CN114331786A - Community management method and system based on Internet of things - Google Patents
Community management method and system based on Internet of things Download PDFInfo
- Publication number
- CN114331786A CN114331786A CN202111646967.3A CN202111646967A CN114331786A CN 114331786 A CN114331786 A CN 114331786A CN 202111646967 A CN202111646967 A CN 202111646967A CN 114331786 A CN114331786 A CN 114331786A
- Authority
- CN
- China
- Prior art keywords
- data
- image
- community
- characteristic data
- audio
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a community management method and a community management system based on the Internet of things, wherein the method comprises the following steps: s01, constructing a community service database; s02, acquiring monitoring image data in a community public area range in real time, and then performing feature recognition on the monitoring image data according to preset conditions to generate image feature data; s03, collecting image characteristic data of the same image monitoring area at different time nodes in the monitored image data to generate an image characteristic data set; s04, acquiring an image characteristic data set according to a preset time frequency, and judging the characteristic data in the image characteristic data set according to a preset condition to generate a judgment result; s05, acquiring and identifying the judgment result, setting the data corresponding to the judgment result as abnormal data and marking the information and generating early warning information to be output when the judgment result does not meet the preset condition, and playing the advocated propaganda data according to the preset condition; the scheme is reliable in implementation, flexible in response and low in cost, and has a great implementation and popularization value.
Description
Technical Field
The invention relates to the technical field of application of the Internet of things, in particular to a community management method and system based on the Internet of things.
Background
The community is an important place for people to live in fixed groups, as the requirements of people on quality of life and environment are higher and higher, the community is used as the environment of group life, harmonious community humanistic environment and clean and comfortable field environment are the working targets of community workers and management departments, as most of current community residents are people from five lakes and four seas, the living habits and the maintenance levels are different, for community managers, how to coordinate and advocate harmonious humanistic and living environments is a common problem, and as the concept of the internet of things is provided, the hardware equipment of community service is not in an independent working state all the day long, various hardware equipment is controlled in a combined mode, the maximum utilization and the optimal matching are the current hot research subject, however, in the aspect of the internet of things service, the current matched hardware part is mainly the combination of a community monitoring system and a security department, the method has the advantages that remote alarm monitoring is realized, other aspects are rare, therefore, most of current community management is a mode of after-the-fact management or after-the-fact relief, measures are usually taken after an uncivilized phenomenon occurs, if intervention can be performed before an uncivilized behavior is expanded from a case to a group normal state, the method has practical significance, on the aspect of hardware support, because Internet of things engineering is a huge and capital-consuming project, the cost required by a common community is high, and therefore, if the cooperation of Internet of things equipment is optimized, the laying cost and the use cost of local hardware are reduced, the method is a problem with practical significance.
Disclosure of Invention
In view of this, the present invention provides a method and a system for community management based on internet of things, which are reliable in implementation, flexible in response, and low in cost.
In order to achieve the technical purpose, the technical scheme adopted by the invention is as follows:
an Internet of things-based community management method, comprising:
s01, constructing a community service database, wherein community monitoring data and advocation and propaganda data are stored in the community service database;
s02, acquiring monitoring image data in a community public area range in real time, and then performing feature recognition on the monitoring image data according to preset conditions to generate image feature data;
s03, collecting image characteristic data of the same image monitoring area at different time nodes in the monitored image data to generate an image characteristic data set;
s04, acquiring an image characteristic data set according to a preset time frequency, and judging the characteristic data in the image characteristic data set according to a preset condition to generate a judgment result;
s05, acquiring and identifying the judgment result, setting the data corresponding to the judgment result as abnormal data and marking the information and generating early warning information to be output when the judgment result does not meet the preset condition, and simultaneously selecting advocated propaganda data according to the preset condition and issuing the advocated propaganda data to the community for playing by public media playing equipment of the community;
s06, constructing a management evaluation model, associating the management evaluation model with a community one to one, setting evaluation elements and weight values of the evaluation elements, acquiring abnormal data of the corresponding community and information marks of the abnormal data, identifying and classifying the information marks or the abnormal data, corresponding identification and classification results with the evaluation elements of the management evaluation model, adjusting the weight values of the evaluation elements, and outputting an early warning instruction according to a preset condition when the weight values of the evaluation elements are higher than a preset threshold value.
As a possible implementation manner, further, in the scheme, S02 further includes monitoring decibel values of sounds in a public area range of a community in real time, recording and storing sounds above 50 or 70 decibels, marking a starting time node of occurrence of the sounds, generating audio feature data, and then collecting audio feature data of the same audio monitoring area at different time nodes to generate an audio feature data set;
s04 further comprises the steps of obtaining an audio characteristic data set, judging the audio characteristic data in the audio characteristic data set according to preset conditions, and generating a judgment result;
the judgment result obtained in S05 includes a judgment result of the audio feature data and a judgment result of the video feature data.
