frame of the second frontal video after detecting the second test start command and / or the second test start action;

[0152] The feature extraction unit is used to obtain head feature data of the child being tested based on the face image; the head feature data includes the head orientation, turning angle and gaze direction of the child being tested.

[0153] The second motion analysis unit is used to determine the second region of interest that the child is looking at and the duration of attention to the region based on the head feature data of the child being tested.

[0154] The second feature assessment unit is used to assess the common attention features of the tested children based on the common attention area, the second interest area, and the duration of attention to the area.

[0155] In one possible implementation, the second feature evaluation unit is specifically used for:

[0156] Calculate the regional deviation between the common region of interest and the second region of interest to determine the attentional bias information of the tested children;

[0157] The duration of motor response of the tested child was determined based on the second area of ​​interest and the duration of attention in that area.

[0158] Based on attentional bias information and action response duration, assess the common attentional characteristics of the tested children.

[0159] In one possible implementation, the test scenario includes a sound response test scenario, which includes at least one controllable sound-emitting object;

[0160] The data acquisition equipment includes a fifth image acquisition device and a third sound acquisition device set in the test scenario; after the child being tested is within a preset range of the controllable sound-emitting object and the controllable sound-emitting object emits sound, the fifth image acquisition device is used to acquire the video of the child's reaction, and the third sound acquisition device is used to acquire the audio of the child's reaction.

[0161] Correspondingly, the behavioral test data includes reaction videos and reaction audios.

[0162] In one possible implementation, the behavioral characteristics include vocal response characteristics; the data analysis module 620 includes:

[0163] The second key point detection unit is used to perform frame analysis on the reaction video and obtain the second human key point data of the tested child contained in each frame of the reaction video.

[0164] The third motion analysis unit is used to determine the posture change information and reaction time of the tested child based on the second human key point data corresponding to each frame image.

[0165] The third feature assessment unit is used to assess the vocal response characteristics of the tested child based on posture change information and reaction time.

[0166] It should be noted that each module in the aforementioned behavioral anomaly assessment device can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of a computer device in hardware form or independent of it, or stored in the memory of a computer device in software form, so that the processor can call and execute the corresponding operations of each module.

[0167] In one exemplary embodiment, this application also provides a computer device. This computer device can serve as a data processing device for implementing the child behavioral abnormality assessment method described in the foregoing embodiments. Figure 7 As shown, the computer device includes one or more processors 710, memory 720, system bus 730 and communication interface 740, and the processors 710, memory 720 and communication interface 740 are connected through system bus 730.

[0168] The processor can be a central processing unit (CPU) or other processing unit with data processing and / or instruction execution capabilities, and can control other components in the computer device to perform the desired functions.

[0169] Optionally, the processor may include application software for data processing and other related functions.

[0170] The memory may include one or more computer program products, which may include various forms of computer-readable storage media, such as volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and / or cache memory. As an example, the non-volatile memory may include read-only memory (ROM), a hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium, and a processor may execute these instructions to implement the child behavioral abnormality assessment methods and / or other desired functions in the embodiments shown above.

[0171] The communication interface of this computer device is used to communicate with external terminals via wired or wireless means. Wireless communication can be achieved through WIFI, carrier networks, NFC (Near Field Communication), or other technologies.

[0172] In some embodiments, the computer device may further include input devices and output devices (not shown), which are interconnected via a bus system and / or other forms of connection mechanisms (not shown). The input device may be a touch layer covering the display screen, a button, trackball, or touchpad mounted on the computer device casing, or an external keyboard, touchpad, or mouse. The output device may output various information to the outside, such as a monitor / display screen, speakers, and communication networks and their connected remote output devices.

[0173] Of course, for the sake of simplicity, Figure 7 Only some of the components of the computer device relevant to the embodiments of this application are shown in the illustration. In addition, the computer device may include any other suitable components depending on the specific application.

[0174] In one exemplary embodiment, this application also provides a computer-readable storage medium storing a computer program that, when executed by a processor in a computer device, can implement the child behavior abnormality assessment method provided in the above embodiments.

[0175] In one exemplary embodiment, this application also provides a computer program product, including a computer program that, when executed by a processor in a computer device, can implement the child behavior abnormality assessment method provided in the above embodiments.

