Intelligent property management system based on Internet of Things

Abstract

The invention discloses an intelligent property management system based on the Internet of Things, relates to the field of property management, and solves the problem of insufficient accuracy of abnormity judgment of existing intelligent property management. The area division module is used for acquiring the monitoring range of the monitoring equipment, performing area division on the community, and performing image analysis on each divided area according to the monitoring information of the monitoring equipment to obtain an area normal image; the exception management module is used for analyzing the monitoring information and extracting event images; collecting a real-time monitoring image; comparing the real-time monitoring image with the event image to obtain an abnormal event, and comparing the real-time monitoring image with the regional normal image to obtain a normal abnormal condition; the real-time response module is used for counting to-be-processed exceptions, planning a processing sequence and transmitting a planning result to processing personnel for orderly processing; according to the invention, the accuracy of abnormity judgment can be effectively improved.

Description

Technical Field

[0001] This invention belongs to the field of property management and involves Internet of Things (IoT) technology, specifically an IoT-based smart property management system. Background Technology

[0002] Existing smart property management systems have the following specific shortcomings when performing management: 1. Existing smart property management systems focus on receiving data, including water and electricity consumption data, vehicle entry and exit data, etc. By intelligently receiving data, the time spent by personnel in data statistics is reduced. This method is effective in terms of intelligence, but it is not comprehensive enough for community management.

[0003] 2. Existing property management has limitations in managing surveillance equipment. Typically, the surveillance data is reviewed after an incident occurs, based on feedback from the parties involved. This method cannot effectively handle abnormal events in the community in a timely manner and cannot guarantee the safety of community residents.

[0004] 3. The current system of handling incidents through property management is not timely enough. Faced with multiple incidents, it is impossible to make comprehensive plans, resulting in a low response rate and failing to guarantee harmony and stability within the community.

[0005] Therefore, we propose an Internet of Things-based smart property management system. Summary of the Invention

[0006] In view of the shortcomings of existing technologies, the purpose of this invention is to provide an Internet of Things-based smart property management system, which aims to improve property management capabilities and ensure the safety of residents in the community.

[0007] To achieve the above objectives, the present invention adopts the following technical solution: a smart property management system based on the Internet of Things, the specific working process of each module is as follows: Data acquisition module: Collects statistics on monitoring equipment within the community and gathers monitoring data from these equipment based on the Internet of Things to obtain monitoring information; Area division module: Obtain the monitoring range of the monitoring equipment, divide the community into areas based on the monitoring range of the monitoring equipment, obtain multiple divided areas, and perform image analysis on each divided area based on the monitoring information of the monitoring equipment to obtain the normal image of the area; Anomaly Management Module: Analyzes monitoring information and extracts event images; collects monitoring data from each monitoring device in real time to obtain real-time monitoring images; compares real-time monitoring images with event images to identify abnormal events; and compares real-time monitoring images with normal area images to identify normal abnormal conditions. Real-time response module: It statistically analyzes abnormal events and normal abnormal conditions to identify anomalies to be processed, plans the processing order of these anomalies, and transmits the planning results to the processing personnel for orderly processing.

[0008] Furthermore, the community is divided into zones, as follows: Obtain the number of surveillance devices in the community and denote the number of surveillance devices as as; obtain the monitoring range of each surveillance device, divide the community into areas based on the monitoring range, and construct the area division centered on the surveillance devices; The monitoring information of the corresponding monitoring equipment in the divided area is extracted. Based on the monitoring information, the monitoring images are extracted and analyzed to obtain the normal image of the area.

[0009] Furthermore, the monitoring equipment is used to construct a zoned area, as detailed below: Obtain the floor plan of the community buildings. Construct a Cartesian coordinate system with the southwest corner of the floor plan as the origin. Obtain the relative position of the monitoring equipment in the Cartesian coordinate system and get the equipment coordinates szb(a). The monitoring range of the monitoring equipment is obtained, and the monitoring range is mapped to a Cartesian coordinate system to obtain the monitoring range bzl(a). The overlapping part of the monitoring range is extracted by the intersection of the monitoring ranges to obtain the overlapping range chf. Based on the overlapping range, multiple overlapping units are extracted; the overlapping units are searched with the monitoring range, and the monitoring devices that overlap with the monitoring range bzl(a) are extracted. The monitoring devices whose monitoring ranges overlap in the overlapping units are denoted as as2; the device coordinates of the above monitoring devices are denoted as szb(a2); the division boundary is preset based on the overlapping units and denoted as hfj; the division ratio hbl(a2) is calculated based on the device coordinates szb(a2) and the division boundary hfj. Set the division ratio hbl(a2) to 1, count the division boundaries that meet the division ratio, and obtain the division area of ​​the monitoring equipment. The normal regional graphs of each region are analyzed based on the division of regions.

