Crowd detection system using CCTV

Abstract

The present invention is a density detection system using CCTV, comprising: a shooting unit that photographs a set area with a camera; an object recognition unit that recognizes stationary or moving objects in an image captured through the shooting unit and identifies the objects using an object detection algorithm based on deep learning or machine learning; a density analysis unit that determines density using information obtained through the object recognition unit, collects object location data to generate a heatmap, and visually and numerically analyzes spatial density based on the distribution and number of objects within a set area; a storage unit that stores object recognition, density analysis data, and original images obtained through the object recognition unit, and, if necessary, performs statistical analysis or situation reconstruction based on past data; and a control unit that is respectively connected to the shooting unit, the object recognition unit, the density analysis unit, and the storage unit to control the operation of the shooting unit, the object recognition unit, the density analysis unit, and the storage unit.

Description

Technology Field

[0001] The present invention relates to a density detection system using CCTV installed in public facilities or public transportation facilities to monitor people and detect density, thereby preventing safety accidents in advance. Background Technology

[0002] Recently, it has become common practice to establish video surveillance systems utilizing CCTV for crime prevention or securing evidence after an incident. This involves installing multiple CCTV cameras in various regions, displaying the footage generated by these cameras on monitors, and storing it on storage devices. When surveillance personnel discover a crime or accident, they can respond immediately and appropriately; furthermore, if necessary, they can search the CCTV footage stored in the storage to secure evidence after the fact.

[0003] However, the reality is that the number of control center personnel is severely insufficient compared to the installation status of CCTV cameras. To effectively perform video surveillance with such limited personnel, simply displaying CCTV footage on a monitor screen is insufficient.

[0004] It is desirable to enable control center personnel to effectively detect objects by detecting object movements in CCTV footage and displaying additional information in the corresponding area in real time. In such cases, rather than monitoring the entire CCTV footage with uniform interest, control center personnel only need to focus their surveillance on areas containing object movements or meaningful additional information.

[0005] Furthermore, object tracking is being treated as an important function in CCTV video surveillance. Object tracking is a feature that automatically detects objects (e.g., people, cars) in CCTV footage and tracks the detected objects through video analysis.

[0006] To achieve this, video analysis devices (e.g., video analysis servers, video analysis computers) must manage object-unit movements by identifying object identities through reference to the position, size, features, etc., of objects within a continuous sequence of CCTV frames.

[0007] In other words, to perform object tracking on CCTV footage, it is necessary to detect objects, define target objects, verify object identity across frames, and manage tracking through CCTV video analysis. At this stage, verifying object identity involves predicting the position and size of the target object based on frame progression, extracting and comparing object features, detecting movement of feature points (e.g., meanshift, optical flow, camshift), and comparing identity. Prior art literature

[0008] Republic of Korea Registered Patent No. 10-2416620 (July 5, 2022) United States Patent Application Publication US2016 / 0269689 (September 15, 2016) The problem to be solved

[0009] Accordingly, the present invention has been devised to solve the above-mentioned problems, and the objective of the present invention is to provide a density detection system using CCTV that includes an algorithm for detecting the density of a specific location and tracking the direction of movement of people in the said specific location over time.

[0010] However, the technical problems to be solved by the present invention are not limited to those mentioned above, and other technical problems not mentioned will be clearly understood by those skilled in the art to which the present invention belongs from the description below. means of solving the problem

[0011] The present invention was created to improve upon the problems of the prior art described above and is a density detection system using CCTV, comprising: a shooting unit that photographs an area set by a camera; an object recognition unit that recognizes stationary or moving objects in the image captured by the shooting unit and identifies the objects based on deep learning or machine learning; a density analysis unit that determines density using information obtained through the object recognition unit, collects object location data to generate a heatmap, and visually and numerically analyzes spatial density based on the distribution and number of objects within the set area; a storage unit that stores object recognition, density analysis data, and original images obtained through the object recognition unit, and, if necessary, performs statistical analysis or situation reconstruction based on past data; and a control unit that is respectively connected to the shooting unit, the object recognition unit, the density analysis unit, and the storage unit to control the operation of the shooting unit, the object recognition unit, the density analysis unit, and the storage unit.

[0012] In addition, in one embodiment, the object recognition unit may include: an image preprocessing unit that converts an original image input from the shooting unit into a form suitable for a deep learning or machine learning-based object detection model; a feature extraction unit that extracts meaningful set visual features from the image preprocessed through the image preprocessing unit; an object candidate region generation unit that sets a location where an object may exist as a candidate region based on a feature map extracted through the feature extraction unit; and an object classification and bounding box adjustment unit that identifies what each object is in the candidate region through the object candidate region generation unit and accurately adjusts its location.

