A method and system for processing indoor abnormal behavior monitoring data

Abstract

The present invention discloses a method and system for processing indoor abnormal behavior monitoring data, including the steps of obtaining indoor abnormal behavior monitoring data, determining the current abnormal level of indoor abnormal behavior according to the indoor abnormal behavior monitoring data and the abnormal level of previous indoor abnormal behavior; and sending an alarm message to a preset address when the current abnormal level of indoor abnormal behavior reaches a preset indoor abnormal behavior alarm level. The present invention can accurately identify indoor abnormal behavior and reduce the misjudgment rate.

Description

Technical Field

[0001] The present invention relates to the field of big data processing, and in particular to a method and system for processing indoor abnormal behavior monitoring data. Background Art

[0002] Abnormal behavior occurring indoors is usually difficult to detect due to its concealed nature.

[0003] Different from the indiscriminate, full-view monitoring of abnormal behavior on the streets, indoor monitoring of abnormal behavior should be sensitive and moderate. It should not be too sensitive, causing rescue workers' efforts to be in vain, nor too slow, causing relevant personnel to be exposed to abnormal behavior for a long time.

[0004] Therefore, in order to meet the above requirements, people need a method and system for processing indoor abnormal behavior monitoring data. Summary of the invention

[0005] In order to solve the above problems, the present invention provides a method for processing indoor abnormal behavior monitoring data, including: obtaining indoor abnormal behavior monitoring data; determining the current indoor abnormal behavior abnormality level based on the indoor abnormal behavior monitoring data and based on previous indoor abnormal behavior abnormality levels; and when the current indoor abnormal behavior abnormality level reaches a preset indoor abnormal behavior alarm level, sending an alarm message to a preset address.

[0006] Preferably, obtaining indoor abnormal behavior monitoring data includes at least one of the following: when the force of the impact on the device exceeds a preset force threshold, the wearable device obtains indoor abnormal behavior degree data; when the crying or screaming of the abnormal behavior participant A exceeds a preset volume threshold, the surveillance camera pre-installed in the house obtains indoor abnormal behavior video data.

[0007] Preferably, determining the abnormality level of the current indoor abnormal behavior includes: judging whether the number of times that sensitive words of indoor abnormal behavior appear in the indoor abnormal behavior monitoring data within a first predetermined time exceeds a preset number threshold; if the judgment result is yes, continuing to judge whether the number of times that the indoor abnormal behavior intensity data exceeds the preset intensity threshold within a second predetermined time exceeds the preset number threshold; if the judgment result is yes, then based on the previous abnormality level of the indoor abnormal behavior, the current abnormality level of the indoor abnormal behavior is increased, otherwise the current abnormality level of the indoor abnormal behavior is maintained unchanged.

[0008] Preferably, before judging whether the number of times that sensitive words for indoor abnormal behavior appear in the indoor abnormal behavior monitoring data within the first predetermined time exceeds a preset number threshold, the method further comprises: obtaining sensitive words for indoor abnormal behavior appearing in the indoor abnormal behavior monitoring data of previous indoor abnormal behavior cases; classifying the abnormal behavior degree of the sensitive words for indoor abnormal behavior into high abnormal behavior degree, medium abnormal behavior degree and low abnormal behavior degree according to the order of the degree of punishment of the abnormal behavior participant B and the degree of tolerance of the abnormal behavior participant A in the previous indoor abnormal behavior case from high to low; and performing deep learning training on the correlation between the voice intonation of the sensitive words for indoor abnormal behavior of the family members of the abnormal behavior participant B and / or the abnormal behavior participant A and the abnormal behavior degree of the sensitive words for indoor abnormal behavior according to the place of the abnormal behavior participant B and / or the abnormal behavior participant A in the previous indoor abnormal behavior case, so as to obtain an indoor abnormal behavior sensitive word library for executing the judgment action of whether the number of times that sensitive words for indoor abnormal behavior appear in the indoor abnormal behavior monitoring data within the first predetermined time exceeds the preset number threshold.

[0009] Preferably, judging whether the number of times the indoor abnormal behavior sensitive words appear in the indoor abnormal behavior monitoring data within the first predetermined time exceeds a preset number threshold comprises: judging whether the voice intonation of the indoor abnormal behavior sensitive words in the indoor abnormal behavior monitoring data is the same as the voice intonation of the indoor abnormal behavior sensitive words of the abnormal behavior participant B and / or the abnormal behavior participant A pre-trained in the indoor abnormal behavior sensitive word library; if the judgment result is yes, obtaining the indoor abnormal behavior sensitive word set corresponding to the place where the abnormal behavior participant B and / or the abnormal behavior participant A belongs from the indoor abnormal behavior sensitive word library; and The indoor abnormal behavior sensitive words appearing in the normal behavior monitoring data are matched with the indoor abnormal behavior sensitive word set, and the number of occurrences of the indoor abnormal behavior sensitive words with high abnormal behavior degree, medium abnormal behavior degree and low abnormal behavior degree are determined respectively, and the occurrence trends of the indoor abnormal behavior sensitive words with high abnormal behavior degree, medium abnormal behavior degree and low abnormal behavior degree are determined; when the number of occurrences of the indoor abnormal behavior sensitive words with high abnormal behavior degree exceeds a preset number threshold within the first predetermined time, and when the occurrence trend of the indoor abnormal behavior sensitive words with high abnormal behavior degree continues to increase, a judgment result of yes is generated.