As a preferred alternative, in S02, preferably, the specific method for performing feature recognition on the monitored image data according to the preset condition to generate the image feature data includes:
0201, acquiring images in a monitoring range corresponding to the monitoring images according to a preset time interval, then guiding the images into a positioning neural network for element positioning, then guiding the elements obtained by positioning into a detection neural network for feature identification, obtaining information tags of the positioned elements, and setting the information tags as reference tags, wherein the positioning areas among the positioned elements are not coincident;
0202, extracting image frames in the monitoring image data from the monitoring image data acquired in real time in the community public area range by taking 1 second as a time interval unit, then importing the image frames into a positioning neural network for element positioning, importing the elements obtained by positioning into a detection neural network for feature identification, acquiring information tags of the positioned elements, and setting the information tags as real-time tags;
and S0203, merging the real-time tag and the reference tag, then excluding the intersection part of the real-time tag and the reference tag to obtain a difference tag, and extracting and associating the difference tag and the corresponding element in the image frame to obtain image characteristic data.
As a preferred alternative, it is preferable that S03 includes: collecting image characteristic data of the same image monitoring area in different time nodes in the monitored image data, selecting and storing the image characteristic data with the same real-time label and the element similarity of more than 90%, and deleting the rest to generate an image characteristic data set.
As a preferred alternative, it is preferable that S04 includes:
s0411, acquiring an image characteristic data set according to the time frequency of each time of 10-20S, and classifying the characteristic data in the same image characteristic data set into people, animals, vehicles and the like;
s0412, the characteristic data classified into the characters further identifies the articles carried by the hand regions, when the identification result is non-empty, the identified articles and the same image characteristic data are classified into other image characteristic data in a centralized manner to be subjected to secondary matching, when the matching probability is more than 80%, the article abandon is output as a judgment result, and the image characteristic data corresponding to the judgment result is output together;
carrying out article identification on the neck region of the characteristic data classified into the animal, outputting an unconstrained animal as a judgment result when the characteristic data is identified as empty, and outputting image characteristic data corresponding to the judgment result;
and further carrying out license plate recognition on the feature data classified into the vehicles, outputting abnormal vehicles as judgment results when the license plates are recognized to be empty, and outputting image feature data corresponding to the judgment results together.
As a preferred alternative, it is preferable that S04 further includes: s0421, leading the audio characteristic data in the audio characteristic data set into a trained time delay neural network-hidden Markov model to identify the audio characteristic data, outputting an audio identification result, and setting the audio identification result as a judgment result, wherein the audio identification result comprises at least one of abnormal life noise and abnormal animal noise;
wherein, the training process of the time-delay neural network-hidden Markov model comprises the following steps:
obtaining each information frame of training audio, and the Mel cepstrum characteristic, the tone characteristic and the current speaker characteristic of each information frame;
and training the speech recognition model by taking the Mel cepstrum characteristic, the tone characteristic and the current speaker characteristic of a current information frame, and a plurality of continuous historical information frames before the current information frame and a plurality of continuous future information frames after the current information frame as input until the speech recognition model is converged.
As a preferred alternative embodiment, in S06, the evaluation element preferably includes at least one of waste abnormal disposal, abnormal animal behavior, abnormal vehicle running, abnormal living noise, and abnormal animal noise;
when the recognition classification result corresponds to an evaluation element of the management evaluation model, the weight value of the evaluation element is adjusted according to the following formula:
X＝X1+Y1(A)
Wherein, X1Is the current weight value of the evaluation element, X is the weight value of the evaluation element after weight adjustment, Y is the single weight adjustment value, and in addition, X is the current weight value of the evaluation element after weight adjustment1Is 0.5, Y1Is 0.01;
when the weighted value of an evaluation element of a management evaluation model is not adjusted within a preset time length, the weighted value of the evaluation element is adjusted according to the following formula:
X’＝X1’－Y2(II)
Wherein, X1'is the current weight value of the evaluation element, X' is the weight value of the evaluation element after weight adjustment, Y2Adjust the value for the one-time weight, in addition, X1' has an initial value of 0.5, Y2Is 0.005.