[0013] Additionally, in one embodiment, the object recognition unit may further include an ID assignment and inter-frame connection unit that, after an object is identified through the object classification and bounding box adjustment unit, assigns a unique ID to each object and connects them to track the same object in consecutive frames.

[0014] Additionally, in one embodiment, the density analysis unit may include: a spatial division and weight setting unit that divides the shooting area into multiple grid cells or spatial blocks serving as units of analysis and assigns weights to each grid cell; an object density calculation unit that calculates a density index by reflecting the number, size, and weight of objects detected within each cell after spatial division and weighting are completed through the spatial division and weight setting unit; a density threshold classification and risk evaluation unit that classifies the density status of the entire shooting area based on the density value per cell calculated through the object density calculation unit and evaluates the risk by considering the size, continuity, and change trends of the dense area; a density change analysis unit that analyzes density changes over time through the density threshold classification and risk evaluation unit; and a density judgment result integration and multidimensional interpretation unit that integrates the density value per individual cell, risk classification, and time-series change analysis results calculated through the density change analysis unit to interpret the density situation of the entire space in a multidimensional manner.

[0015] In addition, in one embodiment, it may further include a movement path analysis unit that is controlled by the control unit and uses information obtained through the object recognition unit to record the location data of an object in chronological order and clusters the movement trajectory data to derive a pattern.

[0016] In addition, in one embodiment, it may further include an integration and visualization unit that is controlled by the control unit and integrates the density value, number of objects, and movement path analyzed by the movement path analysis unit to provide them to the user in an intuitive form.

[0017] Additionally, in one embodiment, the movement path analysis unit may include: a movement path data preprocessing and matching unit that detects, removes, or corrects outliers in location data and applies interpolation to connect trajectories when continuity is broken; a movement path clustering and clustering unit that performs clustering operations by grouping objects with similar movement paths based on movement path data preprocessed through the movement path data preprocessing and matching unit; a movement pattern characteristic analysis and classification unit that analyzes and classifies movement path clusters classified by clustering after passing through the movement path clustering and clustering unit in detail according to their respective characteristics; and an abnormal movement behavior and abnormal pattern detection unit that detects behaviors deviating from a standard set through the movement pattern characteristic analysis and classification unit in real time.

[0018] Additionally, in one embodiment, the integration and visualization unit may include: a data integration and matching unit that consistently combines various forms of data provided from the object recognition unit, the density analysis unit, and the movement path analysis unit; a visualization interface design and implementation unit that visualizes based on the data integrated through the data integration and matching unit; a real-time data feedback and notification linkage unit that rapidly transmits real-time changing density and movement pattern information to the user through the visualization interface design and implementation unit; and a report generation unit that generates various customized reports based on the analysis data through the real-time data feedback and notification linkage unit. Effects of the invention

[0019] According to one embodiment of the present invention, by monitoring people through CCTVs installed in public facilities or public transportation facilities and detecting density using an analysis program or algorithm, the density of a specific location is detected, and the location where people move over time is tracked, thereby having the effect of preventing safety accidents in advance.

[0020] In addition, according to one embodiment of the present invention, an artificial intelligence chip is provided in the control unit, and based on deep learning, the density of the area captured by the shooting unit and recognized through the object recognition unit is specifically analyzed and visualized to show, thereby having the effect of preventing or notifying of safety accidents in advance.

[0021] However, the effects obtainable from the present invention are not limited to those mentioned above, and other unmentioned effects will be clearly understood by those skilled in the art from the description below. Brief explanation of the drawing

[0022] The following drawings attached to this specification illustrate preferred embodiments of the present invention and serve to further enhance understanding of the technical concept of the present invention together with the detailed description of the invention provided below; therefore, the present invention should not be interpreted as being limited only to the matters described in such drawings. FIG. 1 is a conceptual diagram of a density detection system using a CCTV according to an embodiment of the present invention. Figure 2 is a block diagram of the overall components of a density detection system using CCTV. Figure 3 is a block diagram of the components of the object recognition unit. Figure 4 is a block diagram of the components of the density analysis unit. Figure 5 is a block diagram of the components of the space partitioning and weight setting unit. Figure 6 is a block diagram of the components of the object density calculation unit. Figure 7 is a block diagram of the components of the density threshold classification and risk assessment section. Figure 8 is a block diagram of the components of the density change analysis unit. Figure 9 is a block diagram of the components of the density judgment result integration and multidimensional analysis unit. Figure 10 is a block diagram of the components of the movement path analysis unit. Figure 11 is a block diagram of the components of the integration and visualization section. Specific details for implementing the invention