[0010] Preferably, the deep learning training of the correlation between the voice intonation of the indoor abnormal behavior sensitive words of the abnormal behavior participant B and / or the abnormal behavior participant A and the abnormal behavior degree of the indoor abnormal behavior sensitive words includes: dividing the previous indoor abnormal behavior cases into an indoor abnormal behavior sensitive word training set and an indoor abnormal behavior sensitive word test set according to a preset ratio, wherein the indoor abnormal behavior sensitive word training set and the indoor abnormal behavior sensitive word test set both include no less than a first predetermined number of the previous indoor abnormal behavior cases of different locations, and both also include no less than a second predetermined number of indoor abnormal behavior sensitive words of different abnormal behavior degrees. ; Through each of the previous indoor abnormal behavior cases in the indoor abnormal behavior sensitive vocabulary training set, the correlation between the voice intonation of the indoor abnormal behavior sensitive vocabulary of the abnormal behavior participant B and / or the abnormal behavior participant A and the abnormal behavior degree of the indoor abnormal behavior sensitive vocabulary is trained, so as to obtain the indoor abnormal behavior sensitive vocabulary training library for executing the judgment action of whether the number of times the indoor abnormal behavior sensitive vocabulary appears in the indoor abnormal behavior monitoring data within the first predetermined time exceeds the preset number threshold; Through each of the previous indoor abnormal behavior cases in the indoor abnormal behavior sensitive vocabulary test set, the abnormal behavior participant B and / or the abnormal behavior participant A are trained. The correlation between the voice intonation of A's indoor abnormal behavior sensitive words and the abnormal behavior degree of the indoor abnormal behavior sensitive words is tested, wherein the testing step includes: calculating the deviation value between the actual abnormal behavior degree in the previous indoor abnormal behavior case in the indoor abnormal behavior sensitive words test set and the abnormal behavior degree of the indoor abnormal behavior sensitive words corresponding to the voice intonation of the indoor abnormal behavior sensitive words of the abnormal behavior participant B and / or the abnormal behavior participant A, wherein the influence factor of the high abnormal behavior degree on the deviation value is greater than the influence factor of the medium abnormal behavior degree on the deviation value, and the influence factor of the medium abnormal behavior degree on the deviation value is greater than the influence factor of the low abnormal behavior degree The method further comprises the steps of: determining whether the deviation value is greater than a preset deviation threshold, if the judgment result is yes, then incrementing the deviation counter, if the judgment result is no, then keeping the deviation counter unchanged; for each of the previous indoor abnormal behavior cases in the indoor abnormal behavior sensitive vocabulary test set, determining whether the cumulative sum of the deviation counter is greater than a preset cumulative threshold, if the judgment result is yes, then using the indoor abnormal behavior sensitive vocabulary training library as an indoor abnormal behavior sensitive vocabulary library for executing a judgment action of whether the number of times the indoor abnormal behavior sensitive vocabulary appears in the indoor abnormal behavior monitoring data within a first predetermined time exceeds a preset number threshold.

[0011] Preferably, after obtaining the indoor abnormal behavior sensitive word library, the method further includes: performing deep learning training on the correlation between the voice and intonation of the film and television game abnormal behavior sensitive words appearing in the film and television game and the abnormal behavior degree of the indoor abnormal behavior sensitive words, so as to correct the indoor abnormal behavior sensitive word library used to execute the judgment action of whether the number of times the indoor abnormal behavior sensitive words appear in the indoor abnormal behavior monitoring data within the first predetermined time exceeds a preset number threshold.

[0012] Preferably, the method determines whether the number of times that indoor abnormal behavior sensitive words appear in the indoor abnormal behavior monitoring data within the first predetermined time exceeds a preset number threshold, and further includes: before obtaining the indoor abnormal behavior sensitive word set corresponding to the place of residence of the abnormal behavior participant B and / or the abnormal behavior participant A from the indoor abnormal behavior sensitive word library, determining whether the voice and intonation of the indoor abnormal behavior sensitive words in the indoor abnormal behavior monitoring data are the same as the voice and intonation of the film and television game abnormal behavior sensitive words that appear in the pre-trained film and television games in the indoor abnormal behavior sensitive word library; if the judgment result is yes, reducing the current indoor abnormal behavior abnormality level or maintaining the current indoor abnormal behavior abnormality level unchanged; if the judgment result is no, continuing to execute the acquisition action of acquiring the indoor abnormal behavior sensitive word set from the indoor abnormal behavior sensitive word library according to the voice and intonation of the indoor abnormal behavior sensitive words in the indoor abnormal behavior monitoring data.

[0013] Preferably, the deep learning training for the correlation between the voice and intonation of the abnormal behavior sensitive words of the film and television game and the abnormal behavior degree of the indoor abnormal behavior sensitive words appearing in the film and television game comprises: dividing the film and television game into a film and television game abnormal behavior sensitive words training set and a film and television game abnormal behavior sensitive words test set according to a preset ratio, wherein the film and television game abnormal behavior sensitive words training set and the film and television game abnormal behavior sensitive words test set both include film and television games with different voice and intonations of not less than a third predetermined number, and both also include film and television game abnormal behavior sensitive words with different abnormal behavior degrees of not less than a fourth predetermined number; through each of the film and television games in the film and television game abnormal behavior sensitive words training set, the correlation between the voice and intonation of the film and television game abnormal behavior sensitive words and the abnormal behavior degree of the indoor abnormal behavior sensitive words is trained, so as to obtain a modified indoor abnormal behavior sensitive words training library for executing a judgment action of whether the number of times the indoor abnormal behavior sensitive words appear in the indoor abnormal behavior monitoring data within the first predetermined time exceeds a preset number threshold; through each of the film and television games in the film and television game abnormal behavior sensitive words test set, the voice and intonation of the film and television game abnormal behavior sensitive words and the abnormal behavior of the indoor abnormal behavior sensitive words are trained. The test comprises: calculating the abnormal behavior degree of the indoor abnormal behavior sensitive words corresponding to the voice intonation of the film and television game abnormal behavior sensitive words in the film and television game abnormal behavior sensitive words test set, wherein the influence factor of the high abnormal behavior degree on the calculated abnormal behavior degree is greater than the influence factor of the medium abnormal behavior degree on the calculated abnormal behavior degree, and the influence factor of the medium abnormal behavior degree on the calculated abnormal behavior degree is greater than the influence factor of the low abnormal behavior degree on the calculated abnormal behavior degree; judging whether the calculated abnormal behavior degree is greater than the preset film and television game abnormal behavior degree; If the judgment result is yes, the abnormal behavior counter of the film and television game is incremented, if the judgment result is no, the abnormal behavior counter of the film and television game remains unchanged; for each of the film and television games in the film and television game abnormal behavior sensitive vocabulary test set, it is judged whether the cumulative sum of the abnormal behavior counters of the film and television games is greater than a preset cumulative threshold value, if the judgment result is no, the revised indoor abnormal behavior sensitive vocabulary training library is used as the indoor abnormal behavior sensitive vocabulary library for executing the judgment action of whether the number of times the indoor abnormal behavior sensitive vocabulary appears in the indoor abnormal behavior monitoring data within the first predetermined time exceeds the preset number threshold.