[0010] Furthermore, the normal regional graphs for each region are analyzed, as follows: The monitoring information of the monitoring devices in each divided area is acquired, historical monitoring images are extracted based on the monitoring information, and the number of historical monitoring images is counted, denoted as js; the number of pixels in the image is counted as us×vs; the pixel value of each historical monitoring image is acquired, and the pixel value xsz(u,v) is obtained. j ; The historical surveillance images are iterated through, and the mode of the pixel value corresponding to pixel (u, v) is extracted as the normal pixel cxs(u, v). Each pixel is iterated through according to the number of pixels us × vs to obtain normal pixels cxs(1, 1) to cxs(us, vs). These normal pixels cxs(1, 1) to cxs(us, vs) are integrated to construct a predicted normal image. The predicted normal image is then compared with the historical surveillance images to obtain the image similarity value tsz. j ; ; Where: xqz is the range of pixel values; The system iterates through the historical surveillance images, calculating the image pixel values ​​of each historical surveillance image and the predicted normal image to obtain image similarity values ​​tsz1 to tsz. js Extract image similarity values ​​from tsz1 to tsz js The maximum value in the range is used as the corresponding historical monitoring image for the region.

[0011] Furthermore, the abnormal events and normal abnormal conditions are obtained, as detailed below: The monitoring information is analyzed, event images are extracted, the event images are statistically analyzed, an event image list is constructed, the monitoring data of each monitoring device is collected in real time to obtain real-time monitoring images, the real-time monitoring images are compared with the event image list in a loop to obtain event similarity values, a judgment threshold is set, the event similarity values ​​are compared, the corresponding real-time monitoring images are extracted, and abnormal events are identified through manual inspection. Statistical analysis of real-time monitoring images yields multiple real-time monitoring images at different time points. Static features are extracted from these real-time monitoring images at different time points to obtain static feature images. These static feature images are then compared with normal regional images to determine the normal and abnormal conditions.

[0012] Furthermore, the abnormal events were identified as follows: Based on the monitoring information, historical monitoring images are obtained. Through manual judgment, the historical monitoring images that need attention are extracted and special events are marked to obtain event images. The event images are then statistically analyzed to construct an event image list. The system collects monitoring data from each monitoring device in real time to obtain real-time monitoring images. Event images are compared with real-time monitoring images to obtain event similarity values. The event image list is traversed to obtain event similarity values ​​for different events. A preset judgment threshold is used to compare event similarity values. Event warnings are issued based on the comparison results, and real-time monitoring images are output. Events are verified through manual inspection to identify abnormal events. Real-time monitoring images over a period of time are statistically analyzed, and combined with manual inspection, event warnings are screened. The judgment threshold is optimized based on the screening results.

[0013] Furthermore, the event similarity value is calculated as follows: The image pixels of the real-time monitoring image are collected to obtain the real-time pixel value sjk(u, v); the image pixels of the event image are collected to obtain the event pixel value shj(u2, v2); the real-time pixel value and the event pixel value are compared to obtain the event similarity value sxs; The event images are compared with the real-time monitoring images in a loop to extract the maximum event similarity value; similarly, the event image list is traversed, and the real-time monitoring images are compared with different event images to obtain the event similarity value of different events. The event similarity values ​​are counted to construct an event similarity value list. A preset threshold is obtained, and the event similarity value is compared with the preset threshold. If the event similarity value is greater than the preset threshold, the region where the real-time monitoring image is located is extracted, the event is labeled in the region, and the real-time monitoring image is output. If the event similarity value is not greater than the preset threshold, it indicates that no abnormal event has occurred in the region. The preset threshold is optimized based on the output results.