[0023] Below, with reference to the attached drawings, embodiments of the present invention are described in detail so that those skilled in the art can easily implement the invention. However, since the description of the present invention is merely an example for structural or functional explanation, the scope of the present invention should not be interpreted as being limited by the embodiments described in the text. That is, since the embodiments are subject to various modifications and may take various forms, the scope of the present invention should be understood to include equivalents capable of realizing the technical concept. Furthermore, the objectives or effects presented in the present invention do not imply that a specific embodiment must include all of them or only such effects; therefore, the scope of the present invention should not be understood as being limited by them.

[0024] The meaning of the terms described in this invention should be understood as follows.

[0025] Terms such as "first" and "second" are intended to distinguish one component from another, and the scope of rights shall not be limited by these terms. For example, the first component may be named the second component, and similarly, the second component may be named the first component. When a component is referred to as being "connected" to another component, it should be understood that it may be directly connected to that other component, or that there may be other components in between. Conversely, when a component is referred to as being "directly connected" to another component, it should be understood that there are no other components in between. Meanwhile, other expressions describing the relationship between components, such as "between" and "exactly between," or "adjacent to" and "directly adjacent to," shall be interpreted in the same manner.

[0026] A singular expression should be understood to include a plural expression unless the context clearly indicates otherwise, and terms such as "include" or "have" are intended to specify the existence of the set-up features, numbers, steps, actions, components, parts, or combinations thereof, and should be understood not to preclude the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof.

[0027] Unless otherwise defined, all terms used herein have the same meaning as generally understood by those skilled in the art to which this invention pertains. Terms defined in commonly used dictionaries should be interpreted as having meanings consistent with the context of the relevant technology and should not be interpreted as having an ideal or overly formal meaning unless explicitly defined in this invention.

[0028] FIG. 1 is a conceptual diagram of a density detection system using CCTV according to an embodiment of the present invention, FIG. 2 is a block diagram of the overall components of a density detection system using CCTV, FIG. 3 is a block diagram of the components of an object recognition unit, FIG. 4 is a block diagram of the components of a density analysis unit, FIG. 5 is a block diagram of the components of a spatial division and weight setting unit, FIG. 6 is a block diagram of the components of an object density calculation unit, FIG. 7 is a block diagram of the components of a density threshold classification and risk evaluation unit, FIG. 8 is a block diagram of the components of a density change analysis unit, FIG. 9 is a block diagram of the components of a density judgment result integration and multidimensional interpretation unit, FIG. 10 is a block diagram of the components of a movement path analysis unit, and FIG. 11 is a block diagram of the components of an integration and visualization unit.

[0029] As illustrated in FIGS. 1 to 11, the present invention may include a density detection system using CCTV, a shooting unit (100), an object recognition unit (200), a density analysis unit (300), a storage unit (400), and a control unit (500).

[0030] The shooting unit (100) can capture an area set by a camera. Specifically, the shooting unit (100) can collect accurate data even in a low-light environment using at least one of a high-resolution camera, a PTZ camera that provides clear images for object identification, and an IR camera equipped with detailed tracking and zoom functions for a specific area.

[0031] The object recognition unit (200) recognizes stationary or moving objects in an image captured through the shooting unit (100) and can identify objects using an object detection algorithm (50) based on deep learning or machine learning.

[0032] The object recognition unit (200) may include an image preprocessing unit (210), a feature extraction unit (220), an object candidate region generation unit (230), an object classification and bounding box adjustment unit (240), and an ID assignment and inter-frame connection unit (250).

[0033] The image preprocessing unit (210) can convert the original image input from the shooting unit (100) into a form suitable for a deep learning or machine learning-based object detection model.

[0034] The feature extraction unit (220) can extract meaningful set visual features from the image preprocessed through the image preprocessing unit (210).

[0035] The object candidate region generation unit (230) can set a location where an object may exist as a candidate region based on the feature map extracted through the feature extraction unit (220).

[0036] The object classification and bounding box adjustment unit (240) can identify what each object is for the subsequent candidate area through the object candidate area generation unit (230) and accurately adjust its position.

