Intelligent community target identification and tracking method based on big data

Abstract

The invention relates to the technical field of smart communities, and discloses a smart community target identification tracking method based on big data, which comprises the following steps: constructing an omnibearing monitoring network in a community, collecting real-time video and sensor data in the community, obtaining feature data of a target through an edge computing technology, and carrying out target identification tracking on the feature data; transmitting the target feature data to a cloud server through a network, identifying and classifying the target feature data, judging the type and identity of the target, analyzing the current behavior of the target according to a preset target historical behavior pattern library, and determining whether the behavior is abnormal or not; and if the behavior of the target is not matched with the normal mode in the historical behavior mode library, judging that the behavior is an abnormal behavior, and triggering an early warning mechanism. According to the invention, intelligent identification, dynamic tracking and timely early warning of the abnormal behavior of the smart community are realized, and the intelligent level and response efficiency of community safety management are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of smart communities, and in particular to a smart community target identification and tracking method based on big data. Background Art

[0002] A large number of high-definition cameras and various sensors are deployed in smart communities, generating massive amounts of surveillance data. How to accurately identify targets of interest from these data and continuously track their whereabouts is a huge technical challenge. First, there are many types of surveillance equipment in the community, and the data formats are different, so unified data cleaning and standardization are required. Second, the features of targets of interest are diverse, such as face, body shape, gait, etc., and they will change under different environmental conditions (such as lighting, angle, occlusion). Third, the judgment of target behavior needs to be based on continuous observation. How to correlate and analyze multi-source heterogeneous data in time and space to construct the target's behavior trajectory is also a complex problem. Finally, the mobility of people in the community is high, and target identification and behavior judgment need to be completed quickly in massive data, which puts high demands on the timeliness of the algorithm. In summary, the all-weather monitoring and real-time analysis of smart communities requires research on multiple technical levels such as data processing, feature engineering, correlation analysis, and high-performance computing. The core is how to use big data technology to automatically mine valuable information from data, form explainable and traceable analysis results, and provide strong support for community management. Summary of the invention

[0003] The purpose of the present invention is to overcome one or more of the above-mentioned existing technical problems and provide a smart community target identification and tracking method based on big data.

[0004] To achieve the above objectives, the present invention provides a method for identifying and tracking smart community targets based on big data, comprising: Deploy cameras and sensors in the community, build a community monitoring network, and collect video and sensor data in the community in real time; Based on the video and sensor data, preliminary target detection and feature extraction are performed on the front-end device through edge computing methods to obtain the target's feature data; The target’s feature data is transmitted to the cloud server through the network, and the target’s feature data is identified and classified using a deep learning model to determine the target’s type and identity; According to the preset target historical behavior pattern library, the current behavior of the target is analyzed to determine whether its behavior is abnormal. If the target's behavior does not match the normal pattern in the target historical behavior pattern library, it is judged as abnormal behavior and the early warning mechanism is triggered; After adjusting the parameters of the community monitoring network, the target’s video and sensor data are re-acquired, and the target is continuously tracked and its behavior analyzed. If the target’s abnormal behavior persists or escalates, the warning level is further increased according to the preset strategy until the abnormal situation is eliminated; Regularly summarize and analyze the data collected by the community monitoring network, use machine learning algorithms to explore behavioral patterns and abnormal laws, and optimize and adjust the community's monitoring strategies and early warning mechanisms based on the analysis results.

[0005] According to one aspect of the present invention, cameras and sensors are installed at key locations according to the community layout plan to build a community monitoring network covering the entire community; The collected video streams and sensor data are transmitted in real time via the optical fiber network to the server cluster of the community data center for storage; Use the visual algorithm to analyze the video stream in real time, identify target objects such as people and vehicles, and extract their motion trajectory information; Use time series analysis methods to detect abnormal fluctuations in environmental data such as temperature, humidity, and noise collected by sensors; Use learning models to identify abnormal behaviors in videos and trigger security warnings if suspicious situations are detected; Based on video analysis and sensor data, draw a heat map of community traffic to determine the density of people and provide a basis for decision-making in community management; Through the data visualization platform, various monitoring data and analysis results are displayed centrally, and the community security situation is monitored in real time.

[0006] According to one aspect of the present invention, real-time video and sensor data are acquired and the data are transmitted to a front-end device; On the front-end device, edge computing technology is used to process data in real time, and the target is detected in the video and sensor data through the target detection algorithm; For the detected target, a feature extraction algorithm is used to extract the features of the target, and the feature data of the target is obtained according to the extracted target features; Compare the target feature data with the preset feature template to determine the target category, and determine the processing strategy based on the target category and feature data.

[0007] According to one aspect of the present invention, an image acquisition device is used to acquire a target image, and the target image is preprocessed to obtain feature data; Extract key feature points from the preprocessed image and generate feature vectors; According to the preset data transmission protocol, the feature vector data is packaged and encrypted and then transmitted to the cloud server through the network; Decrypt the received feature vector data on the cloud server and convert the data format; Perform forward propagation calculation on the converted feature vector to obtain the target category probability distribution; The probability distribution result is compared with the preset recognition threshold. If the maximum probability value exceeds the threshold, it is determined to be that category, otherwise it is determined to be an unknown category; For the identified target category, the corresponding target identity database is called for feature matching to obtain the target's specific identity information.

[0008] According to one aspect of the present invention, a sensor is used to monitor the target behavior in real time to obtain the target behavior data; The collected original target behavior data is cleaned and standardized through data preprocessing to obtain standardized target behavior feature data; Based on the standardized target behavior feature data, clustering algorithms are used to perform cluster analysis on historical behavior data to establish a target historical behavior pattern library; For the current target behavior feature data, a matching algorithm is used to calculate the similarity with the patterns in the target historical behavior pattern library to obtain a matching score; If the matching score is lower than the preset threshold, it is judged as abnormal behavior and triggers abnormal behavior warning; Classify abnormal behaviors and determine warning levels based on their specific characteristics and the degree of deviation from normal patterns; For different warning levels, the corresponding warning mechanism and handling process are triggered, and targeted prevention and control measures are taken.

[0009] According to one aspect of the present invention, based on the video and sensor data collected by the monitoring device, the target in the video is identified and located, and information such as the position and motion trajectory of the target is obtained; By continuously tracking the target, the target's behavior characteristics at different time points are obtained, and compared with the pre-established normal behavior model to determine whether the target's behavior is abnormal; If abnormal behavior of the target is detected, the parameters of the monitoring equipment are dynamically adjusted according to the severity and duration of the abnormal behavior; The acquired video and sensor data are input into the recognition model, and the deep learning algorithm is used to determine whether the target's abnormal behavior persists or escalates; If the target's abnormal behavior persists or escalates, the warning level will be raised according to the preset strategy until the abnormal situation is eliminated; After the abnormal situation is eliminated, the parameters of the monitoring equipment are dynamically adjusted according to the recovery of the target; Record the target's behavioral data and abnormal situations as training samples for optimizing abnormal behavior recognition models and early warning strategies to optimize the models.

[0010] According to one aspect of the present invention, community data collected by monitoring equipment is obtained, and the data is preprocessed, including data cleaning and feature extraction, to obtain a normalized data set; According to the characteristics of the data set, select the machine learning algorithm and establish a mining model for community behavior patterns and abnormal laws; Use the established mining model to train and test the data set; Apply the trained model to the real-time collected monitoring data to predict and identify the behavior patterns and abnormal situations in the community in real time; Dynamically adjust the monitoring strategy of the smart community based on the results predicted by the model; When abnormal behaviors or events are identified in the community, the early warning mechanism is triggered in time to notify relevant personnel to handle the situation; Continuously monitor and evaluate the system's operation and continuously improve the monitoring strategy based on feedback.

[0011] To achieve the above objectives, the present invention provides a smart community target recognition and tracking system based on big data, comprising: Data monitoring and acquisition module: deploy cameras and sensors in the community, build a community monitoring network, and collect video and sensor data in the community in real time; Feature data acquisition module: Based on video and sensor data, preliminary target detection and feature extraction are performed on the front-end device through edge computing methods to obtain the target's feature data; Identity identification module: transmits the target's feature data to the cloud server through the network, uses the deep learning model to identify and classify the target's feature data, and determines the target's type and identity; Community early warning module: Analyze the current behavior of the target according to the preset target historical behavior pattern library to determine whether its behavior is abnormal. If the target's behavior does not match the normal pattern in the target historical behavior pattern library, it is judged as abnormal behavior and triggers the early warning mechanism; Abnormal monitoring module: After adjusting the parameters of the community monitoring network, the target's video and sensor data are re-acquired, and the target is continuously tracked and analyzed for behavior. If the target's abnormal behavior persists or escalates, the warning level is further increased according to the preset strategy until the abnormal situation is eliminated; Maintenance and update module: Regularly summarize and analyze the data collected by the community monitoring network, use machine learning algorithms to mine behavioral patterns and abnormal laws, and optimize and adjust the community's monitoring strategies and early warning mechanisms based on the analysis results.

[0012] To achieve the above-mentioned purpose, the present invention provides an electronic device, including a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, the above-mentioned method for target identification and tracking in a smart community based on big data is implemented.

[0013] To achieve the above-mentioned purpose, the present invention provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned method for identifying and tracking smart community targets based on big data is implemented.

[0014] Based on this, the beneficial effects of the present invention are: by deploying high-definition cameras and sensors in the community, an all-round monitoring network is built to collect real-time data; Use edge computing technology to perform preliminary target detection and feature extraction on the front-end device, and transmit feature data to the cloud server for deep learning recognition and classification; According to the preset historical behavior pattern library, analyze whether the target's current behavior is abnormal, and trigger an early warning if it does not match; Dynamically adjust monitoring parameters, continuously track and analyze target behaviors, and upgrade warning levels based on circumstances; By regularly summarizing and analyzing data, we can mine behavioral patterns and abnormal laws, and optimize monitoring strategies and early warning mechanisms; It has realized intelligent identification, dynamic tracking and timely warning of abnormal behaviors in smart communities, and improved the intelligence level and response efficiency of community security management. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a flow chart of a method for identifying and tracking a smart community target based on big data according to an exemplary embodiment; Figure 2 The present invention is a flow chart of a smart community target identification and tracking system based on big data according to an exemplary embodiment. DETAILED DESCRIPTION

[0016] The present invention will now be discussed with reference to exemplary embodiments. It should be understood that the embodiments discussed are only for enabling those skilled in the art to better understand and thereby implement the present invention, rather than implying any limitation on the scope of the present invention.

[0017] As used herein, the term “including” and variations thereof are to be interpreted as open-ended terms meaning “including but not limited to.” The term “based on” is to be interpreted as “based, at least in part, on,” and the terms “one embodiment” and “an embodiment” are to be interpreted as “at least one embodiment.”

[0018] According to one embodiment of the present invention, Figure 1is a flow chart of a method for identifying and tracking a smart community target based on big data according to an exemplary embodiment. Figure 1 As shown, to achieve the above purpose, the present invention provides a smart community target identification and tracking method based on big data, including: Deploy cameras and sensors in the community, build a community monitoring network, and collect video and sensor data in the community in real time; Based on the video and sensor data, preliminary target detection and feature extraction are performed on the front-end device through edge computing methods to obtain the target's feature data; The target’s feature data is transmitted to the cloud server through the network, and the target’s feature data is identified and classified using a deep learning model to determine the target’s type and identity; According to the preset target historical behavior pattern library, the current behavior of the target is analyzed to determine whether its behavior is abnormal. If the target's behavior does not match the normal pattern in the target historical behavior pattern library, it is judged as abnormal behavior and the early warning mechanism is triggered; After adjusting the parameters of the community monitoring network, the target’s video and sensor data are re-acquired, and the target is continuously tracked and its behavior analyzed. If the target’s abnormal behavior persists or escalates, the warning level is further increased according to the preset strategy until the abnormal situation is eliminated; Regularly summarize and analyze the data collected by the community monitoring network, use machine learning algorithms to explore behavioral patterns and abnormal laws, and optimize and adjust the community's monitoring strategies and early warning mechanisms based on the analysis results.

[0019] According to one embodiment of the present invention, cameras and sensors are installed at key locations according to the community layout plan to build a community monitoring network covering the entire community; The collected video streams and sensor data are transmitted in real time via the optical fiber network to the server cluster of the community data center for storage; Use the visual algorithm to analyze the video stream in real time, identify target objects such as people and vehicles, and extract their motion trajectory information; Use time series analysis methods to detect abnormal fluctuations in environmental data such as temperature, humidity, and noise collected by sensors; Use learning models to identify abnormal behaviors in videos and trigger security warnings if suspicious situations are detected; Based on video analysis and sensor data, draw a heat map of community traffic to determine the density of people and provide a basis for decision-making in community management; Through the data visualization platform, various monitoring data and analysis results are displayed centrally, and the community security situation is monitored in real time.

[0020] According to one embodiment of the present invention, real-time video and sensor data are acquired and transmitted to a front-end device; On the front-end device, edge computing technology is used to process data in real time, and the target is detected in the video and sensor data through the target detection algorithm; For the detected target, a feature extraction algorithm is used to extract the features of the target, and the feature data of the target is obtained according to the extracted target features; Compare the target feature data with the preset feature template to determine the target category, and determine the processing strategy based on the target category and feature data.

[0021] According to one embodiment of the present invention, an image acquisition device is used to acquire a target image, and the target image is preprocessed to obtain feature data; Extract key feature points from the preprocessed image and generate feature vectors; According to the preset data transmission protocol, the feature vector data is packaged and encrypted and then transmitted to the cloud server through the network; Decrypt the received feature vector data on the cloud server and convert the data format; Perform forward propagation calculation on the converted feature vector to obtain the target category probability distribution; The probability distribution result is compared with the preset recognition threshold. If the maximum probability value exceeds the threshold, it is determined to be that category, otherwise it is determined to be an unknown category; For the identified target category, the corresponding target identity database is called for feature matching to obtain the target's specific identity information.

[0022] According to one embodiment of the present invention, a sensor is used to monitor the target behavior in real time to obtain the target behavior data; The collected original target behavior data is cleaned and standardized through data preprocessing to obtain standardized target behavior feature data; Based on the standardized target behavior feature data, clustering algorithms are used to perform cluster analysis on historical behavior data to establish a target historical behavior pattern library; For the current target behavior feature data, a matching algorithm is used to calculate the similarity with the patterns in the target historical behavior pattern library to obtain a matching score; If the matching score is lower than the preset threshold, it is judged as abnormal behavior and triggers abnormal behavior warning; Classify abnormal behaviors and determine warning levels based on their specific characteristics and the degree of deviation from normal patterns; For different warning levels, the corresponding warning mechanism and handling process are triggered, and targeted prevention and control measures are taken.

[0023] According to one embodiment of the present invention, based on the video and sensor data collected by the monitoring device, the target in the video is identified and located, and information such as the position and motion trajectory of the target is obtained; By continuously tracking the target, the target's behavior characteristics at different time points are obtained, and compared with the pre-established normal behavior model to determine whether the target's behavior is abnormal; If abnormal behavior of the target is detected, the parameters of the monitoring equipment are dynamically adjusted according to the severity and duration of the abnormal behavior; The acquired video and sensor data are input into the recognition model, and the deep learning algorithm is used to determine whether the target's abnormal behavior persists or escalates; If the target's abnormal behavior persists or escalates, the warning level will be raised according to the preset strategy until the abnormal situation is eliminated; After the abnormal situation is eliminated, the parameters of the monitoring equipment are dynamically adjusted according to the recovery of the target; Record the target's behavioral data and abnormal situations as training samples for optimizing abnormal behavior recognition models and early warning strategies to optimize the models.

[0024] According to one embodiment of the present invention, community data collected by monitoring equipment is obtained, and the data is preprocessed, including data cleaning and feature extraction, to obtain a normalized data set; According to the characteristics of the data set, select the machine learning algorithm and establish a mining model for community behavior patterns and abnormal laws; Use the established mining model to train and test the data set; Apply the trained model to the real-time collected monitoring data to predict and identify the behavior patterns and abnormal situations in the community in real time; Dynamically adjust the monitoring strategy of the smart community based on the results predicted by the model; When abnormal behaviors or events are identified in the community, the early warning mechanism is triggered in time to notify relevant personnel to handle the situation; Continuously monitor and evaluate the system's operation and continuously improve the monitoring strategy based on feedback.

[0025] According to one embodiment of the present invention, according to the community layout plan, high-definition cameras are installed in key locations such as main roads, parking lots, and public activity areas in the community, and network cameras with a resolution of 4K are selected, and one monitoring point is arranged every 100 meters on average. At the same time, multi-parameter sensors such as temperature, humidity, PM5, etc. are deployed near environmentally sensitive areas such as gardens and pools, and the sensor sampling frequency is set to once per minute. The internal fiber optic network of the community is constructed using EPON technology to ensure that the video transmission bandwidth is not less than 100Mbps and the sensor data transmission bandwidth is not less than 10Mbps. The video stream and sensor data are converged to the community data center in real time through the fiber optic network, and the server cluster with a distributed storage architecture is used to store and manage the data. The total storage capacity reaches 100TB, which meets the demand for continuous storage for 30 days. The video analysis module uses the YOLOv5 target detection algorithm to process the video frames in real time, and can analyze 30 frames of images per second on average, identify targets such as people and vehicles in the picture, and extract motion trajectories through target tracking. The sensor data analysis module is based on the LSTM time series prediction model. It obtains the normal fluctuation patterns of temperature, humidity, and PM5 through training with a large amount of historical data. If the actual data is detected to exceed the predicted value by 20%, it is judged as abnormal and an early warning is issued. The abnormal behavior recognition module uses a behavior recognition algorithm based on a graph convolutional network to identify abnormal behaviors such as fighting and theft. It draws a heat map of community traffic by clustering analysis of the movement trajectories of people, and divides the density of people into three levels according to the heat value. Finally, various types of data and analysis results are presented through a self-developed data visualization platform, which supports map mode and chart mode, and provides abnormal event playback, data query and other functions for community security managers.

[0026] According to one embodiment of the present invention, in terms of real-time video and sensor data acquisition, high-definition cameras and infrared sensors can be used to collect video data at a speed of 30 frames per second, and environmental data can be collected through sensors such as temperature, humidity, and pressure. The acquired data is transmitted to the front-end device through the 5G network, and the transmission rate can reach 10Gbps. The front-end device adopts edge computing technology and uses high-performance processors such as NVIDIA Jetson Xavier NX to analyze and process data in real time, with a computing power of up to 21 TOPS. Through advanced target detection algorithms such as YOLOv5, targets are quickly and accurately detected in video and sensor data, with an average detection time of less than 50ms. For the detected targets, a feature extraction algorithm based on a convolutional neural network is used to extract the features of the target, and the extraction speed can reach 100 times / second. The extracted target feature data and the preset feature template are calculated for cosine similarity, and the similarity is greater than 9, which is determined to be the same category. According to the determined target category and feature data, the system can automatically determine the subsequent processing strategy, such as storing the data in a cloud database or performing real-time alarms, etc., to achieve full-process automated processing. By adopting the above technical solutions, efficient collection, transmission, processing and analysis of real-time video and sensor data can be achieved, greatly improving the intelligence level and response speed of the system, and providing reliable support for various application scenarios.

[0027] According to one embodiment of the present invention, a high-definition camera is used to obtain a target image, and the resolution is set to 1920×1080. The image is pre-processed by grayscale, histogram equalization, etc. to improve the image quality. The 128-dimensional feature vector of the image is extracted by the SIFT algorithm, and each feature point contains information such as position, scale, and direction. The feature vector data is packaged in the protobuf format, encrypted by the AES-256 algorithm, and transmitted to the cloud via the HTTPS protocol. The cloud server uses the same algorithm to decrypt the data and convert the protobuf format into a numpy array. The feature vector is classified using the pre-trained ResNet-50 model, and the accuracy of the model on the ImageNet data set is more than 95%. According to the probability distribution output by the softmax layer, the threshold is set to 8. If it exceeds, it is determined to be the corresponding category, otherwise it is determined to be an unknown category. For the identified target category, the Euclidean distance metric is used to match the feature template in the database. If the distance is less than 6, it is considered to be the same target. Finally, the category, identity information, and similarity score of the target are output to provide data support for subsequent business applications.

[0028] According to one embodiment of the present invention, infrared sensors and cameras are deployed in the target area to collect target behavior data in real time. The collected raw data is denoised and normalized to obtain a standardized behavior feature vector, such as position coordinates (x, y), moving speed v, residence time t, etc. The historical behavior data is clustered using the K-means clustering algorithm to obtain k normal behavior patterns, each pattern corresponding to the center point of a set of feature vectors. The Euclidean distance between the current behavior feature vector and the center point of each pattern is calculated. If the minimum distance is greater than a threshold (such as 8), it is determined to be an abnormal behavior. Abnormal behavior is divided into three warning levels: low, medium, and high according to the degree of deviation from the normal pattern through the C5 decision tree algorithm. Different warning levels trigger corresponding warning information push and control measures, such as sending prompt information for low-level warnings, dispatching patrol personnel for medium-level warnings, and launching emergency plans for high-level warnings. Through real-time data analysis and pattern matching, combined with abnormal behavior graded warning disposal, a complete behavior anomaly detection and response process is formed to improve the intelligent level of target behavior supervision.

[0029] According to one embodiment of the present invention, a high-definition camera is used to collect video data of a target area, and an infrared sensor is used to collect thermal imaging data of the target. The collected video frame data is input into a YOLOv5 target detection model, and the target in the video is identified through model reasoning, and the position coordinates, motion trajectory and other information of the target are obtained. The Kalman filter algorithm is used to predict and update the position of the target to achieve continuous tracking of the target. During the tracking process, the target's motion speed, direction, trajectory curvature and other behavioral characteristics are extracted, and compared with the normal behavior model established in advance by the clustering algorithm. By setting a threshold, it is determined whether the target behavior is abnormal. If abnormal behavior is detected in the target, such as a sudden increase in motion speed, a large deviation in the motion trajectory, etc., the parameters of the monitoring device are dynamically adjusted according to the severity and duration of the abnormal behavior. For example, when the abnormal behavior lasts for more than 5 seconds and the motion speed exceeds 50% of the normal value, the frame rate of the video is increased to 60 frames / second and the resolution is increased to 4K to obtain higher quality target data. The high-definition video and infrared sensor data are input into the abnormal behavior recognition model, which uses a three-dimensional convolutional neural network to determine whether the target's abnormal behavior persists or escalates by jointly modeling the video frame sequence and thermal imaging data. If the abnormal behavior lasts for more than 30 seconds and the severity of the abnormal behavior increases, such as further increase in movement speed and sudden changes in movement trajectory, the warning level is raised to level 2 according to the preset strategy, and relevant personnel are notified by SMS, email, etc., and the abnormal situation is continuously monitored until the abnormal behavior is eliminated. When the abnormal situation is eliminated, the parameters of the monitoring equipment are dynamically adjusted according to the time and stability of the target's restoration of normal behavior, such as reducing the video frame rate to 30 frames / second and the resolution to 1080P, in order to reduce data processing and storage pressure. The target's behavior data and abnormal situations are recorded in chronological order, and they are annotated and preprocessed as training samples for the abnormal behavior recognition model and early warning strategy. By regularly fine-tuning and optimizing the model, the intelligence level and response efficiency of the system are continuously improved, so that it can detect and deal with abnormal situations more accurately and timely, and improve the practicality and reliability of the monitoring system.

[0030] According to one embodiment of the present invention, after obtaining the community data collected by the monitoring equipment, the data is first preprocessed. Through data cleaning, noise data and outliers are removed, such as using the median filter method to remove data points that exceed 20% of the normal range. Then feature extraction is performed, and key features for behavior pattern recognition, such as target trajectory, movement speed, personnel concentration, etc., are selected. The original data is compressed into a 20-dimensional feature vector using the principal component analysis method to obtain a normalized data set. According to the nonlinear characteristics of the data set, the support vector machine algorithm is selected to establish a community behavior pattern mining model, using the radial basis kernel function, setting the penalty coefficient C=10, and the accuracy of the model on the test set after training 3000 times reaches 95%. The trained model is applied to real-time monitoring data, and behavior pattern recognition is performed by the sliding time window method, with a window size of 10 minutes and a step size of 1 minute. When abnormal behaviors such as mob fights are identified in the community, the system automatically triggers an early warning, notifies community managers and nearby police officers through text messages and voice calls, and pushes the location of the abnormal behavior and the characteristic information of the people involved to their mobile terminals. At the same time, the system dynamically adjusts monitoring strategies based on the frequency and regional distribution of abnormal events, such as increasing the monitoring frequency by 20% in key areas, optimizing the location of cameras, and improving the efficiency of identifying and responding to abnormal behaviors. By continuously monitoring the operation of the system, collecting feedback from managers and police officers, using incremental learning methods to continuously optimize machine learning algorithms, and combining deep learning technology to extract higher-level behavioral features, the system's intelligence level can be further improved to achieve precise management of community safety.

[0031] In addition, in order to achieve the above-mentioned purpose, the present invention also provides a smart community target recognition and tracking system based on big data. Figure 2 is a flow chart of a smart community target identification and tracking system based on big data according to an exemplary embodiment. Figure 2 As shown, a smart community target recognition and tracking system based on big data in the present invention includes: Data monitoring and acquisition module: deploy cameras and sensors in the community, build a community monitoring network, and collect video and sensor data in the community in real time; Feature data acquisition module: Based on video and sensor data, preliminary target detection and feature extraction are performed on the front-end device through edge computing methods to obtain the target's feature data; Identity identification module: transmits the target's feature data to the cloud server through the network, uses the deep learning model to identify and classify the target's feature data, and determines the target's type and identity; Community early warning module: Analyze the current behavior of the target according to the preset target historical behavior pattern library to determine whether its behavior is abnormal. If the target's behavior does not match the normal pattern in the target historical behavior pattern library, it is judged as abnormal behavior and triggers the early warning mechanism; Abnormal monitoring module: After adjusting the parameters of the community monitoring network, the target's video and sensor data are re-acquired, and the target is continuously tracked and analyzed for behavior. If the target's abnormal behavior persists or escalates, the warning level is further increased according to the preset strategy until the abnormal situation is eliminated; Maintenance and update module: Regularly summarize and analyze the data collected by the community monitoring network, use machine learning algorithms to mine behavioral patterns and abnormal laws, and optimize and adjust the community's monitoring strategies and early warning mechanisms based on the analysis results.

[0032] In order to achieve the above-mentioned purpose of the invention, the present invention also provides an electronic device, which includes: a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, the above-mentioned smart community target identification and tracking method based on big data is implemented.

[0033] In order to achieve the above-mentioned purpose of the invention, the present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the above-mentioned smart community target identification and tracking method based on big data is implemented.

[0034] Based on this, the beneficial effect of the present invention lies in that by deploying high-definition cameras and sensors in the community, building a full-range monitoring network, collecting real-time data, and using edge computing technology to perform preliminary target detection and feature extraction on the front-end device, the feature data is transmitted to the cloud server for deep learning recognition and classification, and according to the preset historical behavior pattern library, it is analyzed whether the current behavior of the target is abnormal. If it does not match, an early warning is triggered, the monitoring parameters are dynamically adjusted, the target behavior is continuously tracked and analyzed, and the early warning level is upgraded according to the situation. By regularly summarizing and analyzing data, mining behavior patterns and abnormal laws, optimizing monitoring strategies and early warning mechanisms, the intelligent identification, dynamic tracking and timely early warning of abnormal behaviors in smart communities are realized, and the intelligence level and response efficiency of community safety management are improved.

[0035] Those of ordinary skill in the art will appreciate that the modules and algorithm steps described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0036] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described devices and equipment can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0037] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation, such as multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0038] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, that is, they may be located in one place or distributed on multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the embodiments of the present invention.

[0039] In addition, each functional module in the embodiment of the present invention may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0040] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the energy-saving signal sending / receiving method of each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a ROM, a RAM, a magnetic disk, or an optical disk.

[0041] The above description is only a preferred embodiment of the present application and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the present application is not limited to the technical solution formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the inventive concept. For example, the above features are replaced with (but not limited to) technical features with similar functions disclosed in the present application.

[0042] It should be understood that the size of the serial numbers of each step in the content of the invention and the embodiments of the present invention does not absolutely mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

Claims

1. A smart community target identification and tracking method based on big data, characterized in that: include: Deploy cameras and sensors in the community, build a community monitoring network, and collect video and sensor data in the community in real time; Based on the video and sensor data, preliminary target detection and feature extraction are performed on the front-end device through edge computing methods to obtain the target's feature data; The target’s feature data is transmitted to the cloud server through the network, and the target’s feature data is identified and classified using a deep learning model to determine the target’s type and identity; According to the preset target historical behavior pattern library, the current behavior of the target is analyzed to determine whether its behavior is abnormal. If the target's behavior does not match the normal pattern in the target historical behavior pattern library, it is judged as abnormal behavior and the early warning mechanism is triggered; After adjusting the parameters of the community monitoring network, the target’s video and sensor data are re-acquired, and the target is continuously tracked and its behavior analyzed. If the target’s abnormal behavior persists or escalates, the warning level is further increased according to the preset strategy until the abnormal situation is eliminated; Regularly summarize and analyze the data collected by the community monitoring network, use machine learning algorithms to explore behavioral patterns and abnormal laws, and optimize and adjust the community's monitoring strategies and early warning mechanisms based on the analysis results.

2. A method for identifying and tracking smart community targets based on big data as claimed in claim 1, characterized in that: According to the community layout plan, cameras and sensors are installed at key locations to build a community monitoring network covering the entire community; The collected video streams and sensor data are transmitted in real time via the optical fiber network to the server cluster of the community data center for storage; Use the visual algorithm to analyze the video stream in real time, identify target objects such as people and vehicles, and extract their motion trajectory information; Use time series analysis methods to detect abnormal fluctuations in environmental data such as temperature, humidity, and noise collected by sensors; Use learning models to identify abnormal behaviors in videos and trigger security warnings if suspicious situations are detected; Based on video analysis and sensor data, draw a heat map of community traffic to determine the density of people and provide a basis for decision-making in community management; Through the data visualization platform, various monitoring data and analysis results are displayed centrally, and the community security situation is monitored in real time.

3. A method for identifying and tracking smart community targets based on big data as claimed in claim 2, characterized in that: Acquire real-time video and sensor data and transmit the data to front-end devices; On the front-end device, edge computing technology is used to process data in real time, and the target is detected in the video and sensor data through the target detection algorithm; For the detected target, a feature extraction algorithm is used to extract the features of the target, and the feature data of the target is obtained according to the extracted target features; Compare the target feature data with the preset feature template to determine the target category, and determine the processing strategy based on the target category and feature data.

4. A method for identifying and tracking smart community targets based on big data as claimed in claim 3, characterized in that: An image acquisition device is used to acquire a target image, and the target image is preprocessed to obtain feature data; Extract key feature points from the preprocessed image and generate feature vectors; According to the preset data transmission protocol, the feature vector data is packaged and encrypted and then transmitted to the cloud server through the network; Decrypt the received feature vector data on the cloud server and convert the data format; Perform forward propagation calculation on the converted feature vector to obtain the target category probability distribution; The probability distribution result is compared with the preset recognition threshold. If the maximum probability value exceeds the threshold, it is determined to be that category, otherwise it is determined to be an unknown category; For the identified target category, the corresponding target identity database is called for feature matching to obtain the target's specific identity information.

5. A method for identifying and tracking smart community targets based on big data as claimed in claim 4, characterized in that: Use sensors to monitor target behavior in real time and obtain target behavior data; The collected original target behavior data is cleaned and standardized through data preprocessing to obtain standardized target behavior feature data; Based on the standardized target behavior feature data, clustering algorithms are used to perform cluster analysis on historical behavior data to establish a target historical behavior pattern library; For the current target behavior feature data, a matching algorithm is used to calculate the similarity with the patterns in the target historical behavior pattern library to obtain a matching score; If the matching score is lower than the preset threshold, it is judged as abnormal behavior and triggers abnormal behavior warning; Classify abnormal behaviors and determine warning levels based on their specific characteristics and the degree of deviation from normal patterns; For different warning levels, the corresponding warning mechanism and handling process are triggered, and targeted prevention and control measures are taken.

6. A method for identifying and tracking smart community targets based on big data as claimed in claim 5, characterized in that: Based on the video and sensor data collected by the monitoring equipment, identify and locate the target in the video, and obtain information such as the target's position and movement trajectory; By continuously tracking the target, the target's behavior characteristics at different time points are obtained, and compared with the pre-established normal behavior model to determine whether the target's behavior is abnormal; If abnormal behavior of the target is detected, the parameters of the monitoring equipment are dynamically adjusted according to the severity and duration of the abnormal behavior; The acquired video and sensor data are input into the recognition model, and the deep learning algorithm is used to determine whether the target's abnormal behavior persists or escalates; If the target's abnormal behavior persists or escalates, the warning level will be raised according to the preset strategy until the abnormal situation is eliminated; After the abnormal situation is eliminated, the parameters of the monitoring equipment are dynamically adjusted according to the recovery of the target; Record the target's behavioral data and abnormal situations as training samples for optimizing abnormal behavior recognition models and early warning strategies to optimize the models.

7. A method for identifying and tracking smart community targets based on big data as claimed in claim 6, characterized in that: Obtain community data collected by monitoring equipment and pre-process the data, including data cleaning and feature extraction, to obtain a standardized data set; According to the characteristics of the data set, select the machine learning algorithm and establish a mining model for community behavior patterns and abnormal laws; Use the established mining model to train and test the data set; Apply the trained model to the real-time collected monitoring data to predict and identify the behavior patterns and abnormal situations in the community in real time; Dynamically adjust the monitoring strategy of the smart community based on the results predicted by the model; When abnormal behaviors or events are identified in the community, the early warning mechanism is triggered in time to notify relevant personnel to handle the situation; Continuously monitor and evaluate the system's operation and continuously improve the monitoring strategy based on feedback.

8. A smart community target identification and tracking system based on big data, characterized in that: include: Data monitoring and acquisition module: deploy cameras and sensors in the community, build a community monitoring network, and collect video and sensor data in the community in real time; Feature data acquisition module: Based on video and sensor data, preliminary target detection and feature extraction are performed on the front-end device through edge computing methods to obtain the target's feature data; Identity identification module: The target’s feature data is transmitted to the cloud server through the network, and the target’s feature data is identified and classified using a deep learning model to determine the target’s type and identity. Community early warning module: Analyze the current behavior of the target according to the preset target historical behavior pattern library to determine whether its behavior is abnormal. If the target's behavior does not match the normal pattern in the target historical behavior pattern library, it is judged as abnormal behavior and triggers the early warning mechanism; Abnormal monitoring module: After adjusting the parameters of the community monitoring network, the target's video and sensor data are re-acquired, and the target is continuously tracked and analyzed for behavior. If the target's abnormal behavior persists or escalates, the warning level is further increased according to the preset strategy until the abnormal situation is eliminated; Maintenance and update module: Regularly summarize and analyze the data collected by the community monitoring network, use machine learning algorithms to mine behavioral patterns and abnormal laws, and optimize and adjust the community's monitoring strategies and early warning mechanisms based on the analysis results.

9. An electronic device, characterized in that: It includes a processor, a memory, and a computer program stored in the memory and executable on the processor. When the computer program is executed by the processor, a smart community target identification and tracking method based on big data is implemented as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, it implements a smart community target identification and tracking method based on big data as described in any one of claims 1 to 7.

Images