[0014] Furthermore, the preset threshold is optimized as follows: The system receives the output real-time monitoring images to obtain verification images; it judges the verification images by manual verification to identify abnormal events; if no abnormal events exist, it modifies the event labeling results, records the labeling error status, and adjusts and optimizes the preset threshold based on the labeling error status. Real-time monitoring images over a period of time are statistically analyzed. These images are then manually inspected to extract those showing abnormal events, resulting in verification images. The verification images are then compared with the validation images. If the validation images are present in the validation images, it indicates that the abnormal events can be effectively extracted. If they are not present, it indicates that the abnormal events cannot be extracted. Event warnings are then screened, and the judgment threshold is optimized.

[0015] Furthermore, the following steps are taken to obtain information on normal and abnormal conditions: Acquire real-time monitoring images sjd(t) at multiple different time points within an event and construct a time-series monitoring list; analyze the real-time monitoring images based on the time-series monitoring list, overlay the real-time monitoring images sjd(t) at different time points, extract the static features of the real-time monitoring images to obtain a static feature image jtz, compare the static feature image with the normal regional image qct to obtain the regional outlier value qyc; Record the outliers in a region and their corresponding outlier regions to obtain the normal outlier situation.

[0016] Furthermore, the processing order is planned as follows: The number of pending exceptions is denoted as ys. A weight is set for each pending exception to obtain the processing weight qqz(y). The region where the pending exception is located is obtained. The regions where the pending exception is located are connected, and the connection path of the region is recorded. The connection path is normalized and denoted as the connection path ljl. The processing weight and the connection path are combined to calculate the processing judgment value cpd. The system iterates through the processing order of exceptions to be processed, calculates their corresponding processing judgment values, extracts the processing order corresponding to the largest processing judgment value, processes the exceptions to be processed, obtains the connection paths in the processing order, performs path planning, and transmits the planning results to the processing personnel for orderly processing.

[0017] In summary, due to the adoption of the above technical solution, the beneficial effects of the present invention are: 1. This invention collects monitoring data using monitoring equipment, analyzes the community's basic environment through the monitoring data, and specifically analyzes and processes anomalies in the community's basic environment based on real-time monitoring data, effectively ensuring the cleanliness of the community environment; at the same time, it detects abnormal events in a timely manner by manually analyzing abnormal events in combination with real-time monitoring data, ensuring the safety of life and property of community residents.

[0018] 2. This invention divides the community into areas by monitoring the range of the monitoring equipment, deconstructs the community's property management into atomized units, improves the accuracy of anomaly detection by analyzing local areas, and integrates the local analysis structure to ensure the overall harmony and stability of the community.

[0019] 3. This method is based on regional division and accurately locates abnormal events, which facilitates timely handling of abnormal events by property management. At the same time, in the case of multiple abnormal events, the method analyzes the handling order of abnormal events through path planning, taking into account both the timeliness and importance of event handling, and improving the property management's response. Attached Figure Description

[0020] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.

[0021] Figure 1 This is a functional block diagram of the present invention; Figure 2 This is a schematic diagram of the region division in this invention; Figure 3 This is a schematic diagram of the path planning of the present invention; Detailed Implementation It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0022] This application provides an IoT-based smart property management system. The executing entity of this IoT-based smart property management system includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the system provided in this application embodiment: a server, a terminal, etc. In other words, the IoT-based smart property management system can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster.

[0023] Reference Figure 1 The diagram shown is a functional block diagram of an IoT-based smart property management system according to an embodiment of the present invention. In this embodiment, the IoT-based smart property management system includes: a data acquisition module, a regional division module, an anomaly management module, and a real-time response module. The module mentioned in the present invention can also be called a unit, which refers to a series of computer program segments that can be executed by the processor of an electronic device and can perform a fixed function, and are stored in the memory of the electronic device.

[0024] In this embodiment of the invention, the functions of each module / unit are as follows: Data acquisition module: Collects statistics on monitoring equipment within the community and gathers monitoring data from these equipment based on the Internet of Things to obtain monitoring information; It should be noted that the Internet of Things (IoT) is a network system that connects any object to a network through information sensing devices (such as sensors, RFID tags, etc.) according to agreed protocols, enabling information exchange and communication between objects to achieve intelligent identification, positioning, tracking, and monitoring functions. In this invention, "community" refers to the area managed and maintained by property management.