Based on the above technical solution, the present invention further provides a community management system based on the internet of things, which includes:
the background server is used for constructing a community service database and a management evaluation model, and community monitoring data and advocation propaganda data are stored in the community service database;
the system comprises a plurality of image monitoring devices, a plurality of image monitoring devices and a plurality of image monitoring devices, wherein the image monitoring devices are arranged in a public area range of a community and are used for acquiring monitoring image data in the public area range of the community in real time;
the audio monitoring devices are arranged in a public area range of a community and used for monitoring sound decibel values in the public area range of the community in real time and recording and storing sound above 70 decibels;
the data processing unit is used for carrying out feature recognition on the monitored image data according to preset conditions, generating image feature data, collecting the image feature data of the same image monitoring area in different time nodes in the monitored image data, generating an image feature data set, marking the initial time node of the generation of the audio data stored in the audio monitoring device, generating audio feature data, and collecting the audio feature data of the same audio monitoring area in different time nodes to generate an audio feature data set;
the data detection unit is used for acquiring an image characteristic data set according to a preset time frequency, judging the characteristic data in the image characteristic data set according to a preset condition and generating a judgment result, and is also used for leading the audio characteristic data in the audio characteristic data set into a hidden Markov model loaded with a trained time delay neural network to identify the audio characteristic data and outputting an audio identification result which is set as the judgment result;
the data scheduling unit is used for acquiring and identifying the judgment result, setting the data corresponding to the judgment result as abnormal data and marking information and generating early warning information to be output when the judgment result does not accord with the preset condition, and meanwhile, selecting advocated propaganda data according to the preset condition, issuing the advocated propaganda data to a community and playing the advocated propaganda data by public media playing equipment of the community;
and the data statistics unit is used for associating the management evaluation model with the communities in a one-to-one manner, setting evaluation elements and weight values thereof, acquiring abnormal data of the corresponding communities and information marks of the abnormal data, identifying and classifying the information marks or the abnormal data, adjusting the weight values of the evaluation elements after identifying and classifying results correspond to the evaluation elements of the management evaluation model, and outputting an early warning instruction according to a preset condition when the weight values of the evaluation elements are higher than a preset threshold value.
Based on the foregoing technical solution, the present invention further provides a computer-readable storage medium, where at least one instruction, at least one program, a code set, or an instruction set is stored in the storage medium, and the at least one instruction, the at least one program, the code set, or the instruction set is loaded by a processor and executed to implement the above-mentioned community management method based on the internet of things.
By adopting the technical scheme, compared with the prior art, the invention has the beneficial effects that: the scheme ingeniously constructs a community service database, monitoring and feedback of community abnormal behaviors are realized by utilizing community image monitoring equipment and audio monitoring equipment, meanwhile, the collected images and audio are subjected to characteristic positioning, extraction and identification judgment, abnormal conditions are checked out, the method processes by means of a neural network, not only is labor saved, but also the processing efficiency is high, the data response is fast, high-efficiency feedback can be timely provided for community services, effective support is provided for early intervention of the community abnormal conditions, a management evaluation model is further introduced, comprehensive condition evaluation of the community is realized by utilizing weight value adjustment of evaluation elements of the management evaluation model, early warning instructions are triggered by virtue of weight values of different evaluation elements, for example, when the weight value of abnormal animal actions is larger than a preset threshold value, the community service personnel or other security department personnel can know in advance that the community may have the situation that animals do not drag ropes or a large number of wandering animals, and under the situation, the method takes advance measures on the wandering animals, so that the situation that the subsequent wandering animals cause personal injury or environmental pollution to residents in the community is reduced; the condition of abnormal waste treatment is the same, whether the residents in the community have the phenomenon of unreevidential waste disposal or not can be judged by acquiring the weight value of the evaluation element, and on the basis, corresponding propaganda and advocation videos are called for carrying out important propaganda, so that the overall quality of the residents is improved; the management evaluation model can also be used as evaluation reference for community civilization evaluation and the working effect of community service personnel, so that the effect and measures of community management can be more humanized and timely.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a schematic flow chart of a management method according to the present invention;
fig. 2 is a schematic connection diagram of the system of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples. It is to be noted that the following examples are only illustrative of the present invention, and do not limit the scope of the present invention. Similarly, the following examples are only some but not all examples of the present invention, and all other examples obtained by those skilled in the art without any inventive work are within the scope of the present invention.