[0014] The present invention also provides a processing system for indoor abnormal behavior monitoring data, comprising: an acquisition device for acquiring indoor abnormal behavior monitoring data; a determination device for determining the current abnormal level of indoor abnormal behavior based on the indoor abnormal behavior monitoring data and based on previous abnormal levels of indoor abnormal behavior; and an alarm device for sending an alarm message to a preset address when the current abnormal level of indoor abnormal behavior reaches a preset indoor abnormal behavior alarm level.

[0015] The present invention can accurately identify abnormal indoor behaviors and reduce the misjudgment rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of this application. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0017] Figure 1 is a flow chart of a method for processing indoor abnormal behavior monitoring data according to an embodiment of the present invention;

[0018] Figure 2 is a flow chart of a method for determining an abnormal level of current indoor abnormal behavior according to an embodiment of the present invention;

[0019] Figure 3 is a flow chart of an automatic learning process of a trial-and-error mode of sensitive words for indoor abnormal behavior according to an embodiment of the present invention;

[0020] Figure 4 is a flow chart of a method for determining whether the number of times sensitive words for indoor abnormal behavior appear in indoor abnormal behavior monitoring data within a first predetermined time exceeds a preset number threshold according to an embodiment of the present invention;

[0021] Figure 5 4 is a schematic diagram of a system for processing indoor abnormal behavior monitoring data according to an embodiment of the present invention. DETAILED DESCRIPTION

[0022] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present invention will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0023] According to an embodiment of the present invention, a method for processing indoor abnormal behavior monitoring data is provided. Figure 1 As shown, the process includes the following steps S102 to S106.

[0024] Step S102: Obtaining indoor abnormal behavior monitoring data. Specifically, the indoor abnormal behavior monitoring data can be obtained in the following ways: when the impact force received by the device exceeds a preset force threshold, the wearable device obtains indoor abnormal behavior degree data, for example, the deformation signal sensed when the wearable device is hit or the impact speed and impact force obtained by processing the deformation signal are compared with a preset threshold. Once the threshold is exceeded, the wearable device obtains indoor abnormal behavior degree data; similarly, when the crying or screaming of the abnormal behavior participant A exceeds a preset volume threshold, the surveillance camera pre-installed in the house obtains indoor abnormal behavior video data.

[0025] Step S104: Determine the current abnormal level of indoor abnormal behavior based on the indoor abnormal behavior monitoring data and the abnormal level of previous indoor abnormal behavior.

[0026] Step S106: When the current abnormal level of indoor abnormal behavior reaches the preset indoor abnormal behavior alarm level, an alarm message is sent to a preset address.

[0027] According to another embodiment of the present invention, a method for determining the abnormal level of current indoor abnormal behavior is provided, such as Figure 2 As shown, the process includes the following steps S202 to S206.

[0028] Step S202: determining whether the number of times that sensitive words for indoor abnormal behavior appear in the indoor abnormal behavior monitoring data within the first predetermined time exceeds a preset number threshold, for example, whether the number M of times that sensitive words for indoor abnormal behavior appear in the indoor abnormal behavior monitoring data within 3 minutes exceeds a preset number threshold of 10.

[0029] Step S204: If the judgment result is yes, continue to judge whether the number of times the indoor abnormal behavior intensity data exceeds the preset intensity threshold within the second predetermined time exceeds the preset number threshold, for example, within the subsequent 3 minutes, whether the number N of times the indoor abnormal behavior intensity data exceeds the preset intensity threshold exceeds the preset number threshold 5.

[0030] Step S206: If the judgment result is yes, then based on the previous abnormal level of indoor abnormal behavior, the current abnormal level of indoor abnormal behavior is increased; otherwise, the current abnormal level of indoor abnormal behavior is maintained unchanged.

[0031] In order to improve the accuracy of monitoring, before judging whether the number of times that indoor abnormal behavior sensitive words appear in the indoor abnormal behavior monitoring data within the first predetermined time exceeds a preset number threshold, indoor abnormal behavior sensitive words appearing in the indoor abnormal behavior monitoring data of previous indoor abnormal behavior cases can be obtained; the abnormal behavior degree of indoor abnormal behavior sensitive words is classified into high abnormal behavior degree, medium abnormal behavior degree and low abnormal behavior degree according to the punishment degree of abnormal behavior participant B and the tolerance degree of abnormal behavior participant A in the previous indoor abnormal behavior cases from high to low; because in actual situations, special languages ​​such as dialects may be mixed, in this embodiment, according to the place of the abnormal behavior participant B and / or the abnormal behavior participant A in the previous indoor abnormal behavior case, the voice intonation of the indoor abnormal behavior sensitive words of the abnormal behavior participant B and / or the abnormal behavior participant A and the abnormal behavior degree of the indoor abnormal behavior sensitive words can be deeply learned and trained, so as to obtain an indoor abnormal behavior sensitive word library for executing the judgment action of whether the number of times that indoor abnormal behavior sensitive words appear in the indoor abnormal behavior monitoring data within the first predetermined time exceeds the preset number threshold.