[0025] Area division module: Obtain the monitoring range of the monitoring equipment, divide the community into areas based on the monitoring range of the monitoring equipment, obtain multiple divided areas, and perform image analysis on each divided area based on the monitoring information of the monitoring equipment to obtain the normal image of the area; The specific workflow of the region division module is as follows: Obtain the number of surveillance devices in the community and denote the number of surveillance devices as as; obtain the monitoring range of each surveillance device, divide the community into areas based on the monitoring range, and construct the area division centered on the surveillance devices; The monitoring information of the corresponding monitoring equipment is extracted from the monitoring equipment in the divided area. Based on the monitoring information, the monitoring images are extracted and analyzed to obtain the normal image of the area. The specific process for dividing the region is as follows: Please see Figure 2 Obtain the floor plan of the community buildings (referring to the building distribution based on the community's land area, which visually shows the division of the community into regions). With the southwest corner of the floor plan of the community buildings as the origin, construct a plane rectangular coordinate system, obtain the relative position of the monitoring equipment in the plane rectangular coordinate system, and obtain the equipment coordinates szb(a), where szb(a) represents the equipment coordinates of the a-th monitoring equipment.

[0026] The monitoring range of the monitoring equipment is obtained, and the monitoring range is mapped to a Cartesian coordinate system to obtain the monitoring range bzl(a). The overlapping part of the monitoring range is extracted by the intersection of the monitoring ranges to obtain the overlapping range chf. ; Based on the overlapping range, extract multiple overlapping units (referring to overlapping ranges that are independent of each other in spatial distribution); search the overlapping units and the monitoring range, and extract the monitoring devices that overlap with the monitoring range bzl(a) of the overlapping units, and denote the monitoring devices whose monitoring ranges overlap in the overlapping units as as2; denote the device coordinates of the above monitoring devices as szb(a2); predetermine the division boundary based on the overlapping units, and denote it as hfj; calculate the division ratio hbl(a2) based on the device coordinates szb(a2) and the division boundary hfj. ; Where szb(a2') represents the device coordinates of the a2'th monitoring device; It should be noted that the value of hfj must meet the following conditions: it belongs to the overlapping unit and is within the monitoring range of the a2th monitoring device; Set the division ratio hbl(a2) to 1, and count the division boundaries that meet the division ratio to obtain the division area of ​​the monitoring equipment.

[0027] By dividing the area proportionally, the balance of the area division is ensured, the monitoring range of each monitoring device is relatively consistent, the image processing format is consistent, and the processing effectiveness of the image is improved.

[0028] Analyze the normal regional graphs of each region based on the division of regions; The monitoring information of the monitoring devices in each divided area is acquired. Historical monitoring images are extracted based on the monitoring information, and the number of historical monitoring images is counted, denoted as js. The number of pixels in each image is counted as us×vs; (us×vs represents the number of pixels in the image in rows us and columns vs, where (u, v) is the u-th row and j-th column). The pixel value of each historical monitoring image is obtained as xsz(u, v). j ; pixel value xsz(u,v) j This represents the pixel value corresponding to pixel (u, v) in the j-th historical surveillance image; The historical surveillance images are iterated through, and the mode of the pixel value corresponding to pixel (u, v) is extracted as the normal pixel cxs(u, v). Each pixel is iterated through according to the number of pixels us × vs to obtain normal pixels cxs(1, 1) to cxs(us, vs). These normal pixels cxs(1, 1) to cxs(us, vs) are integrated to construct a predicted normal image. The predicted normal image is then compared with the historical surveillance images to obtain the image similarity value tsz. j ; ; Where: xqz is the range of pixel values, that is, the difference between the maximum and minimum values; It should be noted that, under normal circumstances, the predicted normal image should be consistent with one or more images in the historical monitoring images. The accuracy of the image consistency judgment is ensured by multiplying the images, thereby improving the accuracy of the selection of normal images in the region.

[0029] The system iterates through the historical surveillance images, calculating the image pixel values ​​of each historical surveillance image and the predicted normal image to obtain image similarity values ​​tsz1 to tsz. js Extract image similarity values ​​from tsz1 to tsz js The maximum value in the range is used as the corresponding historical monitoring image for the region.