As shown in fig. 1, the scheme is a community management method based on the internet of things, and the method includes:
s01, constructing a community service database, wherein community monitoring data and advocation and propaganda data are stored in the community service database;
s02, acquiring monitoring image data in a community public area range in real time, and then performing feature recognition on the monitoring image data according to preset conditions to generate image feature data; in addition, the sound decibel value in the community public area range is monitored in real time, sound above 50 or 70 decibels is recorded and stored, meanwhile, after the starting time node of the sound generation is marked, audio characteristic data are generated, then the audio characteristic data of the same audio monitoring area at different time nodes are collected, an audio characteristic data set is generated, wherein the decibel trigger value for recording and storing the sound can be set to be above 70 decibels in the daytime and to be above 50 decibels at night;
s03, collecting image characteristic data of the same image monitoring area at different time nodes in the monitored image data to generate an image characteristic data set;
s04, acquiring an image characteristic data set according to a preset time frequency, and judging the characteristic data in the image characteristic data set according to a preset condition to generate a judgment result; in addition, the step also obtains an audio characteristic data set according to the preset time frequency, judges the audio characteristic data in the audio characteristic data set according to the preset condition and generates a judgment result;
s05, acquiring and identifying a judgment result, setting data corresponding to the judgment result as abnormal data and marking information and generating early warning information to be output when the judgment result does not meet a preset condition, selecting advocated propaganda data according to the preset condition and issuing the advocated propaganda data to a community, and playing the advocated propaganda data by public media playing equipment of the community, wherein the judgment result acquired in the step comprises a judgment result of audio characteristic data and a judgment result of image characteristic data;
s06, constructing a management evaluation model, associating the management evaluation model with a community one to one, setting evaluation elements and weight values of the evaluation elements, acquiring abnormal data of the corresponding community and information marks of the abnormal data, identifying and classifying the information marks or the abnormal data, corresponding identification and classification results with the evaluation elements of the management evaluation model, adjusting the weight values of the evaluation elements, and outputting an early warning instruction according to a preset condition when the weight values of the evaluation elements are higher than a preset threshold value.
In this embodiment, as a preferred alternative, preferably, in S02, the specific method for performing feature recognition on the monitored image data according to the preset condition to generate the image feature data includes:
0201, acquiring images in a monitoring range corresponding to the monitoring images according to a preset time interval, then guiding the images into a positioning neural network for element positioning, then guiding the elements obtained by positioning into a detection neural network for feature identification, obtaining information tags of the positioned elements, and setting the information tags as reference tags, wherein the positioning areas among the positioned elements are not coincident;
0202, extracting image frames in the monitoring image data from the monitoring image data acquired in real time in the community public area range by taking 1 second as a time interval unit, then importing the image frames into a positioning neural network for element positioning, importing the elements obtained by positioning into a detection neural network for feature identification, acquiring information tags of the positioned elements, and setting the information tags as real-time tags;
and S0203, merging the real-time tag and the reference tag, then excluding the intersection part of the real-time tag and the reference tag to obtain a difference tag, and extracting and associating the difference tag and the corresponding element in the image frame to obtain image characteristic data.
Since there may be a large amount of repeated data to be detected in real-time monitoring, in order to keep the feature data captured by the image within a certain time as much as possible while improving the working efficiency, in the present scheme, as a preferred embodiment, preferably, S03 includes: collecting image characteristic data of the same image monitoring area in different time nodes in the monitored image data, selecting and storing the image characteristic data with the same real-time label and the element similarity of more than 90%, and deleting the rest to generate an image characteristic data set; after the duplication removal and simplification are carried out in the mode, the data volume of subsequent processing can be greatly reduced, and meanwhile, most characteristic data can be identified and detected.
In order to improve the security of the data, in the scheme, the image characteristic data set and the audio characteristic data set are further encrypted after being generated, and keys are randomly generated after encryption, the keys are acquired before the image characteristic data set is acquired at S04, namely the encrypted image characteristic data set, the audio characteristic data set and the corresponding keys are transmitted asynchronously, and the keys between different image characteristic data sets and audio characteristic data sets are different, after the image characteristic data set and the audio characteristic data set are acquired, the corresponding keys are analyzed, so that on one hand, the acquired data can be encrypted in time after being acquired, the specific information is directly read after the data outflow is avoided, and therefore adverse effects are caused, on the other hand, the characteristic data set can be prevented from being tampered or replaced, because the encryption strategy of the data is customized, if the data without encryption is adopted for replacement, during subsequent analysis, the situation of information disorder may exist after the data which can be directly read (the replacement data without encryption) is analyzed by the key, so that the encryption, decryption and key transmission mechanism introduced by the scheme is beneficial to improving the safety of the image characteristic data set and the audio characteristic data set.
In this embodiment, as a preferred alternative, S04 preferably includes:
s0411, acquiring an image characteristic data set according to the time frequency of each time of 10-20S, and classifying the characteristic data in the same image characteristic data set into people, animals, vehicles and the like;
s0412, the characteristic data classified into the characters further identifies the articles carried by the hand regions, when the identification result is non-empty, the identified articles and the same image characteristic data are classified into other image characteristic data in a centralized manner to be subjected to secondary matching, when the matching probability is more than 80%, the article abandon is output as a judgment result, and the image characteristic data corresponding to the judgment result is output together;
carrying out article identification on the neck region of the characteristic data classified into the animal, outputting an unconstrained animal as a judgment result when the characteristic data is identified as empty, and outputting image characteristic data corresponding to the judgment result;
further carrying out license plate recognition on the feature data classified into the vehicles, outputting abnormal vehicles as judgment results when the license plates are recognized to be empty, and outputting image feature data corresponding to the judgment results; the identification of the vehicle license plate as empty is carried out under two conditions, one is that the vehicle license plate is not recorded into a community management system and belongs to an unregistered vehicle, the unregistered vehicle possibly brings a certain external risk when entering or exiting a community, and the other is that the vehicle license plate is not hung.