[0037] The ID assignment and inter-frame connection unit (250) can assign a unique ID to each object and connect them so that the same object can be tracked in consecutive frames after the object is identified through the object classification and bounding box adjustment unit (240).

[0038] The density analysis unit (300) determines the density using information obtained through the object recognition unit (200), collects object location data to generate a heatmap, and can visually and numerically analyze the spatial density based on the distribution and number of objects within a set area.

[0039] The density analysis unit (300) may include a spatial partitioning and weight setting unit (310), an object density calculation unit (320), a density threshold classification and risk evaluation unit (330), a density change analysis unit (340), and a density judgment result integration and multidimensional interpretation unit (350).

[0040] The space division and weight setting unit (310) can divide the shooting area into several grid cells or space blocks that serve as units of analysis and assign weights to each grid cell.

[0041] Specifically, the space partitioning and weight setting unit (310) may include a basic space partitioning unit (311), a weight application criteria establishment unit (312), an object-based weight mapping unit (313), and a cell weight integration and normalization unit (314).

[0042] The basic space division unit (311) can divide the entire observation area into grid cells or space blocks of a set size based on the frame of the captured image of the capturing unit (100).

[0043] The weight application criteria establishment unit (312) can define evaluation criteria for setting weights to be assigned to cells through the basic space division unit (311).

[0044] The object-based weight mapping unit (313) can analyze the characteristics of each object detected by the object recognition unit (200) according to the evaluation criteria set through the weight application criteria establishment unit (312), and assign a corresponding weight to reflect it in the cell to which the object belongs.

[0045] The cell weight integration and normalization unit (314) can quantify and make comparable the density distribution across the entire space based on the weights accumulated in each cell through the object-based weight mapping unit (313).

[0046] The object density calculation unit (320) can calculate a density index by reflecting the number, size, and weight of objects detected within each cell after the spatial division and weight setting unit (310) completes the spatial division and weighting.

[0047] Specifically, the object density calculation unit (320) may include an object quantity identification unit (321), an object characteristic-based correction unit (322), a cell density index calculation unit (323), and a cell-to-cell relative comparison and outlier processing unit (324).

[0048] The object quantity determination unit (321) can identify the cell to which the object belongs and calculate the cumulative number of objects per cell based on the center coordinates of each object obtained through the object recognition unit (200).

[0049] The object characteristic-based correction unit (322) can correct the density within the cell by reflecting at least one of the object's size, movement type, dwell time, and directionality through the object quantity identification unit (321).

[0050] The cell density index calculation unit (323) can calculate the final density value of each cell based on the number of objects collected through the object characteristic-based correction unit (322) and the correction weight.

[0051] The cell-to-cell relative comparison and outlier processing unit (324) can set a reference value or comparison cell group for each space type through the cell density index calculation unit (323) and convert it into a relative value within the group.

[0052] The density threshold classification and risk evaluation unit (330) can classify the density status of the entire shooting area based on the density value per cell calculated through the object density calculation unit (320) and evaluate the risk by considering the size, continuity, and change trend of the dense area.

[0053] Specifically, the density threshold classification and risk evaluation unit (330) may include a density standard threshold setting unit (331), a spatial continuity and concentration evaluation unit (332), a temporal change rate and surge detection unit (333), and a comprehensive risk evaluation and grading unit (334).

[0054] The density standard threshold setting unit (331) can set the boundary of the risk level corresponding to the density value of each cell based on a preset standard.

[0055] The spatial continuity and concentration evaluation unit (332) can identify clustered dense areas by analyzing the relationship with adjacent cells through the density standard threshold setting unit (331).

[0056] The temporal change rate and surge detection unit (333) can compare the difference between the predicted value and the actual value in real time using a time series model or an LSTM-based deep learning (or machine learning) model after passing through the spatial continuity and concentration evaluation unit (332), and flag it as an abnormal pattern when it exceeds the error range.

[0057] The risk comprehensive evaluation and grading unit (334) can calculate the final risk level for each cell and the entire space by synthesizing the spatial and temporal analysis results collected through the temporal change rate and surge detection unit (333), and visualize this as a grading system.

[0058] The density change analysis unit (340) can analyze density changes over time through the density threshold classification and risk evaluation unit (330).

[0059] Specifically, the density change analysis unit (340) may include a time series density history collection and normalization unit (341), a time series change pattern analysis unit (342), an anomaly detection and boundary situation determination unit (343), and a spatiotemporal correlation analysis unit (344) of density change.