[0032] According to another embodiment of the present invention, before determining whether the number of times that sensitive words for indoor abnormal behavior appear in the indoor abnormal behavior monitoring data within the first predetermined time exceeds a preset number threshold, it also includes automatically learning the indoor abnormal behavior sensitive word library in a trial-and-error mode. The process is implemented by the following modules: an intelligent operation unit; an environment unit for interacting with the intelligent operation unit; a state unit for indicating the current location or current field of view of the intelligent operation unit; an action unit for taking different actions to achieve different states; a strategy unit for instructing the action unit to take different actions; and a feedback unit for accumulating positive feedback when the intelligent operation unit takes the correct action.

[0033] The interaction between the intelligent operation unit and the environment can be achieved in the following ways: the intelligent operation unit performs corresponding actions based on certain strategies in a specific environment, and then obtains rewards from the environment to migrate to the next state. As deep learning proceeds, the intelligent operation unit will explore what strategy to use to perform actions in the current state to maximize the final reward.

[0034] Specifically, the intelligent operation unit performs an action A in a certain state S. The environment will judge the quality of the action according to the settings, and then feedback the corresponding reward R to the intelligent operation unit, and transfer to the next state S'. Figure 3As shown in the figure, the framework is divided into two interference stages, namely pre-interference and real interference, where real interference mainly includes four modules: feature extraction algorithm, interference algorithm, environment and intelligent operation unit.

[0035] The intelligent operation unit is the main part of the algorithm, and the main modules related to it are as follows:

[0036] (1) State: The state is the input of the intelligent operation unit. In this patent, it refers to the feature-related information of the indoor abnormal behavior sensitive word sample. This patent application introduces a historical frame in the state space, and retains N historical records of features in the state space, where the first record always retains the initial features, and the subsequent N-1 records retain the features generated by the most recent N-1 interferences. The role of the first record is to retain the original information of the sample, which is conducive to the intelligent operation unit to judge the initial features of the sample; the subsequent N-1 records can effectively guide the intelligent operation unit to learn the change trajectory of the sample caused by the historical action, and learn the relationship between each record.

[0037] (2) Strategy: Strategy is the method by which the intelligent operation unit executes decisions. In the algorithm of this patent, the process of the intelligent operation unit executing decisions is completed by the value function neural network and the strategy function neural network.

[0038] (3) Reward: Reward is the feedback from the environment after the intelligent operation unit performs an action. Correct actions will receive positive feedback, which is a relatively large reward value; incorrect actions will receive negative feedback as punishment, which is a relatively small reward value.

[0039] Before generating malicious samples of sensitive words for indoor abnormal behavior to interfere with the model of sensitive words for indoor abnormal behavior, a detector can be pre-trained as the attack target of malicious samples of sensitive words for indoor abnormal behavior, and also as an environmental unit to provide feedback to the intelligent operation unit. The method and system provided by this embodiment can improve the accuracy of identifying sensitive words for indoor abnormal behavior.

[0040] According to another embodiment of the present invention, Figure 4 As shown, a method is provided to determine whether the number of times that sensitive words for indoor abnormal behavior appear in the indoor abnormal behavior monitoring data within a first predetermined time exceeds a preset number threshold, such as Figure 4 As shown, the following steps S402 to S408 are included.

[0041] Step S402: determining whether the voice and intonation of the indoor abnormal behavior sensitive words in the indoor abnormal behavior monitoring data are the same as the voice and intonation of the indoor abnormal behavior sensitive words of the abnormal behavior participant B and / or the abnormal behavior participant A pre-trained in the indoor abnormal behavior sensitive word library.

[0042] Step S404: If the judgment result is yes, then the indoor abnormal behavior sensitive word set corresponding to the location of the abnormal behavior participant B and / or the abnormal behavior participant A is obtained from the indoor abnormal behavior sensitive word library.

[0043] Step S406: Match the indoor abnormal behavior sensitive words appearing in the indoor abnormal behavior monitoring data with the indoor abnormal behavior sensitive word set, determine the number of occurrences of indoor abnormal behavior sensitive words with high abnormal behavior degree, medium abnormal behavior degree and low abnormal behavior degree respectively, and determine the occurrence trends of indoor abnormal behavior sensitive words with high abnormal behavior degree, medium abnormal behavior degree and low abnormal behavior degree.

[0044] Step S408: if the number of occurrences of the indoor abnormal behavior sensitive words with a high abnormal behavior degree exceeds a preset number threshold within the first predetermined time, and if the occurrence trend of the indoor abnormal behavior sensitive words with a high abnormal behavior degree continues to increase, generate a judgment result of yes.