In the identification of the audio feature data, in this embodiment, as a preferred alternative, S04 further includes: s0421, leading the audio characteristic data in the audio characteristic data set into a trained time delay neural network-hidden Markov model to identify the audio characteristic data, outputting an audio identification result, and setting the audio identification result as a judgment result, wherein the audio identification result comprises at least one of abnormal life noise and abnormal animal noise.
Wherein, the training process of the time-delay neural network-hidden Markov model comprises the following steps:
obtaining each information frame of training audio, and the Mel cepstrum characteristic, the tone characteristic and the current speaker characteristic of each information frame;
and training the speech recognition model by taking the Mel cepstrum characteristic, the tone characteristic and the current speaker characteristic of a current information frame, and a plurality of continuous historical information frames before the current information frame and a plurality of continuous future information frames after the current information frame as input until the speech recognition model is converged.
In order to facilitate continuous monitoring and evaluation of the abnormal conditions of the community, in the present scheme, as a preferred optional implementation, in S06, the evaluation element includes at least one of waste abnormal treatment, abnormal animal behavior, abnormal vehicle driving, abnormal living noise, and abnormal animal noise;
when the recognition classification result corresponds to an evaluation element of the management evaluation model, the weight value of the evaluation element is adjusted according to the following formula:
X＝X1+Y1(A)
Wherein, X1Is the current weight value of the evaluation element, X is the weight value of the evaluation element after weight adjustment, Y is the single weight adjustment value, and in addition,X1is 0.5, Y1Is 0.01;
when the weighted value of an evaluation element of a management evaluation model is not adjusted within a preset time length, the weighted value of the evaluation element is adjusted according to the following formula:
X’＝X1’－Y2(II)
Wherein, X1'is the current weight value of the evaluation element, X' is the weight value of the evaluation element after weight adjustment, Y2Adjust the value for the one-time weight, in addition, X1' has an initial value of 0.5, Y2Is 0.005, and in addition, X1The lower limit of' is 0.3, and when 0.3 is reached, the subtraction adjustment is not performed.
By the management evaluation model, the comprehensive condition evaluation of the community is realized by utilizing the weight value adjustment of the evaluation elements, and the early warning instruction is triggered by utilizing the weight values of different evaluation elements, for example, when the weight value of the abnormal action of the animal is greater than a preset threshold value, community service personnel or other security department personnel can know in advance that the situation that the animal is not tied or the number of the wandering animals is large in the community, and in this situation, the method takes the advance measures for the wandering animals, and is beneficial to reducing the situation that the subsequent wandering animals cause personal injury or environmental pollution to community residents; the condition of abnormal waste treatment is the same, whether the residents in the community have the phenomenon of unreevidential waste disposal or not can be judged by acquiring the weight value of the evaluation element, and on the basis, corresponding propaganda and advocation videos are called for carrying out important propaganda, so that the overall quality of the residents is improved; the management evaluation model can also be used as evaluation reference for community civilization evaluation and the working effect of community service personnel, so that the effect and measures of community management can be more humanized and timely; meanwhile, the management evaluation model of the scheme also realizes the effect of automatically reducing (repairing the initial) weight value through a special weight value attenuation mechanism, so that the management evaluation model can indirectly judge that the early abnormal problem is relieved or solved through abnormal result updating without manual intervention, and the data is not required to be reset or corrected.