[0060] The time series density history collection and normalization unit (341) can continuously collect density values ​​according to time intervals set for each grid cell and standardize them to structure time series data aligned along the time axis.

[0061] The time series change pattern analysis unit (342) can analyze the trend of density change over time in each cell based on the time series density data collected through the time series density history collection and normalization unit (341).

[0062] The abnormal change detection and boundary situation determination unit (343) can detect a section in the time series change that has an abnormally rapid change exceeding a set standard through the time series change pattern analysis unit (342) and determine it as a boundary state or a dangerous situation.

[0063] The spatiotemporal correlation analysis unit (344) of density change can analyze density change data over time in conjunction with the distribution pattern of the entire space through the abnormal change detection and boundary situation judgment unit (343).

[0064] The density judgment result integration and multidimensional interpretation unit (350) can integrate the density values ​​per individual cell, risk classification, and time series change analysis results calculated through the density change analysis unit (340) to interpret the density situation of the entire space in a multidimensional way.

[0065] Specifically, the density judgment result integration and multidimensional interpretation unit (350) may include a cell-based density data integration unit (351), a spatiotemporal trend context analysis unit (352), an abnormal sign integration evaluation and complex risk calculation unit (353), and a multidimensional visualization and interpretation result provision unit (354).

[0066] The cell-based density data integration unit (351) can reconstruct at least one of the density values, risk grades, and time series change information of each grid cell obtained from the density change analysis unit (340) into an integrated matrix or data frame.

[0067] The spatiotemporal trend context analysis unit (352) can perform analysis to identify the trend of increase or decrease in density over time and the congestion pattern that occurs repeatedly in the set time period in the data integrated through the cell-based density data integration unit (351).

[0068] The abnormal sign integrated evaluation and composite risk calculation unit (353) multidimensionally evaluates the risk of a dense state based on statistical information by cluster and spatiotemporal change patterns through the spatiotemporal trend context analysis unit (352), and can calculate a comprehensive score by reflecting at least one of the density change rate, duration, spatial diffusion speed and the rapid change of movement patterns.

[0069] The multidimensional visualization and interpretation result providing unit (354) can visually represent and deliver all integrated dense information and analysis results through the abnormal sign integrated evaluation and complex risk calculation unit (353).

[0070] A CCTV system according to one embodiment of the present invention may further include a movement path analysis unit (600) and an integration and visualization unit (700).

[0071] The movement path analysis unit (600) is controlled by the control unit (500) and uses information obtained through the object recognition unit (200) to record the location data of the object in chronological order and cluster the movement trajectory data to derive a pattern.

[0072] The movement path analysis unit (600) may include a movement trajectory data preprocessing and matching unit (610), a movement trajectory clustering and clustering unit (620), a movement pattern characteristic analysis and classification unit (630), and an abnormal movement behavior and abnormal pattern detection unit (640).

[0073] The movement trajectory data preprocessing and matching unit (610) detects and removes or corrects outliers in the position data, and if the continuity is broken, it can apply interpolation to connect the trajectories.

[0074] The movement trajectory clustering and clustering unit (620) can perform clustering operations by grouping objects with similar movement paths based on movement trajectory data preprocessed through the movement trajectory data preprocessing and matching unit (322).

[0075] The movement pattern characteristic analysis and classification unit (630) can analyze and classify the movement path clusters classified by clustering after passing through the movement trajectory clustering and clustering unit (324) in detail according to each characteristic.

[0076] The abnormal movement behavior and abnormal pattern detection unit (640) can detect behaviors that deviate in real time by comparing them with a normal pattern model of the standard set through the movement pattern characteristic analysis and classification unit (326).

[0077] The integration and visualization unit (700) is controlled by the control unit (500) and can integrate the density value, number of objects, and movement path analyzed by the movement path analysis unit (600) and provide them to the user in an intuitive form.

[0078] The integration and visualization unit (700) may include a data integration and matching unit (710), a visualization interface design and implementation unit (720), a real-time data feedback and notification linkage unit (730), and a report generation unit (740).

[0079] The data integration and matching unit (710) can consistently combine various types of data provided from the object recognition unit (200), the density analysis unit (300), and the movement path analysis unit (600).

[0080] The visualization interface design and implementation unit (720) can visualize based on the data integrated through the data integration and matching unit (710).

[0081] The real-time data feedback and notification linkage unit (730) can quickly deliver information on density and movement patterns that change in real time to the user through the visualization interface design and implementation unit (720).