[0030] Anomaly Management Module: Analyzes monitoring information and extracts event images; collects monitoring data from each monitoring device in real time to obtain real-time monitoring images; compares real-time monitoring images with event images to identify abnormal events; and compares real-time monitoring images with normal area images to identify normal abnormal conditions. The specific workflow of the exception management module is as follows: The monitoring information is analyzed, event images are extracted, the event images are statistically analyzed, an event image list is constructed, the monitoring data of each monitoring device is collected in real time to obtain real-time monitoring images, the real-time monitoring images are compared with the event image list in a loop to obtain event similarity values, a judgment threshold is set, the event similarity values ​​are compared, the corresponding real-time monitoring images are extracted, and abnormal events are identified through manual inspection. Statistical analysis of real-time monitoring images over a period of time yields real-time monitoring images at multiple different time points. Static features are extracted from these real-time monitoring images at different time points to obtain static feature images. These static feature images are then compared with normal regional images to determine the normal and abnormal conditions.

[0031] It should be noted that: a period of time refers to backtracking the monitoring based on the current time, such as viewing the monitoring of the previous 10 minutes. The longer the time period, the more accurate the statistical results.

[0032] The specific process for identifying abnormal events is as follows: Based on the monitoring information, historical monitoring images are obtained. Through manual judgment, historical monitoring images that need attention are extracted, and special events (such as fire, damage to public facilities, and climbing walls) are marked to obtain event images. The event images are then statistically analyzed to construct an event image list. The system collects monitoring data from each monitoring device in real time to obtain real-time monitoring images. Event images are compared with real-time monitoring images to obtain event similarity values. The event image list is traversed to obtain event similarity values ​​for different events. A preset judgment threshold is used to compare event similarity values, and event warnings are issued based on the comparison results. Real-time monitoring images are output, and events are verified through manual inspection to identify abnormal events. Real-time monitoring images over a period of time are statistically analyzed, and combined with manual inspection, event warnings are screened. The judgment threshold is then optimized based on the screening results. The image pixels of the real-time monitoring image are collected to obtain the real-time pixel value sjk(u, v); the image pixels of the event image are collected to obtain the event pixel value shj(u2, v2); the real-time pixel value and the event pixel value are compared to obtain the event similarity value sxs; ; Where: sxs(u,v) represents the event similarity value obtained by comparing the event images with (u,v) as the base point; It should be noted that the impact of different environmental conditions on the similarity value of events was corrected through manual judgment regarding environmental impact. If wall climbing behavior occurs, extract the person's wall climbing action; then perform a global traversal search on the surveillance images to extract consistent behavior. The event images are compared with the real-time monitoring images in a loop to extract the maximum event similarity value; similarly, the event image list is traversed, and the real-time monitoring images are compared with different event images to obtain the event similarity value of different events. The event similarity values ​​are counted to construct an event similarity value list. Obtain a preset threshold, compare the event similarity value with the preset threshold. If the event similarity value is greater than the preset threshold, extract the region where the real-time monitoring image is located, annotate the region with the event, and output the real-time monitoring image. If the event similarity value is not greater than the preset threshold, it indicates that no abnormal event has occurred in the region. It should be noted that the preset threshold is manually set and is adjusted and optimized in real time according to the actual situation. For example, the preset threshold is set to 0.5. If multiple images are output based on the preset threshold and the images are manually identified, and the images are different from the expected results, the preset threshold is increased. At the same time, the images that are not output are checked. If the expected results are not output, the preset threshold is decreased.

[0033] The system receives the output real-time monitoring images to obtain verification images; it judges the verification images by manual verification to identify abnormal events; if no abnormal events exist, it modifies the event labeling results, records the labeling error status, and adjusts and optimizes the preset threshold based on the labeling error status. Statistically analyze real-time monitoring images over a period of time, manually inspect these images, and extract those with abnormal events to obtain verification images. Search between the verification images and the validation images. If the validation images are present in the validation images, it indicates that the abnormal events can be effectively extracted. If they are not present, it indicates that the abnormal events cannot be extracted. Screening for event warnings requires optimization of the judgment threshold. Acquire real-time monitoring images sjd(t) at multiple different time points within a certain period and construct a time-series monitoring list; analyze the real-time monitoring images based on the time-series monitoring list, overlay the real-time monitoring images sjd(t) at different time points, extract the static features of the real-time monitoring images to obtain the static feature image jtz, compare the static feature image with the normal regional image qct to obtain the regional outlier value qyc; ; Where σ is a local minimum value to prevent the denominator from being 0 when the images are consistent; qct(u,v) represents the pixel value corresponding to pixel (u,v) in the normal image of the region; jtz(u,v) represents the pixel value corresponding to pixel (u,v) in the static feature image. Record the outliers in a region and their corresponding outlier regions to obtain the normal outlier situation.