As shown in fig. 2, based on the above technical solution, the present solution further provides a community management system based on the internet of things, which includes:
the background server is used for constructing a community service database and a management evaluation model, and community monitoring data and advocation propaganda data are stored in the community service database;
the system comprises a plurality of image monitoring devices, a plurality of image monitoring devices and a plurality of image monitoring devices, wherein the image monitoring devices are arranged in a public area range of a community and are used for acquiring monitoring image data in the public area range of the community in real time;
the audio monitoring devices are arranged in a public area range of a community and used for monitoring sound decibel values in the public area range of the community in real time and recording and storing sound above 70 decibels;
the data processing unit is used for carrying out feature recognition on the monitored image data according to preset conditions, generating image feature data, collecting the image feature data of the same image monitoring area in different time nodes in the monitored image data, generating an image feature data set, marking the initial time node of the generation of the audio data stored in the audio monitoring device, generating audio feature data, and collecting the audio feature data of the same audio monitoring area in different time nodes to generate an audio feature data set;
the data detection unit is used for acquiring an image characteristic data set according to a preset time frequency, judging the characteristic data in the image characteristic data set according to a preset condition and generating a judgment result, and is also used for leading the audio characteristic data in the audio characteristic data set into a hidden Markov model loaded with a trained time delay neural network to identify the audio characteristic data and outputting an audio identification result which is set as the judgment result;
the data scheduling unit is used for acquiring and identifying the judgment result, setting the data corresponding to the judgment result as abnormal data and marking information and generating early warning information to be output when the judgment result does not accord with the preset condition, and meanwhile, selecting advocated propaganda data according to the preset condition, issuing the advocated propaganda data to a community and playing the advocated propaganda data by public media playing equipment of the community;
and the data statistics unit is used for associating the management evaluation model with the communities in a one-to-one manner, setting evaluation elements and weight values thereof, acquiring abnormal data of the corresponding communities and information marks of the abnormal data, identifying and classifying the information marks or the abnormal data, adjusting the weight values of the evaluation elements after identifying and classifying results correspond to the evaluation elements of the management evaluation model, and outputting an early warning instruction according to a preset condition when the weight values of the evaluation elements are higher than a preset threshold value.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be substantially or partially implemented in the form of a software product stored in a storage medium and including instructions for causing a computer device (which may be a personal computer, a server, a network device, or the like) or a processor (processor) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
The above description is only a part of the embodiments of the present invention, and not intended to limit the scope of the present invention, and all equivalent devices or equivalent processes performed by the present invention through the contents of the specification and the drawings, or directly or indirectly applied to other related technical fields, are included in the scope of the present invention.
Claims (10)
1. A community management method based on the Internet of things is characterized by comprising the following steps:
s01, constructing a community service database, wherein community monitoring data and advocation and propaganda data are stored in the community service database;
s02, acquiring monitoring image data in a community public area range in real time, and then performing feature recognition on the monitoring image data according to preset conditions to generate image feature data;
s03, collecting image characteristic data of the same image monitoring area at different time nodes in the monitored image data to generate an image characteristic data set;
s04, acquiring an image characteristic data set according to a preset time frequency, and judging the characteristic data in the image characteristic data set according to a preset condition to generate a judgment result;
and S05, acquiring and identifying the judgment result, setting the data corresponding to the judgment result as abnormal data and marking the information and generating early warning information to be output when the judgment result does not meet the preset condition, and simultaneously selecting the advocated propaganda data according to the preset condition and issuing the advocated propaganda data to the community for playing by public media playing equipment of the community.
2. The community management method based on the internet of things of claim 1, wherein S02 further comprises monitoring decibel values of sounds in a community public area range in real time, recording and storing sounds above 50 or 70 decibels, marking a starting time node of occurrence of the sounds, generating audio characteristic data, and then collecting the audio characteristic data of the same audio monitoring area at different time nodes to generate an audio characteristic data set;
s04 further comprises the steps of obtaining an audio characteristic data set, judging the audio characteristic data in the audio characteristic data set according to preset conditions, and generating a judgment result;
the judgment result obtained in S05 includes a judgment result of the audio feature data and a judgment result of the video feature data.
3. The internet of things-based community management method of claim 2, further comprising:
s06, constructing a management evaluation model, associating the management evaluation model with a community one to one, setting evaluation elements and weight values of the evaluation elements, acquiring abnormal data of the corresponding community and information marks of the abnormal data, identifying and classifying the information marks or the abnormal data, corresponding identification and classification results with the evaluation elements of the management evaluation model, adjusting the weight values of the evaluation elements, and outputting an early warning instruction according to a preset condition when the weight values of the evaluation elements are higher than a preset threshold value.
4. The community management method based on the internet of things of claim 3, wherein in the step S02, the feature recognition is performed on the monitored image data according to the preset condition, and the specific method for generating the image feature data is as follows:
0201, acquiring images in a monitoring range corresponding to the monitoring images according to a preset time interval, then guiding the images into a positioning neural network for element positioning, then guiding the elements obtained by positioning into a detection neural network for feature identification, obtaining information tags of the positioned elements, and setting the information tags as reference tags, wherein the positioning areas among the positioned elements are not coincident;
0202, extracting image frames in the monitoring image data from the monitoring image data acquired in real time in the community public area range by taking 1 second as a time interval unit, then importing the image frames into a positioning neural network for element positioning, importing the elements obtained by positioning into a detection neural network for feature identification, acquiring information tags of the positioned elements, and setting the information tags as real-time tags;
and S0203, merging the real-time tag and the reference tag, then excluding the intersection part of the real-time tag and the reference tag to obtain a difference tag, and extracting and associating the difference tag and the corresponding element in the image frame to obtain image characteristic data.
5. The Internet of things-based community management method of claim 4, wherein S03 comprises: collecting image characteristic data of the same image monitoring area in different time nodes in the monitored image data, selecting and storing the image characteristic data with the same real-time label and the element similarity of more than 90%, and deleting the rest to generate an image characteristic data set.