[0176] The above detailed embodiments further illustrate the purpose, technical solution, and beneficial effects of the embodiments of this application. It should be understood that the above are merely specific embodiments of the embodiments of this application and are not intended to limit the protection scope of the embodiments of this application. Any modifications, equivalent substitutions, improvements, etc., made on the basis of the technical solutions of the embodiments of this application should be included within the protection scope of the embodiments of this application.

Claims

1. A method for assessing abnormal behavior in children, characterized in that, include: In a test room where the child is in at least one test scenario, behavioral test data of the child is acquired by a data acquisition device set up in the test room; the test scenario corresponds to the abnormal behavior type to be evaluated of the child, and the abnormal behavior type includes at least one manifestation behavior corresponding to autism spectrum disorder. The behavioral test data of the tested children are analyzed and processed to determine the behavioral characteristics of the tested children; Based on the behavioral characteristics, assess the probability of behavioral abnormalities in the tested child; The test scenario includes a preference test scenario, which includes a tester and at least one first test prop for attracting the attention of the child being tested; The data acquisition device includes a first image acquisition device, a second image acquisition device, and a first sound acquisition device, all installed in the test scenario. During the process where the child being tested is positioned opposite the tester and the tester is engaging in face-to-face communication with the child being tested, the first image acquisition device is used to acquire a first frontal video of the child being tested, the second image acquisition device is used to acquire a first interactive video of the tester and the child being tested, and the first sound acquisition device is used to acquire a first interactive audio of the tester and the child being tested. Correspondingly, the behavioral test data includes the first frontal video, the first interactive video, and the first interactive audio; The test scenario includes a joint attention test scenario, which includes a tester and multiple second test props set in front of the tester to guide the child being tested to jointly look at the child. The data acquisition device includes a third image acquisition device, a fourth image acquisition device, and a second sound acquisition device installed in the test scenario; While the child being tested is positioned opposite the tester, and the tester is guiding the child to look at the test prop, the third image acquisition device is used to acquire a second frontal video of the child being tested, the fourth image acquisition device is used to acquire a third frontal video of the tester, and the second sound acquisition device is used to acquire a second interactive audio between the tester and the child being tested. Correspondingly, the behavioral test data includes the second frontal video, the third frontal video, and the second interactive audio.

2. The method according to claim 1, characterized in that, The behavioral characteristics include attentional preference characteristics; The process of analyzing and processing the behavioral test data of the tested children to determine their behavioral characteristics includes: Perform frame parsing on the first interactive audio and / or the first interactive video; After detecting a first test start command in the first interactive audio and / or after detecting the tester's first test start action in the first interactive video, the first human key point data of the child under test contained in each frame of the first frontal video is obtained. Based on the first human key point data corresponding to each frame of the image, the limb deviation movement of the tested child and the duration of the movement corresponding to the limb deviation movement are analyzed. The attentional preference characteristics of the tested children were assessed based on the limb deviation movements and the duration of the movements.

3. The method according to claim 2, characterized in that, The assessment of the child's attentional preference characteristics based on the limb deviation movement and the duration of the movement includes: Based on the described limb deviation movements, a focus distribution map of the tested child is determined; the focus distribution map is used to record the location of at least one first region of interest that the tested child is looking at. Based on the duration of the action, the focus duration of the tested child when gazing at each of the first regions of interest is determined; Based on the attention distribution map and the attention duration, the attentional preference characteristics of the tested children were assessed.

4. The method according to claim 1, characterized in that, The behavioral characteristics include shared interests; The process of analyzing and processing the behavioral test data of the tested children to determine their behavioral characteristics includes: Frame parsing is performed on the second interactive audio and / or the third frontal video; After a second test start command is detected in the second interactive audio, and / or after the tester's second test start action is detected in the third frontal video, a common area of ​​attention is obtained from the third frontal video, in which the tester guides the child being tested to focus; the common area of ​​attention is the area where any of the second test props is located. After detecting the second test start command and / or the second test start action, acquire the face images contained in each frame of the second frontal video; Based on the facial image, obtain the head feature data of the child being tested; the head feature data includes the head orientation, turning angle, and gaze direction of the child being tested; Based on the head feature data of the tested child, the second region of interest that the tested child is looking at and the duration of attention to the region are determined; The common attention characteristics of the tested children are assessed based on the common attention area, the second interest area, and the attention duration of the area.