[0045] Further, deep learning training is performed on the correlation between the voice intonation of the indoor abnormal behavior sensitive words of the abnormal behavior participant B and / or the abnormal behavior participant A and the abnormal behavior degree of the indoor abnormal behavior sensitive words, including: dividing the previous indoor abnormal behavior cases into an indoor abnormal behavior sensitive word training set and an indoor abnormal behavior sensitive word test set according to a preset ratio, wherein the indoor abnormal behavior sensitive word training set and the indoor abnormal behavior sensitive word test set both include no less than a first predetermined number of the previous indoor abnormal behavior cases from different locations, and both also include no less than a second predetermined number of indoor abnormal behavior sensitive words of different abnormal behavior degrees. The indoor abnormal behavior sensitive vocabulary training library is used to perform a judgment action of whether the number of times the indoor abnormal behavior sensitive vocabulary appears in the indoor abnormal behavior monitoring data within the first predetermined time exceeds a preset number threshold; the indoor abnormal behavior sensitive vocabulary training library is used to perform a judgment action of whether the number of times the indoor abnormal behavior sensitive vocabulary appears in the indoor abnormal behavior monitoring data within the first predetermined time exceeds a preset number threshold through each of the previous indoor abnormal behavior cases in the indoor abnormal behavior sensitive vocabulary test set. The correlation between the voice intonation of the indoor abnormal behavior sensitive words of person A and the abnormal behavior degree of the indoor abnormal behavior sensitive words is tested, wherein the testing step includes: calculating the deviation value between the actual abnormal behavior degree in the previous indoor abnormal behavior case in the indoor abnormal behavior sensitive words test set and the abnormal behavior degree of the indoor abnormal behavior sensitive words corresponding to the voice intonation of the indoor abnormal behavior sensitive words of abnormal behavior participant B and / or abnormal behavior participant A, wherein the influence factor of the high abnormal behavior degree on the deviation value is greater than the influence factor of the medium abnormal behavior degree on the deviation value, and the influence factor of the medium abnormal behavior degree on the deviation value is greater than the influence factor of the low abnormal behavior degree on the deviation value. is an influence factor of the degree on the deviation value; judging whether the deviation value is greater than a preset deviation threshold, if the judgment result is yes, the deviation counter is incremented, if the judgment result is no, the deviation counter remains unchanged; for each of the previous indoor abnormal behavior cases in the indoor abnormal behavior sensitive vocabulary test set, judging whether the cumulative sum of the deviation counter is greater than a preset cumulative threshold, if the judgment result is yes, using the indoor abnormal behavior sensitive vocabulary training library as an indoor abnormal behavior sensitive vocabulary library for executing a judgment action of whether the number of times the indoor abnormal behavior sensitive vocabulary appears in the indoor abnormal behavior monitoring data within a first predetermined time exceeds a preset number threshold.

[0046] In addition, in order to correct the indoor abnormal behavior sensitive word library used to execute the judgment action of whether the number of times the indoor abnormal behavior sensitive words appear in the indoor abnormal behavior monitoring data within the first predetermined time exceeds a preset number threshold, after obtaining the indoor abnormal behavior sensitive word library, it also includes: deep learning training is performed on the correlation between the voice and intonation of the film and television game abnormal behavior sensitive words appearing in the film and television games and the abnormal behavior degree of the indoor abnormal behavior sensitive words.

[0047] Further, judging whether the number of times that indoor abnormal behavior sensitive words appear in the indoor abnormal behavior monitoring data within the first predetermined time exceeds a preset number threshold also includes: before obtaining the indoor abnormal behavior sensitive word set corresponding to the place of residence of the abnormal behavior participant B and / or the abnormal behavior participant A from the indoor abnormal behavior sensitive word library, judging whether the voice and intonation of the indoor abnormal behavior sensitive words in the indoor abnormal behavior monitoring data are the same as the voice and intonation of the film and television game abnormal behavior sensitive words appearing in the pre-trained film and television games in the indoor abnormal behavior sensitive word library; if the judgment result is yes, reducing the current indoor abnormal behavior abnormality level or maintaining the current indoor abnormal behavior abnormality level unchanged; if the judgment result is no, continuing to execute the acquisition action of acquiring the indoor abnormal behavior sensitive word set from the indoor abnormal behavior sensitive word library according to the voice and intonation of the indoor abnormal behavior sensitive words in the indoor abnormal behavior monitoring data.

[0048] Further, the deep learning training for the correlation between the voice and intonation of the abnormal behavior sensitive words of the film and television game and the abnormal behavior degree of the indoor abnormal behavior sensitive words that appear in the film and television game includes: dividing the film and television game into a film and television game abnormal behavior sensitive words training set and a film and television game abnormal behavior sensitive words test set according to a preset ratio, wherein the film and television game abnormal behavior sensitive words training set and the film and television game abnormal behavior sensitive words test set both include film and television games with different voice and intonations of not less than a third predetermined number, and both also include film and television game abnormal behavior sensitive words with different abnormal behavior degrees of not less than a fourth predetermined number; through each of the film and television games in the film and television game abnormal behavior sensitive words training set, the correlation between the voice and intonation of the film and television game abnormal behavior sensitive words and the abnormal behavior degree of the indoor abnormal behavior sensitive words is trained, so as to obtain a modified indoor abnormal behavior sensitive words training library for executing a judgment action of whether the number of times the indoor abnormal behavior sensitive words appear in the indoor abnormal behavior monitoring data within the first predetermined time exceeds a preset number threshold; through each of the film and television games in the film and television game abnormal behavior sensitive words test set, the voice and intonation of the film and television game abnormal behavior sensitive words and the abnormal behavior of the indoor abnormal behavior sensitive words are trained. The test comprises: calculating the abnormal behavior degree of the indoor abnormal behavior sensitive words corresponding to the voice intonation of the film and television game abnormal behavior sensitive words in the film and television game abnormal behavior sensitive words test set, wherein the influence factor of the high abnormal behavior degree on the calculated abnormal behavior degree is greater than the influence factor of the medium abnormal behavior degree on the calculated abnormal behavior degree, and the influence factor of the medium abnormal behavior degree on the calculated abnormal behavior degree is greater than the influence factor of the low abnormal behavior degree on the calculated abnormal behavior degree; judging whether the calculated abnormal behavior degree is greater than the preset film and television game abnormal behavior degree; If the judgment result is yes, the abnormal behavior counter of the film and television game is incremented, if the judgment result is no, the abnormal behavior counter of the film and television game remains unchanged; for each of the film and television games in the film and television game abnormal behavior sensitive vocabulary test set, it is judged whether the cumulative sum of the abnormal behavior counters of the film and television games is greater than a preset cumulative threshold value, if the judgment result is no, the revised indoor abnormal behavior sensitive vocabulary training library is used as the indoor abnormal behavior sensitive vocabulary library for executing the judgment action of whether the number of times the indoor abnormal behavior sensitive vocabulary appears in the indoor abnormal behavior monitoring data within the first predetermined time exceeds the preset number threshold.