6. The internet-of-things-based community management method of claim 5, wherein S04 comprises:
s0411, acquiring an image characteristic data set according to the time frequency of each time of 10-20S, and classifying the characteristic data in the same image characteristic data set into people, animals, vehicles and the like;
s0412, the characteristic data classified into the characters further identifies the articles carried by the hand regions, when the identification result is non-empty, the identified articles and the same image characteristic data are classified into other image characteristic data in a centralized manner to be subjected to secondary matching, when the matching probability is more than 80%, the article abandon is output as a judgment result, and the image characteristic data corresponding to the judgment result is output together;
carrying out article identification on the neck region of the characteristic data classified into the animal, outputting an unconstrained animal as a judgment result when the characteristic data is identified as empty, and outputting image characteristic data corresponding to the judgment result;
and further carrying out license plate recognition on the feature data classified into the vehicles, outputting abnormal vehicles as judgment results when the license plates are recognized to be empty, and outputting image feature data corresponding to the judgment results together.
7. The internet-of-things-based community management method of claim 6, wherein S04 further comprises: s0421, leading the audio characteristic data in the audio characteristic data set into a trained time delay neural network-hidden Markov model to identify the audio characteristic data, outputting an audio identification result, and setting the audio identification result as a judgment result, wherein the audio identification result comprises at least one of abnormal life noise and abnormal animal noise;
wherein, the training process of the time-delay neural network-hidden Markov model comprises the following steps:
obtaining each information frame of training audio, and the Mel cepstrum characteristic, the tone characteristic and the current speaker characteristic of each information frame;
and training the speech recognition model by taking the Mel cepstrum characteristic, the tone characteristic and the current speaker characteristic of a current information frame, and a plurality of continuous historical information frames before the current information frame and a plurality of continuous future information frames after the current information frame as input until the speech recognition model is converged.
8. The internet-of-things-based community management method of claim 7, wherein in S06, the evaluation element includes one or more of waste abnormal disposal, abnormal animal behavior, abnormal vehicle driving, abnormal living noise, and abnormal animal noise;
when the recognition classification result corresponds to an evaluation element of the management evaluation model, the weight value of the evaluation element is adjusted according to the following formula:
X＝X1+Y1(A)
Wherein, X1Is the current weight value of the evaluation element, X is the weight value of the evaluation element after weight adjustment, Y is the single weight adjustment value, and in addition, X is the current weight value of the evaluation element after weight adjustment1Is 0.5, Y1Is 0.01;
when the weighted value of an evaluation element of a management evaluation model is not adjusted within a preset time length, the weighted value of the evaluation element is adjusted according to the following formula:
X’＝X1’－Y2(II)
Wherein, X1'is the current weight value of the evaluation element, X' is the weight value of the evaluation element after weight adjustment, Y2Adjust the value for the one-time weight, in addition, X1' has an initial value of 0.5, Y2Is 0.005.
9. A community management system based on the Internet of things is characterized by comprising:
the background server is used for constructing a community service database and a management evaluation model, and community monitoring data and advocation propaganda data are stored in the community service database;
the system comprises a plurality of image monitoring devices, a plurality of image monitoring devices and a plurality of image monitoring devices, wherein the image monitoring devices are arranged in a public area range of a community and are used for acquiring monitoring image data in the public area range of the community in real time;
the audio monitoring devices are arranged in a public area range of a community and used for monitoring sound decibel values in the public area range of the community in real time and recording and storing sound above 70 decibels;
the data processing unit is used for carrying out feature recognition on the monitored image data according to preset conditions, generating image feature data, collecting the image feature data of the same image monitoring area in different time nodes in the monitored image data, generating an image feature data set, marking the initial time node of the generation of the audio data stored in the audio monitoring device, generating audio feature data, and collecting the audio feature data of the same audio monitoring area in different time nodes to generate an audio feature data set;
the data detection unit is used for acquiring an image characteristic data set according to a preset time frequency, judging the characteristic data in the image characteristic data set according to a preset condition and generating a judgment result, and is also used for leading the audio characteristic data in the audio characteristic data set into a hidden Markov model loaded with a trained time delay neural network to identify the audio characteristic data and outputting an audio identification result which is set as the judgment result;
the data scheduling unit is used for acquiring and identifying the judgment result, setting the data corresponding to the judgment result as abnormal data and marking information and generating early warning information to be output when the judgment result does not accord with the preset condition, and meanwhile, selecting advocated propaganda data according to the preset condition, issuing the advocated propaganda data to a community and playing the advocated propaganda data by public media playing equipment of the community;
and the data statistics unit is used for associating the management evaluation model with the communities in a one-to-one manner, setting evaluation elements and weight values thereof, acquiring abnormal data of the corresponding communities and information marks of the abnormal data, identifying and classifying the information marks or the abnormal data, adjusting the weight values of the evaluation elements after identifying and classifying results correspond to the evaluation elements of the management evaluation model, and outputting an early warning instruction according to a preset condition when the weight values of the evaluation elements are higher than a preset threshold value.
