A psychological counseling method and system based on AI technology

Abstract

The application relates to the technical field of psychological counseling, in particular to a psychological counseling method and system based on AI technology. The method comprises the following steps: obtaining target voice of a psychological counseling object and performing semantic keyword extraction; analyzing the semantic keywords, determining a dialogue theme and performing preliminary dialogue; tracking the degree of closure of the eyes of the psychological counseling object in real time to obtain a real-time eye closure degree value; comparing the real-time eye closure degree value with a preset eye closure degree value to obtain a comparison result, combining the comparison result and an analysis result to determine a real-time emotional state; according to the real-time emotional state and the dialogue theme, extracting a corresponding counseling scheme to perform psychological counseling. When performing psychological counseling on the psychological counseling object, the degree of closure of the eyes and the analysis result are combined to determine the emotional state of the psychological counseling object, a psychological counseling scheme corresponding to the emotion is extracted, and the psychological counseling object is subjected to psychological counseling, so that the accuracy of the psychological counseling is ensured.

Description

Technical Field

[0001] This application relates to the field of psychological counseling technology, and in particular to a psychological counseling method and system based on AI technology. Background Technology

[0002] With the advancement of artificial intelligence and big data technologies, psychological counseling is gradually extending to online platforms and mobile applications, utilizing intelligent algorithms to provide personalized psychological intervention plans. Simultaneously, interdisciplinary collaboration is increasingly strengthened, with psychological counseling integrating knowledge from medicine, sociology, education, and other fields to help individuals cope with psychological problems in a more comprehensive and systematic way.

[0003] Under current technological conditions, psychological counseling is typically conducted by analyzing the content of the dialogue between the individual and the counselor. However, this method is susceptible to subjectivity, which can lead to omissions in expression and inaccurate counseling. Summary of the Invention

[0004] This application provides a psychological counseling method and system based on AI technology to solve the above problems.

[0005] Firstly, this application provides a psychological counseling method based on AI technology, the method comprising:

[0006] Obtain the target speech input by the psychological counseling subject, and extract semantic keywords from the target speech;

[0007] Analyze the semantic keywords to obtain the semantic keyword analysis results, determine the dialogue topic, and conduct an initial dialogue.

[0008] The degree of eye closure of the psychological counseling subject is tracked in real time to obtain the real-time eye closure degree value;

[0009] The real-time eye closure degree value is compared with the preset eye closure degree value to obtain the comparison result. The real-time emotional state of the psychological counseling object is determined by combining the comparison result and the semantic keyword analysis result.

[0010] Based on the real-time emotional state of the person receiving psychological counseling and the topic of the conversation, a corresponding psychological counseling plan is extracted, and psychological counseling is provided to the person receiving psychological counseling.

[0011] By employing the aforementioned technical solutions, and through real-time acquisition and analysis of the speech of the individuals receiving psychological counseling, semantic keywords can be accurately extracted, and the dialogue topic can be quickly determined, making psychological counseling more targeted and effective. Simultaneously, the degree of eye closure of the individuals receiving psychological counseling is tracked in real time, which helps to gain a more comprehensive understanding of their emotional state and provides a more accurate basis for subsequent psychological counseling. Based on the real-time emotional state of the individuals receiving psychological counseling and the dialogue topic, appropriate psychological counseling plans can be flexibly extracted and applied, effectively helping the individuals adjust their emotions, ensuring the accuracy of psychological counseling, and thus achieving the expected psychological counseling effect.

[0012] Optionally, the real-time tracking of the degree of eye closure of the psychological counseling subject to obtain a real-time eye closure degree value includes:

[0013] The facial information of the psychological counseling recipients is monitored in real time to obtain monitoring results;

[0014] Analyzing the monitoring results, a specific image of the degree of eye closure of the psychological counseling subject was obtained;

[0015] By analyzing the specific images of the degree of eye closure, the real-time eye closure value is obtained.

[0016] By employing the aforementioned technical solution, real-time monitoring of the facial information of the person receiving psychological counseling can directly obtain specific images of the degree of eye closure. This non-contact monitoring method not only improves the convenience and accuracy of data collection but also enhances the comfort of the psychological counseling process. Further analysis of these images yields real-time eye closure values, providing objective and reliable data for accurately assessing the emotional state of the person receiving psychological counseling. This allows for a more comprehensive understanding of the emotional state of the person receiving psychological counseling, enabling the development of more precise and effective psychological counseling plans and improving the overall effectiveness of psychological counseling.

[0017] Optionally, the step of analyzing the specific image of the degree of eye closure to obtain the real-time eye closure value includes:

[0018] By analyzing the specific images of the degree of eye closure, the eyelid area, eye closure time, iris area, and eyelid area of ​​the psychological counseling subject can be obtained.

[0019] Based on the eyelid area, eyelid closure time, eyelid iris area, and eyelid area, the real-time eyelid closure degree value is obtained and calculated using the following formula:

[0020] ;

[0021] in, This indicates the real-time degree of eye closure. This indicates the area of ​​the double eyelids. This indicates the duration of eye closure. This represents the area of ​​the irises in both eyes. This indicates the area of ​​both eyelids.

[0022] By employing the aforementioned technical solutions and conducting in-depth analysis of specific images depicting the degree of eyelid closure, multiple key parameters, including eyelid area, closure time, iris area, and eyelid area, are precisely extracted. This provides a rich data foundation for accurately assessing the real-time eyelid closure degree of individuals receiving psychological counseling. Utilizing these refined physiological indicators, combined with specific calculation formulas, a more accurate and comprehensive value of the real-time eyelid closure degree can be obtained, thus reflecting the emotional state of the individuals receiving psychological counseling more objectively to a certain extent. This improves the targeting and effectiveness of psychological counseling.

[0023] Optionally, comparing the real-time eye closure degree value with a preset eye closure degree value to obtain a comparison result, and combining the comparison result with the semantic keyword analysis result to determine the real-time emotional state of the psychological counseling subject, includes:

[0024] Obtain historical psychological counseling data, and based on the historical psychological counseling data, analyze the degree of eye closure under different emotional states to obtain preset degree of eye closure under different emotional states;

[0025] The preset eye closure degree values ​​under different emotional states are matched with the real-time eye closure degree values ​​to obtain preliminary matching results;

[0026] The preliminary matching results are compared with the semantic keyword analysis results. If the comparison results are consistent, the real-time emotional state is determined based on the preliminary matching results.

[0027] By acquiring and analyzing historical psychological counseling data, the above technical solution obtains preset eyelid closure values ​​for different emotional states, providing a reliable reference benchmark for real-time emotional state recognition. Matching the real-time eyelid closure values ​​with the preset values ​​allows for a rapid preliminary assessment of the emotional state. Furthermore, comparing the preliminary matching results with semantic keyword analysis results increases the accuracy and reliability of emotional state recognition. This improves the precision and efficiency of emotional state recognition and provides more scientific and personalized support for psychological counseling and emotion management, contributing to enhanced targeting and effectiveness of psychological counseling.

[0028] Optionally, comparing the preliminary matching results with the semantic keyword analysis results to determine the real-time emotional state further includes:

[0029] If the comparison results are inconsistent, obtain the response content of the psychological counseling subject at the current moment;

[0030] Based on the content of the response, determine the type of psychological counseling;

[0031] Obtain the internal response speed, and determine the historical eye closure time of the psychological counseling subject based on the internal response speed;

[0032] Based on the type of psychological counseling, the internal response speed, and the historical eye closure time, the content of psychological counseling is adjusted to obtain the adjusted content of psychological counseling.

[0033] The adjusted psychological counseling content is used as the dialogue content for psychological counseling. The dialogue is conducted with the psychological counseling subject based on the dialogue content to ensure that the real-time degree of eye closure of the psychological counseling subject can be monitored within a preset time period to determine the real-time emotional state.

[0034] By employing the aforementioned technical solutions, and through real-time comparison of results and acquisition of the responses from those receiving psychological counseling, the targeted and timely nature of the dialogue can be ensured, making psychological counseling more closely aligned with the actual needs of the individuals. Determining the type of psychological counseling based on the responses helps in precise policy implementation, improving the efficiency and effectiveness of psychological counseling. Adjusting the content of psychological counseling ensures sufficient time to capture the state of the individual's eyes closing once. By monitoring the real-time degree of eye closure of the individual to determine their emotional state, it is possible to adjust psychological counseling strategies in a timely manner, ensuring the effectiveness and safety of psychological counseling.

[0035] Optionally, before obtaining the real-time eye closure degree value by tracking the eye closure degree of the psychological counseling subject in real time, the method further includes:

[0036] Analyze the specific image to obtain image analysis results, and divide the eyes of the psychological counseling subject into regions based on the image analysis results to obtain eye regions;

[0037] Feature extraction is performed on the eye region, and based on the extraction results, it is determined whether there are reflective points in the current eye region;

[0038] If present, the reflective points of each binocular region are collected to obtain the distribution characteristics of the reflective points of each binocular region;

[0039] Based on the distribution characteristics, the density of the reflective points is determined;

[0040] Based on the density of the reflected points, a method for eliminating them is determined;

[0041] The glare points in the bilateral regions are eliminated according to the glare point elimination method.

[0042] Through the above technical solution and image analysis technology, the visual regions of the subjects receiving psychological counseling can be accurately segmented, providing precise localization for subsequent reflection point detection. Feature extraction of the visual regions can effectively determine whether reflection points exist in the current visual regions, improving the accuracy and reliability of reflection point detection. By collecting the distribution characteristics of reflection points in each visual region, the density of reflection points can be accurately determined, providing a strong basis for selecting appropriate reflection point removal methods. Determining the removal method based on the density of reflection points and removing reflection points in the visual regions effectively improves image quality and enhances the accuracy and reliability of subsequent emotional state analysis.

[0043] Optionally, the step of extracting a corresponding emotional psychological counseling plan based on the real-time emotional state of the person receiving psychological counseling and the topic of the dialogue, and then providing psychological counseling to the person receiving psychological counseling, includes:

[0044] The gaze focus of the psychological counseling subject is tracked in real time to determine whether the gaze focus is abnormal. If the gaze focus is abnormal, the facial micro-expressions of the psychological counseling subject are tracked in real time to obtain real-time images of facial micro-expressions.

[0045] Based on the real-time images of facial micro-expressions, the degree of contraction and relaxation of the facial muscles of the psychological counseling subject is analyzed to obtain the analysis results of facial muscle contraction and relaxation.

[0046] Based on the analysis results of facial muscle contraction and relaxation, the first emotional state of the psychological counseling subject is determined;

[0047] By combining the first emotional state with the semantic keyword analysis results, the actual emotional state of the psychological counseling recipient is determined;

[0048] Based on the real-time emotional state of the person receiving psychological counseling and the topic of the conversation, a corresponding psychological counseling plan is extracted, and psychological counseling is provided to the person receiving psychological counseling.

[0049] By employing the aforementioned technical solutions, real-time tracking of the gaze focus of those receiving psychological counseling can promptly detect any abnormalities, providing crucial clues for subsequent emotional state analysis. Real-time tracking of facial micro-expressions captures subtle changes in facial muscles, further enhancing the accuracy and sensitivity of emotional state analysis. Combining facial muscle contraction and relaxation analysis with semantic keyword analysis results provides a more comprehensive understanding of the actual emotional state of those receiving psychological counseling, offering strong support for developing personalized counseling plans. Furthermore, by extracting corresponding counseling plans based on real-time emotional states and dialogue topics, counseling can be provided to those receiving psychological counseling in real time, effectively helping them adjust their emotional state, alleviate stress, or resolve psychological problems, thus improving the targeting and effectiveness of psychological counseling.

[0050] Optionally, the step of analyzing the semantic keywords to obtain the semantic keyword analysis results, determining the dialogue topic, and conducting a preliminary dialogue includes:

[0051] The keyword set was analyzed to obtain the keyword set analysis results;

[0052] Based on the analysis results of the keyword set, a preliminary dialogue topic was determined;

[0053] Based on the meaning expressed by each keyword and the degree of correlation between keywords, the preliminary emotional state of the psychological counseling subject is determined;

[0054] Based on the initial dialogue topic and the initial emotional state, determine the dialogue topic and conduct an initial dialogue.

[0055] By extracting emotional state keywords from the target speech, the current emotional focus of the person receiving psychological counseling can be quickly and accurately grasped, providing a foundation for subsequent analysis. Analyzing the keyword set not only reveals the meaning of the keywords themselves but also uncovers the correlations between them, contributing to a more comprehensive understanding of the emotional state of the person receiving psychological counseling. Determining the initial dialogue topic based on the keyword set analysis results ensures the relevance and focus of the dialogue, making psychological counseling more focused. Judging the initial emotional state through the meaning and correlation of the keywords provides a basis for developing personalized psychological counseling strategies. Combining the initial dialogue topic with the emotional state allows for better guidance of the person receiving psychological counseling to express and vent their emotions, thereby achieving the effect of relieving stress and resolving psychological problems.

[0056] Optionally, the step of extracting a corresponding emotional psychological counseling plan based on the real-time emotional state of the person receiving psychological counseling and the topic of the dialogue, and then providing psychological counseling to the person receiving psychological counseling, includes:

[0057] Obtain behavioral images of the psychological counseling subject, analyze the behavioral images, and obtain the frequency and amplitude of the psychological counseling subject's behavior;

[0058] By combining the frequency and amplitude of the behavior, the real-time degree of eye closure, and the semantic keyword analysis results, the real-time emotional state is determined, and a corresponding psychological counseling plan is extracted based on the real-time emotional state to provide psychological counseling to the target.

[0059] By employing the aforementioned technical solutions, real-time tracking of the gaze focus of those receiving psychological counseling can promptly detect any abnormalities, providing crucial clues for subsequent emotional state analysis. Real-time tracking of facial micro-expressions captures subtle changes in facial muscles, further enhancing the accuracy and sensitivity of emotional state analysis. Combining facial muscle contraction and relaxation analysis with semantic keyword analysis results provides a more comprehensive understanding of the actual emotional state of those receiving psychological counseling, offering strong support for developing personalized counseling plans. Furthermore, by extracting corresponding counseling plans based on real-time emotional states and dialogue topics, counseling can be provided to those receiving psychological counseling in real time, effectively helping them adjust their emotional state, alleviate stress, or resolve psychological problems, thus improving the targeting and effectiveness of psychological counseling.

[0060] Secondly, this application provides a psychological counseling system based on AI technology, the system comprising:

[0061] The keyword extraction module is used to acquire the target speech input by the psychological counseling subject and extract semantic keywords from the target speech;

[0062] The keyword analysis module is used to analyze the semantic keywords, obtain the semantic keyword analysis results, determine the dialogue topic, and conduct preliminary dialogue.

[0063] The degree value determination module is used to track the degree of eye closure of the psychological counseling subject in real time while the initial dialogue is being conducted, and to obtain the real-time eye closure degree value.

[0064] The emotional state determination module is used to compare the real-time eye closure degree value with the preset eye closure degree value, obtain the comparison result, and combine the comparison result with the semantic keyword analysis result to determine the real-time emotional state of the psychological counseling object.

[0065] The psychological counseling module is used to extract corresponding emotional psychological counseling plans based on the real-time emotional state of the person receiving psychological counseling and the topic of the dialogue, and to provide psychological counseling to the person receiving psychological counseling.

[0066] Optionally, the degree value determination module is specifically used for:

[0067] The facial information of the psychological counseling recipients is monitored in real time to obtain monitoring results;

[0068] Analyzing the monitoring results, a specific image of the degree of eye closure of the psychological counseling subject was obtained;

[0069] By analyzing the specific images of the degree of eye closure, the real-time eye closure value is obtained.

[0070] Optionally, the degree value determination module is specifically used for:

[0071] By analyzing the specific images of the degree of eye closure, the eyelid area, eye closure time, iris area, and eyelid area of ​​the psychological counseling subject can be obtained.

[0072] Based on the eyelid area, eyelid closure time, eyelid iris area, and eyelid area, the real-time eyelid closure degree value is obtained and calculated using the following formula:

[0073] ;

[0074] in, This indicates the real-time degree of eye closure. This indicates the area of ​​the double eyelids. This indicates the duration of eye closure. This represents the area of ​​the irises in both eyes. This indicates the area of ​​both eyelids.

[0075] Optionally, the emotional state determination module is specifically used to: acquire historical psychological counseling data, analyze the degree of eye closure under different emotional states based on the historical psychological counseling data, and obtain preset degree of eye closure under different emotional states;

[0076] The preset eye closure degree values ​​under different emotional states are matched with the real-time eye closure degree values ​​to obtain preliminary matching results;

[0077] The preliminary matching results are compared with the semantic keyword analysis results. If the comparison results are consistent, the real-time emotional state is determined based on the preliminary matching results.

[0078] Optionally, the AI-based psychological counseling system further includes a special emotion analysis module, specifically used for:

[0079] If the comparison results are inconsistent, obtain the response content of the psychological counseling subject at the current moment;

[0080] Based on the content of the response, determine the type of psychological counseling;

[0081] The response speed of the psychological counseling subject is obtained, and the historical eye closing time of the psychological counseling subject is determined based on the response speed.

[0082] Based on the type of psychological counseling, the response speed, and the historical eye closure time, the content of the psychological counseling is adjusted to obtain the adjusted psychological counseling content.

[0083] The adjusted psychological counseling content is used as the dialogue content for psychological counseling. The dialogue is conducted with the psychological counseling subject based on the dialogue content to ensure that the real-time degree of eye closure of the psychological counseling subject can be monitored within a preset time period to determine the real-time emotional state.

[0084] Optionally, the AI-based psychological counseling system further includes a reflective spot elimination module, specifically used for:

[0085] Analyze the specific image to obtain image analysis results, and divide the eyes of the psychological counseling subject into regions based on the image analysis results to obtain eye regions;

[0086] Feature extraction is performed on the eye region, and based on the extraction results, it is determined whether there are reflective points in the current eye region;

[0087] If present, the reflective points of each binocular region are collected to obtain the distribution characteristics of the reflective points of each binocular region;

[0088] Based on the distribution characteristics, the density of the reflective points is determined;

[0089] Based on the density of the reflected points, a method for eliminating them is determined;

[0090] The glare points in the bilateral regions are eliminated according to the glare point elimination method.

[0091] Optionally, the psychological counseling module is specifically used for:

[0092] The gaze focus of the psychological counseling subject is tracked in real time to determine whether the gaze focus is abnormal. If the gaze focus is abnormal, the facial micro-expressions of the psychological counseling subject are tracked in real time to obtain real-time images of facial micro-expressions.

[0093] Based on the real-time images of facial micro-expressions, the degree of contraction and relaxation of the facial muscles of the psychological counseling subject is analyzed to obtain the analysis results of facial muscle contraction and relaxation.

[0094] Based on the analysis results of facial muscle contraction and relaxation, the first emotional state of the psychological counseling subject is determined;

[0095] By combining the first emotional state with the semantic keyword analysis results, the actual emotional state of the psychological counseling recipient is determined;

[0096] Based on the real-time emotional state of the person receiving psychological counseling and the topic of the conversation, a corresponding psychological counseling plan is extracted, and psychological counseling is provided to the person receiving psychological counseling.

[0097] Optionally, the keyword analysis module is specifically used for:

[0098] Based on the target speech of the psychological counseling recipient, keywords related to emotional state are extracted from the target speech to obtain a keyword set;

[0099] The keyword set was analyzed to obtain the keyword set analysis results;

[0100] Based on the analysis results of the keyword set, a preliminary dialogue topic was determined;

[0101] Based on the meaning expressed by each keyword and the degree of correlation between keywords, the preliminary emotional state of the psychological counseling subject is determined;

[0102] Based on the initial dialogue topic and the initial emotional state, determine the dialogue topic and conduct an initial dialogue.

[0103] Optionally, the psychological counseling module is specifically used for:

[0104] Obtain behavioral images of the psychological counseling subject, analyze the behavioral images, and obtain the frequency and amplitude of the psychological counseling subject's behavior;

[0105] Analyze the behavioral images of the psychological counseling subject to obtain the degree of eye closure of the psychological counseling subject;

[0106] The real-time eye closure degree value is obtained based on the eye closure degree.

[0107] By combining the frequency and amplitude of the behavior, the real-time degree of eye closure, and the semantic keyword analysis results, the real-time emotional state is determined, and a corresponding psychological counseling plan is extracted based on the real-time emotional state to provide psychological counseling to the target. Attached Figure Description

[0108] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0109] Figure 1 This is a schematic diagram illustrating an application scenario provided in one embodiment of this application;

[0110] Figure 2 A flowchart illustrating an AI-based psychological counseling method provided in one embodiment of this application;

[0111] Figure 3 This is a schematic diagram of the structure of a psychological counseling system based on AI technology, provided as an embodiment of this application. Detailed Implementation

[0112] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0113] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article, unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship.

[0114] The embodiments of this application will now be described in further detail with reference to the accompanying drawings.

[0115] Under current technological conditions, psychological counseling is typically conducted by analyzing the content of the dialogue between the individual and the counselor. However, this method is susceptible to subjectivity, which can lead to omissions in expression and inaccurate counseling.

[0116] Based on this, this application provides a psychological counseling method and system based on AI technology. It acquires the target speech of the psychological counseling subject, extracts semantic keywords from the target speech, analyzes the speech keywords to obtain semantic keyword analysis results, determines the initial dialogue topic based on the semantic keywords, which is beneficial for adjusting the subsequent psychological counseling plan. While conducting the initial dialogue, it tracks the degree of eye closure of the psychological counseling subject in real time to obtain the real-time eye closure degree value. Combining the semantic keyword analysis results and the real-time eye closure degree value, it determines the real-time emotional state of the psychological counseling subject, and extracts the corresponding psychological counseling plan based on the real-time emotional state and dialogue topic to provide psychological counseling to the subject. This avoids the one-sidedness of relying solely on dialogue content for psychological counseling and ensures the accuracy of psychological counseling.

[0117] Figure 1 This application provides an illustration of an application scenario. When using AI technology for online psychological counseling, the method provided in this application is applied to improve the accuracy of psychological counseling and achieve better results.

[0118] Specifically, the method provided in this application is applied to any server. The server interacts with the psychological counseling object, extracts keywords and analyzes the keywords from the target voice of the psychological counseling object to obtain a keyword analysis result, determines a preliminary conversation topic, tracks the degree of eye closure of the psychological counseling object during the preliminary conversation to obtain a real-time eye closure degree value, determines the real-time emotional state according to the keyword analysis result and the real-time eye closure degree value, and extracts the corresponding psychological counseling plan according to the real-time emotional state and the conversation topic to conduct psychological counseling on the psychological counseling object. By combining the real-time emotional state, the current emotional state is accurately confirmed, and the corresponding psychological counseling plan is extracted according to the current emotional state for psychological counseling, ensuring the accuracy of psychological counseling, thereby effectively conducting psychological counseling and achieving the expected psychological counseling effect.

[0119] The specific implementation method can refer to the following embodiments.

[0120] Figure 2 The following is a flowchart of a psychological counseling method based on AI technology provided by an embodiment of this application. The method of this embodiment can be applied to the server in the above scenario. As Figure 2 shown, this method includes:

[0121] S201. Obtain the target voice input by the psychological counseling object, and extract semantic keywords from the target voice.

[0122] The target voice can be all the voices uttered by the psychological counseling object at the beginning of psychological counseling.

[0123] Semantic keywords can be all the key words in the target voice.

[0124] Specifically, use an audio acquisition device to collect the target voice to obtain the target voice of the psychological counseling object, perform noise removal and signal enhancement operations on the target voice of the psychological counseling object, convert the target voice signal after noise removal and signal enhancement operations into text and output it to obtain the text containing what the psychological counseling object said, identify the text of what was said, perform word segmentation and词性标注 (POS tagging) on the identified text to identify each word and its词性 in the sentence, remove irrelevant stop words (such as "的", "了", "在", etc.), retain the words with actual meaning, calculate the word frequency of each word, sort the word frequencies of each word according to the calculation results of the word frequencies of each word, and take the words with higher frequencies as the words with higher importance in the current text, that is, semantic keywords.

[0125] S202. Analyze the semantic keywords to obtain a semantic keyword analysis result, determine the conversation topic and conduct a preliminary conversation.

[0126] The topic of the dialogue can be the topic of the current psychological counseling session, obtained from semantic keywords.

[0127] Specifically, the extracted semantic keywords are converted into vector representations to obtain keyword vectors. These keyword vectors are then clustered (keywords with similar semantics are grouped together) to obtain clustering results. Topics are extracted from these clustering results, with each topic corresponding to a set of semantically related keywords. A topic name is automatically generated based on the keywords in each topic. Combined with the contextual information of the target speech, semantic analysis and sentiment analysis are performed on the keywords and topics to obtain semantic keyword analysis results. Based on the semantic keyword analysis results, the emotional intensity expressed by the psychological counseling recipient, the frequency of each semantic keyword, and contextual relevance are determined. Based on the determined emotional intensity expressed by the psychological counseling recipient, the frequency of each semantic keyword, and contextual relevance, different dialogue topics are obtained and extracted. The extracted topics are ranked to obtain a topic ranking result. Based on the topic ranking result, the most relevant topic is selected as the dialogue topic. An initial dialogue is conducted with the psychological counseling recipient based on the semantic keyword analysis results and the dialogue topic.

[0128] S203. Track the degree of eye closure of the psychological counseling subject in real time and obtain the real-time eye closure degree value.

[0129] The degree of eyelid closure can be measured by the area of ​​the eyelids, the time of eyelid closure, the area of ​​the iris, and the area of ​​the eyelids of the person receiving psychological counseling.

[0130] The real-time eye closure degree value can be a numerical value reflecting the real-time eye closure degree obtained from the eyelid area, eye closure time, eyelid iris area, and eyelid area.

[0131] Specifically, firstly, monitoring equipment can be used to acquire facial images and video streams of the subject receiving psychological counseling. Then, a computer vision library can be used to adjust the resolution and correct the color of the video stream. Afterward, a facial landmark detection algorithm can be used to locate the eye region on the face. Based on the facial video stream, the binocular closure time is obtained. From the located eye region, the area of ​​the eyelids, irises, and eyelids of the subject receiving psychological counseling is extracted. Based on the eyelid area, binocular closure time, iris area, and eyelid area, the real-time degree of eye closure is calculated.

[0132] S204. Compare the real-time eye closure degree value with the preset eye closure degree value to obtain the comparison result. Combine the comparison result with the semantic keyword analysis result to determine the real-time emotional state of the psychological counseling object.

[0133] The preset eye closure degree value can be extracted from historical data and is related to the eye closure degree value reflecting emotional state.

[0134] Real-time emotional state can refer to the emotional state of the person receiving psychological counseling during the counseling process.

[0135] Specifically, historical psychological counseling data is acquired, and the degree of eyelid closure under different emotional states is analyzed. Based on the analysis results, preset eyelid closure values ​​for different emotional states are calculated. The real-time eyelid closure values ​​are compared with the preset eyelid closure values ​​to obtain the comparison results. Combining the comparison results with the semantic keyword analysis results mentioned above, the real-time emotional state of the psychological counseling recipient is determined.

[0136] S205. Based on the real-time emotional state and dialogue topic of the person receiving psychological counseling, extract the corresponding psychological counseling plan and provide psychological counseling to the person receiving psychological counseling.

[0137] Specifically, all possible emotional states obtained above are categorized (e.g., "relaxed," "tense," "drowsy," "anxious"), and all possible dialogue topics are categorized (e.g., work stress, family conflicts), resulting in emotional state categorization and dialogue topic categorization results. Based on these results, a psychological counseling solution database is established, containing different emotional states and dialogue topics. Each solution in the database includes a series of applicable counseling methods, techniques, and scripts. The counseling solutions are categorized and tagged according to emotional state and dialogue topic for easy retrieval.

[0138] Based on the emotional state and conversation topic of the current psychological counseling recipient, a matching psychological counseling plan is retrieved and extracted from the psychological counseling plan database, and then psychological counseling is provided to the recipient.

[0139] The method provided in this embodiment, by acquiring and analyzing the voice of the person receiving psychological counseling in real time, can accurately extract semantic keywords and quickly determine the topic of the conversation, making psychological counseling more targeted and effective. During the initial conversation, the degree of eye closure of the person receiving psychological counseling is tracked in real time, which helps to understand the person's emotional state more comprehensively and provides a more accurate basis for subsequent psychological counseling. Based on the real-time emotional state of the person receiving psychological counseling and the topic of the conversation, appropriate psychological counseling plans can be flexibly extracted and applied, which can effectively help the person receiving psychological counseling adjust their emotions, ensure the accuracy of psychological counseling, and thus achieve the expected psychological counseling effect.

[0140] In some embodiments, the facial information of the psychological counseling subject is monitored in real time to obtain monitoring results; the monitoring results are analyzed to obtain a specific image of the degree of eye closure of the psychological counseling subject; the specific image of the degree of eye closure is analyzed to obtain a real-time eye closure value.

[0141] The monitoring results can be the results of monitoring all facial features of the person receiving psychological counseling.

[0142] The specific image could be an image showing the degree to which the eyes of the person receiving psychological counseling are closed.

[0143] Specifically, monitoring equipment is used to acquire real-time facial videos of the individuals receiving psychological counseling. The video stream undergoes resolution adjustment and color correction. A pre-trained facial detection model can be used to detect facial regions within video frames, extracting sub-images containing the face to obtain monitoring results. Based on these results, within the detected facial regions, an eye detection algorithm can be used to detect the position and area of ​​the eyes, obtaining a specific image of the degree of eyelid closure. From this image, the eyelid area, closure time, iris area, and eyelid area are extracted. Based on these parameters, the real-time eyelid closure degree value is calculated.

[0144] The method provided in this embodiment allows for real-time monitoring of the facial information of the person receiving psychological counseling, directly obtaining images of the degree of eye closure. This non-contact monitoring method not only improves the convenience and accuracy of data collection but also enhances the comfort of the psychological counseling process. Further analysis of these images yields real-time eye closure values, providing objective and reliable data for accurately assessing the emotional state of the person receiving psychological counseling. This allows for a more comprehensive understanding of the emotional state of the person receiving psychological counseling, enabling the development of more precise and effective psychological counseling plans and improving the overall effectiveness of psychological counseling.

[0145] In some embodiments, specific images of the degree of eye closure are analyzed to obtain the eyelid area, eye closure time, eyelid iris area, and eyelid area of ​​the psychological counseling subject; based on the eyelid area, eye closure time, eyelid iris area, and eyelid area, the real-time degree of eye closure is obtained and calculated using the following formula (1):

[0146] (1)

[0147] in, This indicates the real-time degree of eye closure. Indicates the area of ​​both eyelids. Indicates the duration of eye closure. Indicates the area of ​​the iris in both eyes. This indicates the area of ​​both eyelids.

[0148] The area of ​​a double eyelid can be the surface area of ​​the eyelid, including the upper and lower eyelids.

[0149] The duration of eye closure is the time it takes for the eyes to close completely during a single blink.

[0150] The iris area of ​​both eyes can be the surface area of ​​the iris of the person receiving psychological counseling.

[0151] The area of ​​both eyelids can be the surface area of ​​the eyelids of the person receiving psychological counseling, including the total area of ​​the upper and lower eyelids.

[0152] Specifically, image processing technology can be used to detect the eyelid area and calculate the area of ​​both eyelids. The binocular closure time can be measured using the timestamps of video frames. The area of ​​both irises can be obtained by detecting the iris region and calculating it. The area of ​​both eyelids can be calculated by detecting the eyelid region. Since the time it takes for the eyes to close is closely related to the area of ​​the eyelids, and the area of ​​the eyelids has a certain influence on the degree of eye closure, and the area of ​​the iris is also closely related to the area of ​​the eyelids, the real-time degree of eye closure can be calculated using formula (1) to provide an accurate basis for the extraction of subsequent psychological counseling plans.

[0153] The method provided in this embodiment, through in-depth analysis of specific images of the degree of eyelid closure, accurately extracts multiple key parameters such as eyelid area, eyelid closure time, eyelid iris area, and eyelid area, providing a rich data foundation for accurately assessing the real-time degree of eyelid closure in psychological counseling subjects. Using these refined physiological indicators, combined with specific calculation formulas, a more accurate and comprehensive value of the real-time degree of eyelid closure can be obtained, thus reflecting the emotional state of the psychological counseling subject more objectively to a certain extent. This improves the targeting and effectiveness of psychological counseling.

[0154] In some embodiments, historical psychological counseling data is acquired, and based on the historical psychological counseling data, the degree of eye closure under different emotional states is analyzed to obtain preset eye closure values ​​under different emotional states; the preset eye closure values ​​under different emotional states are matched with the real-time eye closure values ​​to obtain preliminary matching results; the preliminary matching results are compared with the semantic keyword analysis results, and if the comparison results are consistent, the real-time emotional state is determined based on the preliminary matching results.

[0155] Historical psychological counseling data can be obtained by searching relevant historical documents on psychological counseling.

[0156] Specifically, a large amount of historical psychological counseling data was retrieved, and the data was cleaned and noise removed to ensure data quality. The historical psychological counseling data was categorized into emotional states (e.g., "relaxed," "tense," "drowsy," "anxious," etc.). For each emotional state, the distribution of eyelid closure values ​​was statistically analyzed, and the mean and standard deviation of the eyelid closure values ​​were calculated to obtain the preset closure value for that emotional state. The eyelid closure values ​​of the psychological counseling subjects were monitored in real time via video stream, and the real-time eyelid closure values ​​were calculated. The real-time eyelid closure values ​​were matched with the preset closure values ​​for different emotional states, and the similarity was calculated. Based on the similarity calculation results, the emotional state with the highest similarity was used as the preliminary matching result. The preliminary matching result was compared with the semantic keyword analysis results that can express emotional states. If the comparison results were consistent, the real-time emotional state was determined based on the preliminary matching result.

[0157] The method provided in this embodiment acquires and analyzes historical psychological counseling data to obtain preset eyelid closure values ​​for different emotional states, providing a reliable reference benchmark for real-time emotional state identification. By matching the real-time eyelid closure values ​​with the preset eyelid closure values, a preliminary emotional state judgment can be quickly obtained. Furthermore, comparing the preliminary matching results with semantic keyword analysis results increases the accuracy and reliability of emotional state identification. This improves the precision and efficiency of emotional state identification and provides more scientific and personalized support for psychological counseling and emotion management, helping to enhance the targeting and effectiveness of psychological counseling.

[0158] In some embodiments, if the comparison results are inconsistent, the response content of the psychological counseling subject at the current moment is obtained; the psychological counseling type is determined based on the response content; the internal response speed is obtained, and the historical eye closure time of the psychological counseling subject is determined based on the internal response speed; the psychological counseling content is adjusted based on the psychological counseling type, internal response speed, and historical eye closure time to obtain the adjusted psychological counseling content; the adjusted psychological counseling content is used as the dialogue content for psychological counseling, and a dialogue is conducted with the psychological counseling subject based on the dialogue content to ensure that the real-time eye closure degree of the psychological counseling subject can be monitored within a preset time period to determine the real-time emotional state.

[0159] Internal response speed can be defined as the rate at which AI outputs language per unit of time when providing psychological counseling.

[0160] The preset time period can be the time it takes for the eyes to close once during historical psychological counseling.

[0161] The response can include any content provided by the person receiving psychological counseling.

[0162] Psychological counseling can be categorized by the emotional state and the type of problem of the person receiving counseling.

[0163] The content of psychological counseling can include the form of counseling for the person receiving psychological counseling and the content of the dialogue.

[0164] Specifically, audio acquisition equipment is used to collect the content and internal response speed of the psychological counseling subject at the current moment. Noise removal and signal enhancement preprocessing are performed on the collected response content to obtain preprocessed response content. The type of psychological counseling is determined based on the preprocessed response content. The historical eye closure time of the psychological counseling subject is determined based on the internal response speed and dialogue content. Based on the determined psychological counseling type, internal response speed, and historical eye closure time, the psychological counseling content is adjusted. Based on the adjusted psychological counseling content, a dialogue is conducted with the psychological counseling subject to ensure sufficient time to track the real-time degree of eye closure and calculate the real-time eye closure degree value to determine the real-time emotional state.

[0165] By comparing results in real time and obtaining the responses of the individuals receiving psychological counseling, the targeted and timely nature of the dialogue can be ensured, making psychological counseling more closely aligned with the actual needs of the individuals. Determining the type of psychological counseling based on the responses helps in precise policy implementation, improving the efficiency and effectiveness of psychological counseling. Adjusting the content of psychological counseling ensures sufficient time to capture the state of the individuals' eyes closing once. By monitoring the real-time degree of eye closure of the individuals receiving psychological counseling to determine their emotional state, it is helpful to adjust psychological counseling strategies in a timely manner, ensuring the effectiveness and safety of psychological counseling.

[0166] In some embodiments, a specific image is analyzed to obtain image analysis results. Based on the image analysis results, the eyes of the psychological counseling subject are divided into regions to obtain eye regions. Feature extraction is performed on the eye regions. Based on the extraction results, it is determined whether there are reflective points in the current eye regions. If they exist, reflective points in each eye region are collected to obtain the distribution characteristics of reflective points in each eye region. Based on the distribution characteristics, the density of reflective points is determined. Based on the density, a reflective point elimination method is determined. The reflective points in the eye regions are eliminated according to the reflective point elimination method.

[0167] Specifically, the current frame is obtained from a specific image. Noise removal and grayscale conversion are performed on the image. Facial detection algorithms can be used to detect facial regions. Within the detected facial regions, eye detection algorithms are first used to locate and segment the eye region. Then, feature point detection algorithms are used to extract key points from the eye region. Based on these key points, a sub-image containing the eyes is cropped. Extraction is then performed on the cropped sub-image to obtain the extraction result. Within the eye region, bright spot detection algorithms are used to identify potential reflective points in the extraction result. The specific location of the reflective points is determined based on brightness and positional features. The number and location of reflective points within each eye region are counted to obtain the distribution characteristics of the reflective points. Based on the distribution characteristics, the density of the reflective points is calculated. An appropriate removal method (such as image filtering, reflective point filling, image reconstruction, etc.) is selected based on the density. The selected removal method is then applied to the reflective points within each eye region to ensure the naturalness and integrity of the image after reflective point removal.

[0168] The method provided in this embodiment, through image analysis technology, can accurately delineate the binocular regions of the psychological counseling subject, providing precise localization for subsequent reflection point detection; feature extraction of the eye regions can effectively determine whether reflection points exist in the current eye regions, improving the accuracy and reliability of reflection point detection; by collecting the distribution characteristics of reflection points in each binocular region, the density of reflection points can be accurately determined, providing a strong basis for selecting an appropriate reflection point removal method; determining the removal method based on the density of reflection points and removing reflection points in the binocular regions effectively improves image quality and enhances the accuracy and reliability of subsequent emotional state analysis.

[0169] In some embodiments, the gaze focus of the person receiving psychological counseling is tracked in real time to determine if the gaze focus is abnormal. If the gaze focus is abnormal, the facial micro-expressions of the person receiving psychological counseling are tracked in real time to obtain real-time images of facial micro-expressions. Based on the real-time images of facial micro-expressions, the degree of contraction and relaxation of the facial muscles of the person receiving psychological counseling is analyzed to obtain the analysis results of facial muscle contraction and relaxation. Based on the analysis results of facial muscle contraction and relaxation, the first emotional state of the person receiving psychological counseling is determined. Combining the first emotional state with the semantic keyword analysis results, the actual emotional state of the person receiving psychological counseling is determined. Based on the real-time emotional state of the person receiving psychological counseling and the dialogue topic, a corresponding psychological counseling plan is extracted, and psychological counseling is provided to the person receiving psychological counseling.

[0170] The degree to which the subject of psychological counseling focuses their gaze can be measured by the intensity of that focus.

[0171] Facial micro-expressions can be the state of activity of facial organs and tissues.

[0172] Specifically, monitoring equipment is used to acquire real-time facial video streams of the individuals receiving psychological counseling. Eye-tracking algorithms can be used to detect and track their gaze in real time, analyzing the gaze data to determine if there are any abnormalities (such as prolonged gaze deviation or wandering). If the gaze focus is abnormal, it indicates that judging the real-time emotional state of the individual solely by the degree of eye closure may be inaccurate. Therefore, real-time tracking of the individual's facial micro-expressions is necessary to obtain real-time images of these expressions. Based on these images, the movement and changes of facial feature points are calculated to assess muscle state and obtain analysis results of facial muscle contraction and relaxation. Based on the analysis results of facial muscle contraction and relaxation, emotion recognition algorithms (such as machine learning models) can be used to determine the individual's initial emotional state. Semantic keywords extracted during the counseling process are then combined to analyze the dialogue content. Combining the initial emotional state with the semantic keyword analysis results, the actual emotional state of the individual is determined. Based on the actual emotional state and the dialogue topic, a corresponding counseling plan is extracted from a pre-set counseling plan library to provide psychological counseling to the individual.

[0173] The method provided in this embodiment, by tracking the gaze focus of the person receiving psychological counseling in real time, can promptly detect any abnormalities in their gaze, providing important clues for subsequent emotional state analysis. Real-time tracking of facial micro-expressions captures subtle changes in facial muscles, further improving the accuracy and sensitivity of emotional state analysis. This avoids the inability to accurately determine the degree of eye closure when abnormal gaze is present, thus preventing inaccurate psychological counseling. Combining facial muscle contraction and relaxation analysis results with semantic keyword analysis results allows for a more comprehensive understanding of the actual emotional state of the person receiving psychological counseling, providing strong support for developing personalized psychological counseling plans. Based on the real-time emotional state and dialogue topic, corresponding psychological counseling plans are extracted for real-time guidance, effectively helping the person adjust their emotional state, relieve stress, or resolve psychological problems, thus improving the targeting and effectiveness of psychological counseling.

[0174] In some embodiments, based on the target speech of the psychological counseling subject, keywords related to emotional state are extracted from the target speech to obtain a keyword set; the keyword set is analyzed to obtain keyword set analysis results; based on the keyword set analysis results, a preliminary dialogue topic is determined; based on the meaning expressed by each keyword and the correlation between keywords, the preliminary emotional state of the psychological counseling subject is judged; based on the preliminary dialogue topic and the preliminary emotional state, a dialogue topic is determined and a preliminary dialogue is conducted.

[0175] Specifically, speech recognition technology can be used to convert the speech of the person receiving psychological counseling into text. First, natural language processing technology can be used to extract keywords related to emotional states from the transcribed text. These extracted keywords are then grouped together to form a keyword set. Next, sentiment analysis algorithms can be used to analyze the sentiment distribution of the keyword set. Based on the sentiment and content of the keywords, they are categorized. Based on the categorization results, a preliminary dialogue topic is inferred. The correlation between keywords is analyzed to determine their logical relationships and emotional connections. Based on the sentiment and correlation of the keywords, the preliminary emotional state of the person receiving psychological counseling is determined. Based on the preliminary dialogue topic and the preliminary emotional state, the dialogue topic is finalized, and an initial dialogue is conducted.

[0176] By extracting emotional state keywords from the target speech, this embodiment enables a quick and accurate grasp of the current emotional focus of the person receiving psychological counseling, providing a foundation for subsequent analysis. Analyzing the keyword set not only reveals the meaning of the keywords themselves but also uncovers the correlations between them, facilitating a more comprehensive understanding of the emotional state of the person receiving psychological counseling. Determining the initial dialogue topic based on the keyword set analysis results ensures the relevance and focus of the dialogue, making psychological counseling more targeted. Judging the initial emotional state through the meaning and correlation of the keywords provides a basis for developing personalized psychological counseling strategies. Combining the initial dialogue topic with the emotional state allows for better guidance of the person receiving psychological counseling to express and vent their emotions, thereby achieving the effect of relieving stress and resolving psychological problems.

[0177] In some embodiments, behavioral images of the psychological counseling subject are acquired, the behavioral images are analyzed to obtain the frequency and amplitude of the psychological counseling subject's behavior, and the real-time emotional state is determined by combining the behavioral frequency and amplitude, the real-time eye closure degree value and semantic keyword analysis results. Based on the real-time emotional state, a corresponding psychological counseling plan is extracted and psychological counseling is provided to the psychological counseling subject.

[0178] Behavioral images can be images corresponding to the behaviors of the person receiving psychological counseling.

[0179] Behavioral frequency can be the repetition frequency of several behavioral actions of the person receiving psychological counseling.

[0180] The range of behavior can refer to the intensity of several actions performed by the person receiving psychological counseling.

[0181] Specifically, video equipment is used to capture real-time videos of the behavior of those receiving psychological counseling. Continuous image frames are extracted from the video stream for analysis. First, behavior recognition algorithms are used to analyze the behavior in the images, identifying specific actions and statistically analyzing the frequency and amplitude of these actions. Then, sentiment analysis algorithms are used, combining the frequency and amplitude of the behavior, the real-time eye closure degree, and semantic keyword analysis results to analyze emotional tendencies. This yields the real-time emotional state of the individuals receiving psychological counseling.

[0182] Based on the real-time emotional state, the system extracts the corresponding counseling plan from the preset counseling plan library. Depending on the specific situation, the details of the counseling plan are adjusted to adapt to the current emotions and needs of the person receiving counseling. The enriched counseling content is then used to engage in dialogue with the person receiving counseling and to provide counseling.

[0183] By acquiring and analyzing behavioral images of the individuals receiving psychological counseling, the frequency and amplitude of their behavior can be objectively and accurately captured, providing rich data support for real-time judgment of their emotional state. Combining behavioral image analysis results, semantic keyword analysis results, and real-time eye closure values, a comprehensive assessment of the emotional state of the individuals receiving psychological counseling is achieved from multiple dimensions, including behavior, expression, and language, improving the accuracy and real-time nature of emotion recognition. Based on the real-time emotional state, corresponding psychological counseling plans are extracted, ensuring the targeted and timely nature of the counseling, and effectively alleviating the negative emotions of the individuals receiving psychological counseling. The comprehensive use of image processing, semantic analysis, and other technologies provides a more scientific and comprehensive method for psychological counseling, improving its effectiveness and efficiency.

[0184] Figure 3 This application provides a schematic diagram of the structure of an AI-based psychological counseling system according to an embodiment of the present application. Figure 3 As shown, the AI-based psychological counseling system 300 in this embodiment includes: a keyword extraction module 301, a keyword analysis module 302, a degree value determination module 303, an emotional state determination module 304, and a psychological counseling module 305.

[0185] The keyword extraction module 301 is used to acquire the target speech input by the psychological counseling subject and extract semantic keywords from the target speech;

[0186] Keyword analysis module 302 is used to analyze the semantic keywords, obtain the semantic keyword analysis results, determine the dialogue topic, and conduct preliminary dialogue.

[0187] The degree value determination module 303 is used to track the degree of eye closure of the psychological counseling subject in real time while the initial dialogue is being conducted, and to obtain the real-time eye closure degree value.

[0188] The emotional state determination module 304 is used to compare the real-time eye closure degree value with the preset eye closure degree value, obtain the comparison result, and combine the comparison result with the semantic keyword analysis result to determine the real-time emotional state of the psychological counseling object.

[0189] The psychological counseling module 305 is used to extract a corresponding psychological counseling plan based on the real-time emotional state of the psychological counseling object and the topic of the dialogue, and to provide psychological counseling to the psychological counseling object.

[0190] Optionally, the degree value determination module 303 is specifically used for:

[0191] The facial information of the psychological counseling recipients is monitored in real time to obtain monitoring results;

[0192] Analyzing the monitoring results, a specific image of the degree of eye closure of the psychological counseling subject was obtained;

[0193] By analyzing the specific images of the degree of eye closure, the real-time eye closure value is obtained.

[0194] Optionally, the degree value determination module 303 is specifically used for:

[0195] By analyzing the specific images of the degree of eye closure, the eyelid area, eye closure time, iris area, and eyelid area of ​​the psychological counseling subject can be obtained.

[0196] Based on the eyelid area, eyelid closure time, eyelid iris area, and eyelid area, the real-time eyelid closure degree value is obtained and calculated using the following formula:

[0197] ;

[0198] in, This indicates the real-time degree of eye closure. This indicates the area of ​​the double eyelids. This indicates the duration of eye closure. This represents the area of ​​the irises in both eyes. This indicates the area of ​​both eyelids.

[0199] Optionally, the emotional state determination module 304 is specifically used to: acquire historical psychological counseling data, analyze the degree of eye closure under different emotional states based on the historical psychological counseling data, and obtain preset degree of eye closure under different emotional states;

[0200] The preset eye closure degree values ​​under different emotional states are matched with the real-time eye closure degree values ​​to obtain preliminary matching results;

[0201] The preliminary matching results are compared with the semantic keyword analysis results. If the comparison results are consistent, the real-time emotional state is determined based on the preliminary matching results.

[0202] Optionally, the AI-based psychological counseling system 300 further includes a special emotion analysis module 306, specifically used for:

[0203] If the comparison results are inconsistent, obtain the response content of the psychological counseling subject at the current moment;

[0204] Based on the content of the response, determine the type of psychological counseling;

[0205] The internal response speed is obtained, and the historical eye closure time of the psychological counseling subject is determined based on the internal response speed.

[0206] Based on the type of psychological counseling, the internal response speed, and the historical eye closure time, the content of psychological counseling is adjusted to obtain the adjusted content of psychological counseling.

[0207] The adjusted psychological counseling content is used as the dialogue content for psychological counseling. The dialogue is conducted with the psychological counseling subject based on the dialogue content to ensure that the real-time degree of eye closure of the psychological counseling subject can be monitored within a preset time period to determine the real-time emotional state.

[0208] Optionally, the AI-based psychological counseling system 300 further includes a reflective spot elimination module 307, specifically used for:

[0209] Analyze the specific image to obtain image analysis results, and divide the eyes of the psychological counseling subject into regions based on the image analysis results to obtain eye regions;

[0210] Feature extraction is performed on the eye region, and based on the extraction results, it is determined whether there are reflective points in the current eye region;

[0211] If present, the reflective points of each binocular region are collected to obtain the distribution characteristics of the reflective points of each binocular region;

[0212] Based on the distribution characteristics, the density of the reflective points is determined;

[0213] Based on the density of the reflected points, a method for eliminating them is determined;

[0214] The glare points in the bilateral regions are eliminated according to the glare point elimination method.

[0215] Optionally, the psychological counseling module 305 is specifically used for:

[0216] The gaze focus of the psychological counseling subject is tracked in real time to determine whether the gaze focus is abnormal. If the gaze focus is abnormal, the facial micro-expressions of the psychological counseling subject are tracked in real time to obtain real-time images of facial micro-expressions.

[0217] Based on the real-time images of facial micro-expressions, the degree of contraction and relaxation of the facial muscles of the psychological counseling subject is analyzed to obtain the analysis results of facial muscle contraction and relaxation.

[0218] Based on the analysis results of facial muscle contraction and relaxation, the first emotional state of the psychological counseling subject is determined;

[0219] By combining the first emotional state with the semantic keyword analysis results, the actual emotional state of the psychological counseling recipient is determined;

[0220] Based on the real-time emotional state of the person receiving psychological counseling and the topic of the conversation, a corresponding psychological counseling plan is extracted, and psychological counseling is provided to the person receiving psychological counseling.

[0221] Optionally, the keyword analysis module 302 is specifically used for:

[0222] Based on the target speech of the psychological counseling recipient, keywords related to emotional state are extracted from the target speech to obtain a keyword set;

[0223] The keyword set was analyzed to obtain the keyword set analysis results;

[0224] Based on the analysis results of the keyword set, a preliminary dialogue topic was determined;

[0225] Based on the meaning expressed by each keyword and the degree of correlation between keywords, the preliminary emotional state of the psychological counseling subject is determined;

[0226] Based on the initial dialogue topic and the initial emotional state, determine the dialogue topic and conduct an initial dialogue.

[0227] Optionally, the psychological counseling module 305 is specifically used for:

[0228] Obtain behavioral images of the psychological counseling subject, analyze the behavioral images, and obtain the frequency and amplitude of the psychological counseling subject's behavior;

[0229] By combining the frequency and amplitude of the behavior, the real-time degree of eye closure, and the semantic keyword analysis results, the real-time emotional state is determined, and a corresponding psychological counseling plan is extracted based on the real-time emotional state to provide psychological counseling to the target.

[0230] The system in this embodiment can be used to execute the methods of any of the above embodiments, and its implementation principle and technical effect are similar, so they will not be described again here.

Claims

1. A psychological counseling method based on AI technology, characterized in that, include: Obtain the target speech input by the psychological counseling subject, and extract semantic keywords from the target speech; Analyze the semantic keywords to obtain semantic keyword analysis results, determine the dialogue topic, and conduct an initial dialogue. The degree of eye closure of the psychological counseling subject is tracked in real time to obtain the real-time eye closure degree value; The real-time eye closure degree value is compared with the preset eye closure degree value to obtain the comparison result. The real-time emotional state of the psychological counseling object is determined by combining the comparison result and the semantic keyword analysis result. Based on the real-time emotional state of the person receiving psychological counseling and the topic of the conversation, a corresponding psychological counseling plan is extracted, and psychological counseling is provided to the person receiving psychological counseling. The step of comparing the real-time eye closure degree value with a preset eye closure degree value to obtain a comparison result, and combining the comparison result with the semantic keyword analysis result to determine the real-time emotional state of the psychological counseling subject, includes: Obtain historical psychological counseling data, and based on the historical psychological counseling data, analyze the degree of eye closure under different emotional states to obtain preset degree of eye closure under different emotional states; The preset eye closure degree values ​​under different emotional states are matched with the real-time eye closure degree values ​​to obtain preliminary matching results; The preliminary matching results are compared with the semantic keyword analysis results. If the comparison results are consistent, the real-time emotional state is determined based on the preliminary matching results. The step of comparing the preliminary matching result with the semantic keyword analysis result to determine the real-time emotional state further includes: If the comparison results are inconsistent, obtain the response content of the psychological counseling subject at the current moment; Based on the content of the response, determine the type of psychological counseling; Obtain the internal response speed, and determine the historical eye closure time of the psychological counseling subject based on the internal response speed; Based on the type of psychological counseling, the internal response speed, and the historical eye closure time, the content of psychological counseling is adjusted to obtain the adjusted content of psychological counseling. The adjusted psychological counseling content is used as the dialogue content for psychological counseling. The dialogue is conducted with the psychological counseling subject to ensure that the real-time degree of eye closure of the psychological counseling subject can be monitored within a preset time period to determine the real-time emotional state. The step of extracting a corresponding emotional counseling plan based on the real-time emotional state of the person receiving psychological counseling and the topic of the conversation, and then providing psychological counseling to the person receiving psychological counseling, includes: The gaze focus of the psychological counseling subject is tracked in real time to determine whether the gaze focus is abnormal. If the gaze focus is abnormal, the facial micro-expressions of the psychological counseling subject are tracked in real time to obtain real-time images of facial micro-expressions. Based on the real-time images of facial micro-expressions, the degree of contraction and relaxation of the facial muscles of the psychological counseling subject is analyzed to obtain the analysis results of facial muscle contraction and relaxation. Based on the analysis results of facial muscle contraction and relaxation, the first emotional state of the psychological counseling subject is determined; By combining the first emotional state with the semantic keyword analysis results, the actual emotional state of the psychological counseling recipient is determined; Based on the real-time emotional state of the person receiving psychological counseling and the topic of the conversation, a corresponding psychological counseling plan is extracted, and psychological counseling is provided to the person receiving psychological counseling.

2. The method according to claim 1, characterized in that, The real-time tracking of the degree of eye closure of the person receiving psychological counseling, to obtain a real-time eye closure degree value, includes: The facial information of the psychological counseling recipients is monitored in real time to obtain monitoring results; Analyzing the monitoring results, a specific image of the degree of eye closure of the psychological counseling subject was obtained; By analyzing the specific images of the degree of eye closure, the real-time eye closure value is obtained.

3. The method according to claim 2, characterized in that, The analysis of the specific image of the degree of eye closure to obtain the real-time eye closure value includes: By analyzing the specific images of the degree of eye closure, the eyelid area, eye closure time, iris area, and eyelid area of ​​the psychological counseling subject can be obtained. Based on the eyelid area, eyelid closure time, eyelid iris area, and eyelid area, the real-time eyelid closure degree value is obtained and calculated using the following formula: ； Wherein, C represents the real-time binocular closure degree value. This indicates the area of ​​the double eyelids. This indicates the duration of eye closure. This represents the area of ​​the irises of both eyes. This indicates the area of ​​both eyelids.

4. The method according to claim 2, characterized in that, Before obtaining the real-time eye closure degree value by tracking the eye closure degree during the psychological counseling, the method further includes: Analyze the specific image to obtain image analysis results, and divide the eyes of the psychological counseling subject into regions based on the image analysis results to obtain eye regions; Feature extraction is performed on the eye region, and based on the extraction results, it is determined whether there are reflective points in the current eye region; If present, the reflective points of each binocular region are collected to obtain the distribution characteristics of the reflective points of each binocular region; Based on the distribution characteristics, the density of the reflective points is determined; Based on the density of the reflected points, a method for eliminating them is determined; The glare points in the bilateral regions are eliminated according to the glare point elimination method.

5. The method according to claim 1, characterized in that, The process of analyzing the semantic keywords, obtaining the semantic keyword analysis results, determining the dialogue topic, and conducting preliminary dialogue includes: Analyze the semantic keywords to identify keywords related to emotional states, and then group these keywords together to obtain a keyword set. The keyword set was analyzed to obtain the keyword set analysis results; Based on the analysis results of the keyword set, a preliminary dialogue topic was determined; Based on the meaning expressed by each keyword and the degree of correlation between keywords, the preliminary emotional state of the psychological counseling subject is determined; Based on the initial dialogue topic and the initial emotional state, determine the dialogue topic and conduct an initial dialogue.

6. The method according to claim 1, characterized in that, The step of extracting a corresponding emotional counseling plan based on the real-time emotional state of the person receiving psychological counseling and the topic of the conversation, and then providing psychological counseling to the person receiving psychological counseling, includes: Obtain behavioral images of the psychological counseling subject, analyze the behavioral images, and obtain the frequency and amplitude of the psychological counseling subject's behavior; By combining the frequency and amplitude of the behavior, the real-time degree of eye closure, and the semantic keyword analysis results, the real-time emotional state is determined, and a corresponding psychological counseling plan is extracted based on the real-time emotional state to provide psychological counseling to the target.

7. A psychological counseling system based on AI technology, characterized in that, The method applied to any one of claims 1-6 includes: The keyword extraction module is used to acquire the target speech input by the psychological counseling subject and extract semantic keywords from the target speech; The keyword analysis module is used to analyze the semantic keywords, obtain the semantic keyword analysis results, determine the dialogue topic, and conduct preliminary dialogue. The degree value determination module is used to track the degree of eye closure of the psychological counseling subject in real time while the initial dialogue is being conducted, and to obtain the real-time eye closure degree value. The emotional state determination module is used to compare the real-time eye closure degree value with the preset eye closure degree value, obtain the comparison result, and combine the comparison result with the semantic keyword analysis result to determine the real-time emotional state of the psychological counseling object. The psychological counseling module is used to extract corresponding emotional psychological counseling plans based on the real-time emotional state of the person receiving psychological counseling and the topic of the dialogue, and to provide psychological counseling to the person receiving psychological counseling.

Images