[0049] Furthermore, in order to prevent power outages, a power-off battery can be equipped. Even if there is a sudden power outage, the backup battery can be used for more than 48 hours. After power is turned on, the battery will automatically store electricity.

[0050] Furthermore, in order to prevent the monitored person from dismantling the device without authorization, an anti-dismantling alarm can be installed on the system. Once the system detects abnormal dismantling behavior, it will immediately connect to the preset address to send an alarm message. After receiving the alarm message, the preset address will visit the site according to the system's location address to check the situation.

[0051] According to another embodiment of the present invention, Figure 5 As shown, a system for processing indoor abnormal behavior monitoring data is provided, including: an acquisition device for acquiring indoor abnormal behavior monitoring data; a determination device for determining the current abnormal level of indoor abnormal behavior based on the indoor abnormal behavior monitoring data and based on previous abnormal levels of indoor abnormal behavior; and an alarm device for sending an alarm message to a preset address when the current abnormal level of indoor abnormal behavior reaches a preset indoor abnormal behavior alarm level.

[0052] It is worth noting that the specific method of obtaining indoor abnormal behavior monitoring data, determining the abnormal level of the current indoor abnormal behavior, and sending an alarm message to a preset address is consistent with the method described in the above specific embodiment.

[0053] It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0054] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for processing indoor abnormal behavior monitoring data, characterized in that: include: Obtain indoor abnormal behavior monitoring data; Determining whether the number of times that sensitive words for indoor abnormal behavior appear in the indoor abnormal behavior monitoring data within the first predetermined time exceeds a preset number threshold; If the result of the determination is yes, continue to determine whether the number of times the indoor abnormal behavior intensity data exceeds the preset intensity threshold within the second predetermined time exceeds the preset number threshold; If the judgment result is yes, then based on the previous abnormal level of indoor abnormal behavior, the current abnormal level of indoor abnormal behavior is increased, otherwise the current abnormal level of indoor abnormal behavior is maintained unchanged; When the abnormal level of the current indoor abnormal behavior reaches a preset indoor abnormal behavior alarm level, sending an alarm message to a preset address; Before determining whether the number of times that sensitive words for indoor abnormal behavior appear in the indoor abnormal behavior monitoring data within the first predetermined time exceeds a preset number threshold, the method further includes: Acquire sensitive words of indoor abnormal behavior appearing in indoor abnormal behavior monitoring data of previous indoor abnormal behavior cases; According to the order of the degree of punishment of the abnormal behavior participant B and the degree of tolerance of the abnormal behavior participant A in the previous indoor abnormal behavior case, the abnormal behavior degree of the indoor abnormal behavior sensitive words is classified into high abnormal behavior degree, medium abnormal behavior degree and low abnormal behavior degree; Dividing the previous indoor abnormal behavior cases into an indoor abnormal behavior sensitive vocabulary training set and an indoor abnormal behavior sensitive vocabulary test set according to a preset ratio, wherein the indoor abnormal behavior sensitive vocabulary training set and the indoor abnormal behavior sensitive vocabulary test set both include no less than a first predetermined number of the previous indoor abnormal behavior cases from different locations, and both include no less than a second predetermined number of indoor abnormal behavior sensitive vocabulary of different abnormal behavior degrees; Through each of the previous indoor abnormal behavior cases in the indoor abnormal behavior sensitive word training set, the correlation between the voice intonation of the indoor abnormal behavior sensitive words of the abnormal behavior participant B and / or the abnormal behavior participant A and the abnormal behavior degree of the indoor abnormal behavior sensitive words is trained, so as to obtain the indoor abnormal behavior sensitive word training library for executing the judgment action of whether the number of times the indoor abnormal behavior sensitive words appear in the indoor abnormal behavior monitoring data within the first predetermined time exceeds a preset number threshold; Through each of the previous indoor abnormal behavior cases in the indoor abnormal behavior sensitive word test set, the correlation between the voice intonation of the indoor abnormal behavior sensitive words of the abnormal behavior participant B and / or the abnormal behavior participant A and the abnormal behavior degree of the indoor abnormal behavior sensitive words is tested, including: Calculate the deviation between the actual abnormal behavior degree in the previous indoor abnormal behavior cases in the indoor abnormal behavior sensitive word test set and the abnormal behavior degree of the indoor abnormal behavior sensitive words corresponding to the voice intonation of the indoor abnormal behavior sensitive words of the abnormal behavior participant B and / or the abnormal behavior participant A, wherein the influence factor of the high abnormal behavior degree on the deviation value is greater than the influence factor of the medium abnormal behavior degree on the deviation value, and the influence factor of the medium abnormal behavior degree on the deviation value is greater than the influence factor of the low abnormal behavior degree on the deviation value; Determine whether the deviation value is greater than a preset deviation threshold, if the determination result is yes, the deviation counter is incremented, if the determination result is no, the deviation counter remains unchanged; For each of the previous indoor abnormal behavior cases in the indoor abnormal behavior sensitive vocabulary test set, it is determined whether the accumulated sum of the deviation counter is less than a preset accumulated threshold value. If the determination result is yes, the indoor abnormal behavior sensitive vocabulary training library is used as an indoor abnormal behavior sensitive vocabulary library for executing a judgment action of whether the number of times the indoor abnormal behavior sensitive vocabulary appears in the indoor abnormal behavior monitoring data within a first predetermined time exceeds a preset number threshold value.

2. The method for processing indoor abnormal behavior monitoring data according to claim 1, characterized in that: Acquiring indoor abnormal behavior monitoring data includes at least one of the following: When abnormal behavior participant A activates the function on his wearable device for preventing abnormal indoor behavior, the wearable device obtains audio data of abnormal indoor behavior; When the force of hitting the wearable device of abnormal behavior participant A exceeds a preset force threshold, the wearable device acquires indoor abnormal behavior degree data; When the crying or screaming of the abnormal behavior participant A exceeds a preset volume threshold, the surveillance camera pre-installed in the house obtains indoor abnormal behavior video data.

3. The method for processing indoor abnormal behavior monitoring data according to claim 1, characterized in that: Determining whether the number of times that sensitive words for indoor abnormal behavior appear in the indoor abnormal behavior monitoring data within the first predetermined time exceeds a preset number threshold includes: Determine whether the voice and intonation of the indoor abnormal behavior sensitive words in the indoor abnormal behavior monitoring data are the same as the voice and intonation of the indoor abnormal behavior sensitive words of the abnormal behavior participant B and / or the abnormal behavior participant A pre-trained in the indoor abnormal behavior sensitive word library; If the judgment result is yes, then obtaining the indoor abnormal behavior sensitive word set corresponding to the place where the abnormal behavior participant B and / or the abnormal behavior participant A belongs from the indoor abnormal behavior sensitive word library; Matching the indoor abnormal behavior sensitive words appearing in the indoor abnormal behavior monitoring data with the indoor abnormal behavior sensitive word set, respectively determining the number of occurrences of indoor abnormal behavior sensitive words with high abnormal behavior degree, medium abnormal behavior degree and low abnormal behavior degree, and determining the occurrence trends of indoor abnormal behavior sensitive words with high abnormal behavior degree, medium abnormal behavior degree and low abnormal behavior degree; If the number of occurrences of the indoor abnormal behavior sensitive words with a high abnormal behavior degree exceeds a preset number threshold within the first predetermined time, and if the occurrence trend of the indoor abnormal behavior sensitive words with a high abnormal behavior degree continues to increase, a judgment result is generated as yes.

4. The method for processing indoor abnormal behavior monitoring data according to claim 1, characterized in that: After obtaining the indoor abnormal behavior sensitive word library, the method further includes: performing deep learning training on the correlation between the voice and intonation of the film and television game abnormal behavior sensitive words appearing in the film and television game and the abnormal behavior degree of the indoor abnormal behavior sensitive words, so as to correct the indoor abnormal behavior sensitive word library used to perform the judgment action of whether the number of times the indoor abnormal behavior sensitive words appear in the indoor abnormal behavior monitoring data within the first predetermined time exceeds a preset number threshold.

5. The method for processing indoor abnormal behavior monitoring data according to claim 4, characterized in that: The method further comprises: determining whether the number of times that sensitive words for indoor abnormal behavior appear in the indoor abnormal behavior monitoring data within the first predetermined time exceeds a preset number threshold; Before obtaining the indoor abnormal behavior sensitive word set corresponding to the place of the abnormal behavior participant B and / or the abnormal behavior participant A from the indoor abnormal behavior sensitive word library, determining whether the voice and intonation of the indoor abnormal behavior sensitive words in the indoor abnormal behavior monitoring data are the same as the voice and intonation of the film and television game abnormal behavior sensitive words that appear in the pre-trained film and television games in the indoor abnormal behavior sensitive word library; If the judgment result is yes, then the abnormality level of the current indoor abnormal behavior is reduced or the abnormality level of the current indoor abnormal behavior is maintained unchanged; If the judgment result is no, then continue to execute the acquisition action of acquiring the indoor abnormal behavior sensitive word set from the indoor abnormal behavior sensitive word library according to the voice intonation of the indoor abnormal behavior sensitive words in the indoor abnormal behavior monitoring data.