10. A computer-readable storage medium, characterized in that: the storage medium stores at least one instruction, at least one program, a code set, or a set of instructions, which is loaded and executed by a processor to implement the method for community management based on internet of things according to claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111646967.3A CN114331786A (en) | 2021-12-29 | 2021-12-29 | Community management method and system based on Internet of things |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111646967.3A CN114331786A (en) | 2021-12-29 | 2021-12-29 | Community management method and system based on Internet of things |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114331786A true CN114331786A (en) | 2022-04-12 |
Family
ID=81016208
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111646967.3A Pending CN114331786A (en) | 2021-12-29 | 2021-12-29 | Community management method and system based on Internet of things |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114331786A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114881463A (en) * | 2022-05-07 | 2022-08-09 | 内蒙古云科数据服务股份有限公司 | Intelligent energy management method and system based on Internet of things |
| CN116226527A (en) * | 2023-03-03 | 2023-06-06 | 中浙信科技咨询有限公司 | Digital community treatment method for realizing behavior prediction through resident big data |
| CN117544657A (en) * | 2024-01-09 | 2024-02-09 | 河北万巷互联科技有限公司 | Intelligent community intelligent security method and system based on Internet of things |
| CN116226527B (en) * | 2023-03-03 | 2024-06-07 | 中浙信科技咨询有限公司 | Digital community treatment method for realizing behavior prediction through resident big data |
-
2021
- 2021-12-29 CN CN202111646967.3A patent/CN114331786A/en active Pending
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114881463A (en) * | 2022-05-07 | 2022-08-09 | 内蒙古云科数据服务股份有限公司 | Intelligent energy management method and system based on Internet of things |
| CN114881463B (en) * | 2022-05-07 | 2024-03-08 | 内蒙古云科数据服务股份有限公司 | Intelligent energy management method and system based on Internet of things |
| CN116226527A (en) * | 2023-03-03 | 2023-06-06 | 中浙信科技咨询有限公司 | Digital community treatment method for realizing behavior prediction through resident big data |
| CN116226527B (en) * | 2023-03-03 | 2024-06-07 | 中浙信科技咨询有限公司 | Digital community treatment method for realizing behavior prediction through resident big data |
| CN117544657A (en) * | 2024-01-09 | 2024-02-09 | 河北万巷互联科技有限公司 | Intelligent community intelligent security method and system based on Internet of things |
| CN117544657B (en) * | 2024-01-09 | 2024-03-15 | 河北万巷互联科技有限公司 | Intelligent community intelligent security method and system based on Internet of things |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114331786A (en) | Community management method and system based on Internet of things | |
| CN111914256B (en) | Defense method for machine learning training data under toxic attack | |
| CN108881194B (en) | Method and device for detecting abnormal behaviors of users in enterprise | |
| CN110837615A (en) | Artificial intelligent checking system for advertisement content information filtering | |
| US6975346B2 (en) | Method for suspect identification using scanning of surveillance media | |
| CN110276068B (en) | Legal case analysis method and device | |
| Jansen | Data driven policing in the context of Europe | |
| CN106650799A (en) | Electronic evidence classification extraction method and system | |
| WO2021218312A1 (en) | Method and apparatus for constructing service fraud identification database, and computer device | |
| CN111818198A (en) | Domain name detection method, domain name detection device, equipment and medium | |
| Drossos et al. | Language modelling for sound event detection with teacher forcing and scheduled sampling | |
| CN111276133B (en) | Audio recognition method, system, mobile terminal and storage medium | |
| CN116684128A (en) | Alarm noise reduction method and system based on network attack behavior self-adaptive classification | |
| CN108682157A (en) | Video analysis and method for early warning and system | |
| Yadhunath et al. | Identification of criminal activity hotspots using machine learning to aid in effective utilization of police patrolling in cities with high crime rates | |
| CN112487972B (en) | Anti-riot early warning method, device and storage medium for office places in financial industry | |
| CN111899763B (en) | Nursing identification and evaluation method based on audio analysis | |
| CN111931186B (en) | Software risk identification method and device | |
| CN116865994A (en) | Network data security prediction method based on big data | |
| CN111986681B (en) | Warehouse supervision management system | |
| WO2023281278A1 (en) | Threat assessment system | |
| KR20230025553A (en) | Livestock disease control system and method | |
| Csaba et al. | PPCU Sam: Open-source face recognition framework | |
| CN116975854B (en) | Financial information intelligent storage supervision system and method based on big data | |
| CN118018330B (en) | Data analysis method and system based on artificial intelligence |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |