阅读说明：本技术 心理危机预警系统 (Psychological crisis early warning system ) 是由 吴婷 于 2020-11-30 设计创作，主要内容包括：本发明涉及心理分析技术领域,具体涉及一种心理危机预警系统,包括：采集模块,用于实时采集监控视频；语音模块,用于提取监控视频中检测对象的语音数据,并根据语音的特点初步判断检测对象是否可能出现心理危机；视频模块,用于提取监控视频中的图像数据,分析图像中检测对象的肢体动作,并根据肢体动作的特点再次判断检测对象是否可能出现心理危机；特征模块,用于提取监控视频中的面部图像,分析检测对象的人体特征数据,判断检测对象是否处于紧张状态；预警模块,用于接收指令,并进行预警。本发明只有当检测对象处于紧张状态时,才判定检测对象出现心理危机,解决了现有技术只对语音和视频进行分析不能精确识别出心理危机的技术问题。(The invention relates to the technical field of psychological analysis, in particular to a psychological crisis early warning system, which comprises: the acquisition module is used for acquiring the monitoring video in real time; the voice module is used for extracting voice data of a detected object in the monitoring video and preliminarily judging whether the detected object possibly has a psychological crisis or not according to the characteristics of the voice; the video module is used for extracting image data in the monitoring video, analyzing the limb movement of the detected object in the image, and judging whether the detected object possibly has a psychological crisis again according to the characteristics of the limb movement; the characteristic module is used for extracting a face image in the monitoring video, analyzing human body characteristic data of the detected object and judging whether the detected object is in a tension state; and the early warning module is used for receiving the instruction and carrying out early warning. The invention judges that the detected object has psychological crisis only when the detected object is in a nervous state, and solves the technical problem that the psychological crisis can not be accurately identified only by analyzing voice and video in the prior art.)

1. Psychological crisis early warning system, its characterized in that includes:

the acquisition module is used for acquiring the monitoring video in real time;

the voice module is used for extracting voice data of the detected object in the monitoring video and preliminarily judging whether the detected object possibly has a psychological crisis or not according to the characteristics of the voice: if the detected object is unlikely to have psychological crisis, judging that early warning is not needed; on the contrary, if the detected object possibly has psychological crisis, sending an instruction to the video module for judging again;

the video module is used for extracting image data in the monitoring video, analyzing the limb movement of the detected object in the image, and judging whether the detected object possibly has a psychological crisis again according to the characteristics of the limb movement: if the detected object is unlikely to have psychological crisis, judging that early warning is not needed; on the contrary, if the detected object possibly has psychological crisis, sending an instruction to the characteristic module for final judgment;

the characteristic module is used for extracting the face image in the monitoring video, analyzing the human body characteristic data of the detection object and judging whether the detection object is in a tension state: if the detected object is not in a tension state, judging that the detected object has no psychological crisis and does not need early warning; on the contrary, if the detection object is in a tension state, judging that the detection object has a psychological crisis, needing early warning, and sending an instruction to an early warning module for early warning;

and the early warning module is used for receiving the instruction and carrying out early warning.

2. The psychological crisis warning system of claim 1, wherein the human characteristic data includes micro-expression data and micro-action data.

3. The psychological crisis warning system according to claim 2, characterized in that the tiny changes of the face of the detection object in the surveillance video are input to the neural network for deep learning, and the micro-expression data is obtained; and amplifying the micro-motion of the detected object in the monitoring video, detecting the change rule of the micro-motion, and acquiring micro-motion data.

4. The psychological crisis warning system of claim 3, wherein the human characteristic data further includes heart rate, blood oxygen concentration and blood pressure difference.

5. The psychological crisis warning system of claim 4, wherein the heart rate and blood oxygen concentration of the subject are obtained by detecting the variation law of capillary congestion during respiration; and acquiring the blood pressure difference of the detection object by detecting the phase difference of the blood flow velocity of the human body part.

6. The psychological crisis warning system of claim 5, wherein the current status of the detected object is determined according to the body movement, and includes a general status, an approaching status, a dispute status, and a fighting status.

7. The psychological crisis warning system of claim 6, wherein the tendency of the body movement of the subject is predicted according to the judged current state.

8. The psychological crisis warning system of claim 7, wherein the test subjects are classified according to the human body characteristic data after it is determined that the test subjects are at psychological crisis.

9. The psychological crisis warning system of claim 8, wherein the detected subjects are classified in the form of grades, including generally severe, relatively severe, quite severe and particularly severe.

10. The psychological crisis warning system of claim 9, wherein when it is determined that the psychological crisis occurs in the test subject, the core characteristics of the test subject are analyzed based on the human body characteristic data.

Technical Field

The invention relates to the technical field of psychological analysis, in particular to a psychological crisis early warning system.

Background

Psychological crisis refers to the psychological response of an individual when the individual encounters an emergency or faces significant frustration and difficulty, and the individual cannot avoid the psychological response and cannot solve the psychological response in a resource and stress manner. The psychological crisis not only affects the physical development and personality development of students, but also may affect the classroom education and teaching, so it is necessary to early warn, discover and intervene the psychological crisis of students.

For example, document CN109101933A discloses an emotional behavior visualization analysis method based on artificial intelligence, which includes the steps of: the method comprises the steps of obtaining and storing image data and voice data collected by shooting equipment in real time, obtaining a teaching model, inputting the teaching model into a convolutional neural network, mining and analyzing the image data and the voice data by adopting the convolutional neural network to obtain and store a teacher-student emotion analysis result and a teacher-student behavior analysis result, generating and storing a visual report according to the teacher-student emotion analysis result and the teacher-student behavior analysis result, performing data interpretation on the visual report by adopting the convolutional neural network to obtain and store a data interpretation result, and performing teaching diagnosis and early warning according to the data interpretation result.

Though, the artificial intelligence convolutional neural network is combined to realize visual analysis and interpretation of the emotion and the behavior of teachers and students, and therefore early warning is carried out according to interpretation results. However, the forming process of the psychological crisis of the human being is quite complex, the external information such as voice and video cannot reflect the change of the physiological characteristics of the human body, and the physiological characteristics can reflect the real psychological condition better than the external information. That is, if only voice and video are analyzed, psychological crisis cannot be accurately recognized, and thus early warning measures cannot be taken in time.

Disclosure of Invention

The invention provides a psychological crisis early warning system, which solves the technical problem that only voice and video are analyzed in the prior art, and the psychological crisis cannot be accurately identified.

The basic scheme provided by the invention is as follows: psychological crisis early warning system includes:

the acquisition module is used for acquiring the monitoring video in real time;

the voice module is used for extracting voice data of the detected object in the monitoring video and preliminarily judging whether the detected object possibly has a psychological crisis or not according to the characteristics of the voice: if the detected object is unlikely to have psychological crisis, judging that early warning is not needed; on the contrary, if the detected object possibly has psychological crisis, sending an instruction to the video module for judging again;

the video module is used for extracting image data in the monitoring video, analyzing the limb movement of the detected object in the image, and judging whether the detected object possibly has a psychological crisis again according to the characteristics of the limb movement: if the detected object is unlikely to have psychological crisis, judging that early warning is not needed; on the contrary, if the detected object possibly has psychological crisis, sending an instruction to the characteristic module for final judgment;

the characteristic module is used for extracting the face image in the monitoring video, analyzing the human body characteristic data of the detection object and judging whether the detection object is in a tension state: if the detected object is not in a tension state, judging that the detected object has no psychological crisis and does not need early warning; on the contrary, if the detection object is in a tension state, judging that the detection object has a psychological crisis, needing early warning, and sending an instruction to an early warning module for early warning;

and the early warning module is used for receiving the instruction and carrying out early warning.

The working principle and the advantages of the invention are as follows: whether two levels need to be included is judged: firstly, the external performance of the detection object is judged, namely whether the detection object is possible to have a psychological crisis is judged according to the voice characteristics and the body action characteristics of the detection object, and if the detection object is not possible to have the psychological crisis, the judgment is directly made without early warning. Because the voice characteristics and the body actions can reflect most psychological crisis states, most external expressions of the detected objects can be interpreted in such a way, and the efficiency of psychological crisis recognition is improved. Secondly, if psychological crisis may occur from the external expression of the detection object, then the judgment is performed from the internal state of the detection personnel, that is, whether the detection object is in a tension state is judged according to the human body characteristic data of the detection object, and only when the detection object is in the tension state, the detection object is judged to have the psychological crisis and needs to be warned. The human body characteristic data is not influenced by subjective consciousness of people to a great extent and is not easy to be covered by the detected object, so that the truth degree is higher.

The invention judges whether the detection object is in a tension state or not according to the human body characteristic data of the detection object, and judges that the detection object has a psychological crisis only when the detection object is in the tension state, thereby solving the technical problem that the psychological crisis can not be accurately identified only by analyzing voice and video in the prior art.

Further, the human characteristic data includes micro-expression data and micro-action data.

Has the advantages that: the related psychological studies show that the micro-expression and the micro-action can be regarded as almost natural exposure of the mind, and are difficult to be deliberately disguised, and the psychological state of the detected object can be accurately interpreted by the micro-expression data and the micro-action data.

Further, inputting the tiny changes of the face of the detection object in the monitoring video into a neural network for deep learning, and acquiring micro-expression data; and amplifying the micro-motion of the detected object in the monitoring video, detecting the change rule of the micro-motion, and acquiring micro-motion data.

Has the advantages that: because the micro expression and the micro action have small amplitude and change and are not easily recognized by naked eyes, the accuracy of recognizing the micro expression and the micro action can be improved in such a way.

Further, the human characteristic data also includes heart rate, blood oxygen concentration and blood pressure difference.

Has the advantages that: as known from the physiological related knowledge, the heart rate, the blood oxygen concentration and the blood pressure are closely related to the psychological condition and the state of tension, and are not controlled by the consciousness of people, and the mental condition of the detected object can be objectively and accurately interpreted through the data.

Further, the heart rate and the blood oxygen concentration of the detected object are obtained by detecting the change rule of the blood congestion of the capillary vessels during respiration; and acquiring the blood pressure difference of the detection object by detecting the phase difference of the blood flow velocity of the human body part.

Has the advantages that: by the method, the facial image area of the detection object can be analyzed, a mechanical device is not needed for measurement, and the detection efficiency is improved.

Further, the current state of the detection object is judged according to the limb actions, and the current state comprises a common state, an approaching state, a dispute state and a fighting state.

Has the advantages that: by the method, the emotional condition of the detected object can be qualitatively read, and the result is visual and convenient to analyze.

And further, pre-judging the trend of the limb movement of the detection object according to the judged current state.

Has the advantages that: in this way, the development process of the limb movement can be analyzed in advance, so that the psychological crisis can be recognized in advance to prevent the occurrence of the disease.

Further, after the psychological crisis of the detected object is judged, the detected object is classified according to the human body characteristic data.

Has the advantages that: in this way, the category of the detection object with psychological crisis can be determined, and therefore, the targeted early warning is facilitated.

Further, the test subjects are classified in a hierarchical manner, including generally severe, more severe, quite severe, and particularly severe.

Has the advantages that: by the method, more effective early warning measures can be facilitated, and responsibility consciousness is improved.

Further, when the psychological crisis of the detected object is judged, the core characteristics of the detected object are analyzed according to the human body characteristic data.

Has the advantages that: by the mode, subsequent targeted education and guidance are facilitated, and the probability of occurrence of over-excited behaviors and uncontrolled behaviors is reduced.

Drawings

Fig. 1 is a block diagram of a system structure of an embodiment of the psychological crisis warning system according to the present invention.

Detailed Description

The following is further detailed by the specific embodiments:

example 1

The embodiment of the psychological crisis early warning system of the invention is basically as shown in the attached figure 1, and comprises:

the acquisition module is used for acquiring the monitoring video in real time;

the voice module is used for extracting voice data of the detected object in the monitoring video and preliminarily judging whether the detected object possibly has a psychological crisis or not according to the characteristics of the voice: if the detected object is unlikely to have psychological crisis, judging that early warning is not needed; on the contrary, if the detected object possibly has psychological crisis, sending an instruction to the video module for judging again;

the video module is used for extracting image data in the monitoring video, analyzing the limb movement of the detected object in the image, and judging whether the detected object possibly has a psychological crisis again according to the characteristics of the limb movement: if the detected object is unlikely to have psychological crisis, judging that early warning is not needed; on the contrary, if the detected object possibly has psychological crisis, sending an instruction to the characteristic module for final judgment;

the characteristic module is used for extracting the face image in the monitoring video, analyzing the human body characteristic data of the detection object and judging whether the detection object is in a tension state: if the detected object is not in a tension state, judging that the detected object has no psychological crisis and does not need early warning; on the contrary, if the detection object is in a tension state, judging that the detection object has a psychological crisis, needing early warning, and sending an instruction to an early warning module for early warning;

and the early warning module is used for receiving the instruction and carrying out early warning.

In this embodiment, the acquisition module is a camera, the voice module, the video module and the feature module are all integrated on the server, the functions are realized through software/programs/codes, and the early warning module is a buzzer.

The specific implementation process is as follows:

and S1, the acquisition module acquires the monitoring video in real time.

The camera is installed in a classroom, monitoring videos in the classroom are collected in real time, and the collected monitoring videos are sent to the server.

S2, the voice module extracts the voice data of the detection object in the monitoring video, and preliminarily judges whether the detection object possibly has a psychological crisis according to the characteristics of the voice: if the detected object is unlikely to have psychological crisis, judging that early warning is not needed; on the contrary, if the detected object may have psychological crisis, an instruction is sent to the video module for judging again.

Specifically, Praat voice analysis software is loaded on the server, and the Praat voice analysis software can acquire words with higher voice, words with longer speaking time and words with higher occurrence frequency in the talking process of students in the teacher. In this embodiment, Praat voice analysis software is used to analyze the voice data in the monitoring video to obtain an analysis result.

For example, if the analysis results of the Praat speech analysis software show that the classroom is quiet and substantially free of any sounds. In this case, the probability of the occurrence of an overexcitation behavior by the students in the classroom is small, so that it can be directly determined that there is no possibility of a psychological crisis occurring by the students in the classroom.

For another example, if the analysis result of Praat speech analysis software indicates that some crude, dirty, or abusive speech is mentioned, the pitch of one or several students is higher than 10% of the average pitch of the whole speech, which indicates that there may be students who are quarreling, and that the quarreling students may be psychologically crisis and thus may be overexcited. At this time, the voice module judges that the detected object may have a psychological crisis, so that an instruction is sent to the video module for judging again.

S3, the video module extracts image data in the monitoring video, analyzes the limb movement of the detected object in the image, and judges whether the detected object may have a psychological crisis again according to the characteristics of the limb movement: if the detected object is unlikely to have psychological crisis, judging that early warning is not needed; otherwise, if the detected object possibly has psychological crisis, sending an instruction to the characteristic module for final judgment.

Specifically, the emotional conditions of the students are qualitatively interpreted, namely the current states of the students are judged according to the limb actions, wherein the current states comprise a common state, an approaching state, a dispute state and a shelving state; and then prejudging the trend of the limb actions of the student according to the judged current state. For example, the body actions of students in a classroom and the distance between two adjacent students are acquired through an image recognition technology, the states of the students are analyzed according to the distance between the students, and the current states of the students are judged: more than 1m in a normal state, 0.5-1 m in an approaching state, 0.3-0.5 m in a dispute state, and 0-0.3 m in a shelving state. If the state of the student is a common state, the probability of conflict is low due to the long distance, so that the possibility that the student is in psychological crisis can be directly judged.

On the contrary, if the student is in a close state, the collision may occur due to the close distance, and the trend of the limb movement of the student is predicted. In this embodiment, FaceReader software is loaded on the server, and the software can identify expressions such as pleasure, sadness, fear, disgust, surprise, anger, nature, slight and the like, identify facial expressions of students through the FaceReader software, and predict the trend of the limb movement according to the facial expressions. If the facial expression of the student is 'happy' or 'natural', a 'good' prejudgment is made on the trend of the limb movement; if the student's facial expression is "angry" or "dislike", a prediction of "worsening" is made on the trend of the limb movements.

If the trend of the limb movement is 'deterioration', the situation shows that the students may have psychological crisis, but final judgment is still needed, namely, an instruction is sent to the characteristic module for final judgment.

S4, extracting the face image in the monitoring video by the feature module, analyzing the human body feature data of the detection object, and judging whether the detection object is in a tension state: if the detected object is not in a tension state, judging that the detected object has no psychological crisis and does not need early warning; on the contrary, if the detection object is in a tension state, the detection object is judged to have a psychological crisis, early warning is needed, and an instruction is sent to the early warning module to carry out early warning.

Specifically, when a final judgment instruction sent by the video module is received, the feature module extracts a facial image in the monitoring video, analyzes human body feature data of a student and judges whether the student is in a tension state; and then, judging whether the students are in psychological crisis or not according to whether the students are in a nervous state or not and whether early warning is needed or not.

In the embodiment, the human body characteristic data comprises micro expression data and micro action data, and micro changes of the faces of the students in the monitoring video are input to the neural network for deep learning, so that the micro expression data is obtained; and amplifying the micro-motion of the students in the monitoring video and detecting the change rule of the micro-motion, thereby acquiring micro-motion data. If the time for a certain student to crumple the eyebrow is about 1/25 seconds in the micro expression data or the time for a certain student to close the fist is about 1/25 seconds in the micro action data, the student can be judged to be in a tension state, namely the student is in a psychological crisis. At this time, the characteristic module sends an instruction to the early warning module to carry out early warning.

In addition, the human body characteristic data in the embodiment further includes heart rate, blood oxygen concentration and blood pressure difference, the heart rate, the blood oxygen concentration and the blood pressure are closely related to psychological conditions and tension states and are not controlled by consciousness of people, and the internal heart conditions of the detected object can be objectively and accurately interpreted through the data. For example, the facial image area of the student is selected, and the change rule of the capillary blood vessel congestion during respiration is detected, so that the heart rate and the blood oxygen concentration of the detected object can be obtained; the blood pressure difference of the detected object can be obtained by detecting the phase difference of the blood flow velocity of the human body part, and the prior art can be specifically referred. After obtaining student's heart rate, blood oxygen concentration and blood pressure difference's data, judge one by one whether these three data lie in its standard interval, the standard interval can refer to medical data, say that normal people heart rate is generally 60 ~ 100 times/minute.

If the three data are all located in the standard interval, the student is judged not to be in a tension state, namely the student does not have a psychological crisis and does not need early warning; on the contrary, if more than one of the three data is not in the standard interval, the student is judged to be in a tension state, namely the student has a psychological crisis and needs to be warned.

And S5, early warning by an early warning module.

And after receiving the instruction, the early warning module carries out early warning. For example, the buzzer sounds "there is a situation in the classroom and the teacher is asked to quickly process" the situation.

Example 2

The difference from embodiment 1 is only that the detection object is classified according to the human body feature data after it is determined that the student has a psychological crisis. For example, if the student's heart rate is 140 beats/minute, which indicates that the fluctuation of his mood is large, he may be prone to violence, and it is defined as "serious" to prompt the teacher to pay attention. Meanwhile, the violent tendency is taken as the core characteristic of the student, and the follow-up teachers educate the student in a targeted manner, so that the overstimulation behavior of the student is reduced as much as possible.

Example 3

The difference from the embodiment 2 is only that the role recognition is performed on the student through the interactive inquiry, namely, whether the student is mild or aggressive is determined; wherein, mild students are cool and quiet in case of accidents, are not easy to excited, and have small emotional fluctuation; the radical students are motivated by accidents, are easy to excited and have large emotional fluctuation.

In this embodiment, the server is also provided with a microphone and FaceReader software, and asks students some questions by a guiding method, for example, "when seeing that a friend is beaten by a bady, should the friend rush to help, or alarm? "; meanwhile, the camera shoots the photos of the facial expressions of the students when answering the questions, the photos are sent to the server, and FaceReader software analyzes the facial expressions of the students.

FaceReader software can automatically analyze the facial expressions of students: if the answer of the student is 'alarm' and the facial expression is 'pleasure' or 'natural', the student is cool and quiet in meeting, not easy to excite, small in emotion fluctuation and mild; on the contrary, if the answer of the student is ' rush to help, and the facial expression is ' disgust ', ' anger ' or ' slight libel ', the student is motivated by a chance, is easy to excite, has large emotion fluctuation, and is an aggressive student, the student should be monitored in an important way so as to give an early warning in time.

The foregoing is merely an example of the present invention, and common general knowledge in the field of known specific structures and characteristics is not described herein in any greater extent than that known in the art at the filing date or prior to the priority date of the application, so that those skilled in the art can now appreciate that all of the above-described techniques in this field and have the ability to apply routine experimentation before this date can be combined with one or more of the present teachings to complete and implement the present invention, and that certain typical known structures or known methods do not pose any impediments to the implementation of the present invention by those skilled in the art. It should be noted that, for those skilled in the art, without departing from the structure of the present invention, several changes and modifications can be made, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicability of the patent. The scope of the claims of the present application shall be determined by the contents of the claims, and the description of the embodiments and the like in the specification shall be used to explain the contents of the claims.