5. The method according to claim 4, characterized in that, The assessment of the shared attention characteristics of the tested child based on the shared attention area, the second region of interest, and the duration of attention to the region includes: Calculate the regional deviation between the common region of interest and the second region of interest to determine the attentional deviation information of the tested child; The action response duration of the tested child is determined based on the second region of interest and the duration of attention to the region; Based on the attentional bias information and the action response duration, the common attentional characteristics of the tested children are assessed.

6. The method according to any one of claims 1 to 3, characterized in that, The test scenario includes a sound response test scenario, which includes at least one controllable sound-emitting object; The data acquisition device includes a fifth image acquisition device and a third sound acquisition device installed in the test scenario; After the child being tested is within a preset range of the controllable sound-emitting object, and the controllable sound-emitting object emits a sound, the fifth image acquisition device is used to acquire the video of the child's reaction, and the third sound acquisition device is used to acquire the audio of the child's reaction. Correspondingly, the behavioral test data includes the reaction video and the reaction audio.

7. The method according to claim 6, characterized in that, The behavioral characteristics include vocal response characteristics; The process of analyzing and processing the behavioral test data of the tested children to determine their behavioral characteristics includes: The reaction video is analyzed frame by frame to obtain the second human key point data of the tested child contained in each frame of the reaction video; Based on the second human key point data corresponding to each frame of the image, the posture change information and reaction time of the tested child are determined; The vocal response characteristics of the tested child are evaluated based on the posture change information and the reaction time.

8. A system for assessing abnormal behavior in children, characterized in that, It includes a main control room and a testing room. The main control room includes data processing equipment, and the testing room is equipped with data acquisition equipment. When the child being tested is presenting at least one test scenario in the test room, the data acquisition device is used to acquire the behavioral test data of the child being tested and send the behavioral test data to the data processing device; the test scenario corresponds to the abnormal behavior type to be evaluated of the child being tested, and the abnormal behavior type includes at least one manifestation behavior corresponding to autism spectrum disorder. The data processing device is used to analyze and process the behavioral test data of the tested child, determine the behavioral characteristics of the tested child, and assess the probability of the tested child's abnormal behavior based on the behavioral characteristics. The test scenarios include at least one of the following: preference test scenario, common attention test scenario, and voice response test scenario; The preference test scenario includes a tester and at least one first test prop for attracting the gaze of the child being tested; the joint attention test scenario includes a tester and multiple second test props placed in front of the tester to guide the child being tested to jointly look at the same object; the sound response test scenario includes at least one controllable sound-emitting object.

9. The system according to claim 8, characterized in that, If the test scenario is a preference test scenario, then the data acquisition device includes a first image acquisition device, a second image acquisition device, and a first sound acquisition device disposed in the test scenario; During the process where the child being tested is positioned opposite the tester and the tester is engaging in face-to-face communication with the child being tested, the first image acquisition device is used to acquire a first frontal video of the child being tested, the second image acquisition device is used to acquire a first interactive video of the tester and the child being tested, and the first sound acquisition device is used to acquire a first interactive audio of the tester and the child being tested. Correspondingly, the behavioral test data includes the first frontal video, the first interactive video, and the first interactive audio.

10. The system according to claim 8, characterized in that, If the test scenario is a common interest test scenario, then the data acquisition device includes a third image acquisition device, a fourth image acquisition device, and a second sound acquisition device set in the test scenario; While the child being tested is positioned opposite the tester, and the tester is guiding the child to look at the test prop, the third image acquisition device is used to acquire a second frontal video of the child being tested, the fourth image acquisition device is used to acquire a third frontal video of the tester, and the second sound acquisition device is used to acquire a second interactive audio between the tester and the child being tested. Correspondingly, the behavioral test data includes the second frontal video, the third frontal video, and the second interactive audio.

11. The system according to claim 8, characterized in that, If the test scenario is a sound response test scenario, then the data acquisition device includes a fifth image acquisition device and a third sound acquisition device disposed in the test scenario; After the child being tested is within a preset range of the controllable sound-emitting object, and the controllable sound-emitting object emits a sound, the fifth image acquisition device is used to acquire the video of the child's reaction, and the third sound acquisition device is used to acquire the audio of the child's reaction. Correspondingly, the behavioral test data includes the reaction video and the reaction audio.

Images