[0034] It should be noted that for outliers in a region, it is only necessary to determine whether the image is consistent or not, to count the inconsistent pixels, and to use the inconsistent pixels to provide feedback on the anomalies, thereby reducing the timely consumption of data. It should be noted that: by comparing images to determine the abnormal situation in each area, and separating abnormal events from area abnormalities, the accuracy of handling different abnormalities is improved; by introducing manual inspection to verify the abnormality judgment, the effective judgment of abnormalities is ensured; this method improves the work efficiency of property management under traditional methods and reduces the false judgment rate of abnormalities in fully automated mode.

[0035] Real-time response module: It statistically analyzes abnormal events and normal abnormal conditions, identifies anomalies to be processed, plans the processing order of anomalies to be processed, and transmits the planning results to the processing personnel for orderly processing. The specific workflow of the real-time response module is as follows: Please see Figure 3 ; Obtain the number of exceptions to be processed, denoted as ys, set a weight for each exception to be processed, and obtain the processing weight qqz(y); Obtain the region where the exception to be processed is located, connect the regions where the exception to be processed is located, record the connection path of the region, and normalize the connection path, denoted as the connection path ljl; Combine the processing weight with the connection path to calculate and obtain the processing judgment value cpd. ; Where: ljl(y, y') is the connection path between the region where the y-th exception to be processed and the region where the y'-th exception to be processed are located; It should be noted that the processing weight refers to the importance of handling anomalies as determined by humans. The weight is set manually, and the higher the value, the more timely the handling is required. For example, if a fire occurs, the processing weight is 1; if there is a blockage on a community road, the processing weight is set to 0.5; if there is garbage on the community road at night, it does not require timely handling, and the processing weight is 0.

[0036] At the same time, the path is normalized to make it consistent with the range of weight values, and the reciprocal of the connection path is taken to make it consistent with the trend of the weight. The smaller the connection path, the earlier it is processed. If a fire and a road blockage occur simultaneously within the community, obtain the connection path between the current location and the fire area (100 meters), the connection path between the current location and the road blockage area (200 meters), and the connection path between the fire area and the road blockage area (200 meters); normalize these paths to obtain the corresponding connection paths as 0.5, 1, 1; obtain the processing weight of 1 for the fire and 0.5 for the community road blockage. The processing judgment value is calculated according to the order of fire handling, and the result is cpd = 1 × (1 / 0.5) + 0.5 × (1 / 1) = 2.5; The processing judgment value is calculated according to the order of handling road congestion, and we get cpd = 0.5 × (1 / 1) + 1 × (1 / 1) = 1.5.

[0037] The system iterates through the processing order of exceptions to be processed, calculates their corresponding processing judgment values, extracts the processing order corresponding to the largest processing judgment value, processes the exceptions to be processed, obtains the connection paths in the processing order, performs path planning, and transmits the planning results to the processing personnel for orderly processing.

[0038] By planning processing paths and weights, the timeliness of anomaly handling can be effectively improved, the efficiency of anomaly response can be increased, and the harmony and stability within the community can be guaranteed.

[0039] Compared to the problems described in the background technology, this invention provides comprehensive monitoring of public areas in the community through monitoring equipment. The community is divided into zones based on the monitoring range, and management is implemented according to these zones. By managing individual zones, the accuracy and real-time performance of management are effectively improved. Furthermore, static features of each zone are extracted to construct a normalized image of that zone. By comparing real-time monitoring images with these normalized images, the community environment is managed, ensuring its stability and cleanliness. Abnormal events are manually extracted to construct an event image list. By comparing these event images with monitoring images, the occurrence of abnormal events within the community is determined, and early warnings are issued based on the occurrence status, ensuring the safety and harmony of the community. Finally, this invention combines the urgency of event handling with the location and path of the event to comprehensively analyze the event handling sequence, optimizing the property management response when multiple events occur, and balancing the timeliness and importance of event handling. Therefore, the IoT-based smart property management system provided by this invention can enhance property management capabilities and strengthen the security of the managed community.

[0040] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0041] Finally, it should be noted that deleting any one of the above embodiments does not affect the technical solutions of other embodiments. The above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A smart property management system based on the Internet of Things, characterized in that, include: Data acquisition module: Collects statistics on monitoring equipment within the community and gathers monitoring data from these equipment based on the Internet of Things to obtain monitoring information; Area division module: Obtain the monitoring range of the monitoring equipment, divide the community into areas based on the monitoring range of the monitoring equipment, obtain multiple divided areas, and perform image analysis on each divided area based on the monitoring information of the monitoring equipment to obtain the normal image of the area; Anomaly management module: Analyzes monitoring information and extracts event images; Real-time monitoring data from each monitoring device is collected to obtain real-time monitoring images; By comparing real-time monitoring images with event images, abnormal events can be identified; by comparing real-time monitoring images with normal regional images, abnormal normal conditions can be identified. Real-time response module: It statistically analyzes abnormal events and normal abnormal conditions to identify anomalies to be processed, plans the processing order of these anomalies, and transmits the planning results to the processing personnel for orderly processing.

2. The smart property management system based on the Internet of Things according to claim 1, characterized in that, The community is divided into zones as follows: Obtain the number of surveillance devices in the community and denote the number of surveillance devices as as; obtain the monitoring range of each surveillance device, divide the community into areas based on the monitoring range, and construct the area division centered on the surveillance devices; The monitoring information of the corresponding monitoring equipment in the divided area is extracted. Based on the monitoring information, the monitoring images are extracted and analyzed to obtain the normal image of the area.

3. The smart property management system based on the Internet of Things according to claim 2, characterized in that, The monitoring equipment is used to create a zoned layout, as detailed below: Obtain the floor plan of the community buildings. Construct a Cartesian coordinate system with the southwest corner of the floor plan as the origin. Obtain the relative position of the monitoring equipment in the Cartesian coordinate system and get the equipment coordinates szb(a). The monitoring range of the monitoring equipment is obtained, and the monitoring range is mapped to a Cartesian coordinate system to obtain the monitoring range bzl(a). The overlapping part of the monitoring range is extracted by the intersection of the monitoring ranges to obtain the overlapping range chf. Based on the overlapping range, multiple overlapping units are extracted; the overlapping units are searched with the monitoring range, and the monitoring devices that overlap with the monitoring range bzl(a) are extracted. The monitoring devices whose monitoring ranges overlap in the overlapping units are denoted as as2; the device coordinates of the above monitoring devices are denoted as szb(a2); the division boundary is preset based on the overlapping units and denoted as hfj; the division ratio hbl(a2) is calculated based on the device coordinates szb(a2) and the division boundary hfj. Set the division ratio hbl(a2) to 1, count the division boundaries that meet the division ratio, and obtain the division area of ​​the monitoring equipment. The normal regional graphs of each region are analyzed based on the division of regions.

4. The smart property management system based on the Internet of Things according to claim 3, characterized in that, The analysis of the normal regional graphs for each region is as follows: The monitoring information of the monitoring devices in each divided area is acquired, historical monitoring images are extracted based on the monitoring information, and the number of historical monitoring images is counted, denoted as js; the number of pixels in the image is counted as us×vs; the pixel value of each historical monitoring image is acquired, and the pixel value xsz(u,v) is obtained. j ; The historical surveillance images are iterated through, and the mode of the pixel value corresponding to pixel (u, v) is extracted as the normal pixel cxs(u, v). Each pixel is iterated through according to the number of pixels us × vs to obtain normal pixels cxs(1, 1) to cxs(us, vs). These normal pixels cxs(1, 1) to cxs(us, vs) are integrated to construct a predicted normal image. The predicted normal image is then compared with the historical surveillance images to obtain the image similarity value tsz. j ; ； Where: xqz is the range of pixel values; The system iterates through the historical surveillance images, calculating the image pixel values ​​of each historical surveillance image and the predicted normal image to obtain image similarity values ​​tsz1 to tsz. js Extract image similarity values ​​from tsz1 to tsz js The maximum value in the range is used as the corresponding historical monitoring image for the region.

5. A smart property management system based on the Internet of Things according to claim 1, characterized in that, The details of abnormal events and normal abnormal conditions are as follows: The monitoring information is analyzed, event images are extracted, the event images are statistically analyzed, an event image list is constructed, the monitoring data of each monitoring device is collected in real time to obtain real-time monitoring images, the real-time monitoring images are compared with the event image list in a loop to obtain event similarity values, a judgment threshold is set, the event similarity values ​​are compared, the corresponding real-time monitoring images are extracted, and abnormal events are identified through manual inspection. Statistical analysis of real-time monitoring images yields multiple real-time monitoring images at different time points. Static features are extracted from these real-time monitoring images at different time points to obtain static feature images. These static feature images are then compared with normal regional images to determine the normal and abnormal conditions.

6. A smart property management system based on the Internet of Things according to claim 5, characterized in that, The abnormal event was identified as follows: Based on the monitoring information, historical monitoring images are obtained. Through manual judgment, the historical monitoring images that need attention are extracted and special events are marked to obtain event images. The event images are then statistically analyzed to construct an event image list. The monitoring data of each monitoring device is collected in real time to obtain real-time monitoring images. The event images are compared with the real-time monitoring images to obtain event similarity values. The event image list is traversed to obtain event similarity values ​​for different events. The system presets a judgment threshold, compares event similarity values, issues event warnings based on the comparison results, outputs real-time monitoring images, and verifies events through manual inspection to identify abnormal events. By analyzing real-time monitoring images over a period of time and combining them with manual inspection, event warnings are screened, and the judgment thresholds are optimized based on the screening results.

7. A smart property management system based on the Internet of Things according to claim 6, characterized in that, The similarity score of events is calculated as follows: The image pixels of the real-time monitoring image are acquired to obtain the real-time pixel value sjk(u, v); the image pixels of the event image are acquired to obtain the event pixel value shj(u2, v2). The event similarity value sxs is obtained by comparing the real-time pixel value with the event pixel value. The event images are compared with the real-time monitoring images in a loop to extract the event similarity value; Similarly, the event image list is traversed, and the real-time monitoring images are compared with different event images to obtain the event similarity values ​​of different events. The event similarity values ​​are then statistically analyzed to construct an event similarity value list. A preset threshold is obtained, and the event similarity value is compared with the preset threshold. If the event similarity value is greater than the preset threshold, the region where the real-time monitoring image is located is extracted, the event is labeled in the region, and the real-time monitoring image is output. If the event similarity value is not greater than the preset threshold, it indicates that no abnormal event has occurred in the region. The preset threshold is optimized based on the output results.

8. A smart property management system based on the Internet of Things according to claim 7, characterized in that, The preset threshold is optimized as follows: The system receives the output real-time monitoring images to obtain verification images; it judges the verification images by manual verification to identify abnormal events; if no abnormal events exist, it modifies the event labeling results, records the labeling error status, and adjusts and optimizes the preset threshold based on the labeling error status. Statistically analyze real-time monitoring images over a period of time, manually inspect the real-time monitoring images, extract real-time monitoring images with abnormal events, and obtain verification images; The verification image and the validation image are searched. If the validation image exists in the validation image, it means that the abnormal event can be effectively extracted. If it does not exist, it means that the abnormal event cannot be extracted. The event warning is screened and the judgment threshold is optimized.

9. A smart property management system based on the Internet of Things according to claim 5, characterized in that, To obtain information on normal and abnormal conditions, see the following: Acquire real-time monitoring images sjd(t) at multiple different time points within an event and construct a time-series monitoring list; analyze the real-time monitoring images based on the time-series monitoring list, overlay the real-time monitoring images sjd(t) at different time points, extract the static features of the real-time monitoring images to obtain a static feature image jtz, compare the static feature image with the normal regional image qct to obtain the regional outlier value qyc; Record the outliers in a region and their corresponding outlier regions to obtain the normal outlier situation.

10. A smart property management system based on the Internet of Things according to claim 1, characterized in that, The processing order is planned as follows: The number of pending exceptions is denoted as ys. A weight is set for each pending exception to obtain the processing weight qqz(y). The region where the pending exception is located is obtained. The regions where the pending exception is located are connected, and the connection path of the region is recorded. The connection path is normalized and denoted as the connection path ljl. The processing weight and the connection path are combined to calculate the processing judgment value cpd. The system iterates through the processing order of exceptions to be processed, calculates their corresponding processing judgment values, extracts the processing order corresponding to the largest processing judgment value, processes the exceptions to be processed, obtains the connection paths in the processing order, performs path planning, and transmits the planning results to the processing personnel for orderly processing.

Images