6. The method for processing indoor abnormal behavior monitoring data according to claim 4 or 5, characterized in that: The deep learning training for the correlation between the intonation of the abnormal behavior sensitive words in the film and television games and the abnormal behavior degree of the indoor abnormal behavior sensitive words includes: Dividing the film and television games into a film and television game abnormal behavior sensitive vocabulary training set and a film and television game abnormal behavior sensitive vocabulary test set according to a preset ratio, wherein the film and television game abnormal behavior sensitive vocabulary training set and the film and television game abnormal behavior sensitive vocabulary test set both include film and television games with different voice tones not less than a third predetermined number, and both also include film and television game abnormal behavior sensitive vocabulary with different abnormal behavior degrees not less than a fourth predetermined number; Through each of the film and television games in the film and television game abnormal behavior sensitive vocabulary training set, the correlation between the voice intonation of the film and television game abnormal behavior sensitive vocabulary and the abnormal behavior degree of the indoor abnormal behavior sensitive vocabulary is trained, so as to obtain a modified indoor abnormal behavior sensitive vocabulary training library for executing a judgment action of whether the number of times the indoor abnormal behavior sensitive vocabulary appears in the indoor abnormal behavior monitoring data within a first predetermined time exceeds a preset number threshold; Through each of the film and television games in the film and television game abnormal behavior sensitive vocabulary test set, the correlation between the voice intonation of the film and television game abnormal behavior sensitive vocabulary and the abnormal behavior degree of the indoor abnormal behavior sensitive vocabulary is tested, wherein the testing step includes: Calculating the abnormal behavior degree of the indoor abnormal behavior sensitive words corresponding to the voice intonation of the film and television game abnormal behavior sensitive words of the film and television game in the film and television game abnormal behavior sensitive words test set, wherein the influence factor of the high abnormal behavior degree on the calculated abnormal behavior degree is greater than the influence factor of the medium abnormal behavior degree on the calculated abnormal behavior degree, and the influence factor of the medium abnormal behavior degree on the calculated abnormal behavior degree is greater than the influence factor of the low abnormal behavior degree on the calculated abnormal behavior degree; Determine whether the calculated abnormal behavior degree is greater than a preset threshold value of abnormal behavior degree of a video game. If the determination result is yes, the abnormal behavior counter of the video game is incremented. If the determination result is no, the abnormal behavior counter of the video game remains unchanged. For each of the film and television games in the film and television game abnormal behavior sensitive vocabulary test set, it is determined whether the accumulated total of the film and television game abnormal behavior counters is greater than a preset accumulated threshold value; if the determination result is no, the revised indoor abnormal behavior sensitive vocabulary training library is used as the indoor abnormal behavior sensitive vocabulary library for executing a judgment action of whether the number of times the indoor abnormal behavior sensitive vocabulary appears in the indoor abnormal behavior monitoring data within the first predetermined time exceeds a preset number threshold value.

7. A processing system for indoor abnormal behavior monitoring data, characterized in that: include: The second acquisition device is used to acquire indoor abnormal behavior monitoring data; The determination device is used to determine whether the number of times that the indoor abnormal behavior sensitive words appear in the indoor abnormal behavior monitoring data within the first predetermined time exceeds a preset number threshold; if the determination result is yes, continue to determine whether the number of times that the indoor abnormal behavior intensity data exceeds the preset intensity threshold within the second predetermined time exceeds the preset number threshold; if the determination result is yes, then based on the previous indoor abnormal behavior abnormality level, the current indoor abnormal behavior abnormality level is increased, otherwise the current indoor abnormal behavior abnormality level is maintained unchanged; An alarm device is used to send an alarm message to a preset address when the abnormal level of the current indoor abnormal behavior reaches a preset indoor abnormal behavior alarm level; Before determining whether the number of times that sensitive words for indoor abnormal behavior appear in the indoor abnormal behavior monitoring data within the first predetermined time exceeds a preset number threshold, the system further includes: The first acquisition device is used to acquire indoor abnormal behavior sensitive words appearing in indoor abnormal behavior monitoring data of previous indoor abnormal behavior cases; A classification device, for classifying the abnormal behavior degree of the indoor abnormal behavior sensitive words into high abnormal behavior degree, medium abnormal behavior degree and low abnormal behavior degree according to the order of the punishment degree of the abnormal behavior participant B and the tolerance degree of the abnormal behavior participant A in the previous indoor abnormal behavior case from high to low; a dividing device, for dividing the previous indoor abnormal behavior cases into a training set of indoor abnormal behavior sensitive words and a test set of indoor abnormal behavior sensitive words according to a preset ratio, wherein the training set of indoor abnormal behavior sensitive words and the test set of indoor abnormal behavior sensitive words both include no less than a first predetermined number of the previous indoor abnormal behavior cases of different locations, and both include no less than a second predetermined number of indoor abnormal behavior sensitive words of different abnormal behavior degrees; A training device, for training the correlation between the voice intonation of the indoor abnormal behavior sensitive words of the abnormal behavior participant B and / or the abnormal behavior participant A and the abnormal behavior degree of the indoor abnormal behavior sensitive words through each of the previous indoor abnormal behavior cases in the indoor abnormal behavior sensitive words training set, so as to obtain an indoor abnormal behavior sensitive words training library for executing a judgment action of whether the number of times the indoor abnormal behavior sensitive words appear in the indoor abnormal behavior monitoring data within a first predetermined time exceeds a preset number threshold; The testing device is used to test the correlation between the voice intonation of the indoor abnormal behavior sensitive words of the abnormal behavior participant B and / or the abnormal behavior participant A and the abnormal behavior degree of the indoor abnormal behavior sensitive words through each of the previous indoor abnormal behavior cases in the indoor abnormal behavior sensitive words test set, including: Calculate the deviation between the actual abnormal behavior degree in the previous indoor abnormal behavior cases in the indoor abnormal behavior sensitive word test set and the abnormal behavior degree of the indoor abnormal behavior sensitive words corresponding to the voice intonation of the indoor abnormal behavior sensitive words of the abnormal behavior participant B and / or the abnormal behavior participant A, wherein the influence factor of the high abnormal behavior degree on the deviation value is greater than the influence factor of the medium abnormal behavior degree on the deviation value, and the influence factor of the medium abnormal behavior degree on the deviation value is greater than the influence factor of the low abnormal behavior degree on the deviation value; Determine whether the deviation value is greater than a preset deviation threshold, if the determination result is yes, the deviation counter is incremented, if the determination result is no, the deviation counter remains unchanged; For each of the previous indoor abnormal behavior cases in the indoor abnormal behavior sensitive vocabulary test set, it is determined whether the accumulated sum of the deviation counter is less than a preset accumulated threshold value. If the determination result is yes, the indoor abnormal behavior sensitive vocabulary training library is used as an indoor abnormal behavior sensitive vocabulary library for executing a judgment action of whether the number of times the indoor abnormal behavior sensitive vocabulary appears in the indoor abnormal behavior monitoring data within a first predetermined time exceeds a preset number threshold value.

Images