[0082] The report generation unit (740) can generate various customized reports based on analysis data through the real-time data feedback and notification linkage unit (730).

[0083] The storage unit (400) stores object recognition, density analysis data, and original images obtained through the object recognition unit (200), and, if necessary, allows for statistical analysis or situation reconstruction based on past data. The storage unit (400) is configured to store at least one of the object's unique identifier and location information as an object identification result output from the object recognition unit (200), the cell-by-cell density index and risk classification result regarding the density state output from the density analysis unit (300), and the original images obtained from the shooting unit (100).

[0084] The control unit (500) is connected to the shooting unit (100), object recognition unit (200), density analysis unit (300), and storage unit (400), respectively, and can control the operation of the shooting unit (100), object recognition unit (200), density analysis unit (300), and storage unit (400).

[0085] Specifically, the control unit (500) is equipped with a wireless communication module (not shown) so that it can be linked with an external mobile terminal, etc., to control operation or check the control status in real time from the outside. The wireless communication module can be implemented with various communication technologies. That is, Wi-Fi, WCDMA (Wideband CDMA), HSDPA (High Speed ​​Downlink Packet Access), HSUPA (High Speed ​​Uplink Packet Access), HSPA (High Speed ​​Packet Access), Mobile WiMAX, WiBro, LTE (Long Term Evolution), 5G, Bluetooth, infrared communication (IrDA, infrared data association), NFC (Near Field Communication), Zigbee, wireless LAN technology, etc., can be applied. In addition, when providing services connected to the Internet, it may follow TCP / IP, which is a standard protocol for information transmission on the Internet.

[0086] As described above, the detailed description of the preferred embodiments of the present invention disclosed is provided to enable those skilled in the art to implement and practice the present invention. Although the present invention has been described with reference to preferred embodiments, those skilled in the art will understand that various modifications and changes can be made to the present invention without departing from the scope of the invention. For example, those skilled in the art may utilize each configuration described in the embodiments described above in combination with one another. Accordingly, the present invention is not intended to be limited to the embodiments shown herein, but to be given the broadest scope consistent with the principles and novel features disclosed herein.

[0087] The present invention may be embodied in other specific forms without departing from the spirit and essential features of the invention. Accordingly, the above detailed description should not be interpreted restrictively in all respects but should be considered exemplary. The scope of the invention shall be determined by a reasonable interpretation of the appended claims, and all modifications within the equivalent scope of the invention are included within the scope of the invention. The invention is not intended to be limited to the embodiments shown herein, but to be given the broadest possible scope consistent with the principles and novel features disclosed herein. Furthermore, embodiments may be constructed by combining claims that are not explicitly related in the claims, or by including them as new claims through amendments made after filing. Explanation of the symbols

[0088] 10: CCTV System 100 : Filming Department 200 : Object Recognition Unit 210 : Image Preprocessing Unit 220 : Feature extraction unit 230 : Object candidate region generation unit 240 : Object Classification and Bounding Box Adjustment Unit 250: ID assignment and inter-frame connection 300 : Density Analysis Department 310: Spatial partitioning and weight setting section 311 : Basic space partition 312 : Weight Application Criteria Establishment Unit 313: Object-based weight mapping section 314: Cell Weight Consolidation and Normalization 320 : Object density calculation unit 321 : Object Quantity Identification Unit 322 : Object characteristic-based correction unit 323 : Cell Density Indicator Calculation Section 324 : Inter-cell relative comparison and outlier handling unit 330: Density Threshold Classification and Risk Assessment Section 331 : Density standard threshold setting section 332 : Spatial Continuity and Concentration Evaluation Section 333: Temporal Rate of Change and Surge Detection Unit 334 : Comprehensive Risk Assessment and Grading Department 340 : Density Change Analysis Section 341: Time Series Density History Collection and Normalization Section 342 : Time Series Change Pattern Analysis Section 343: Anomaly Detection and Boundary Situation Determination Unit 344 : Analysis of Spatiotemporal Correlation of Density Changes 350 : Density Judgment Result Integration and Multidimensional Analysis Unit 351 : Cell-based Density Data Integration Unit 352 : Spatiotemporal Trend Context Analysis Department 353: Integrated Anomaly Assessment and Complex Risk Calculation Unit 354 : Multidimensional Visualization and Interpretation Result Provision Section 400 : Storage unit 500 : Control unit 600 : Movement Path Analysis Unit 610: Movement Trajectory Data Preprocessing and Matching Section 620: Movement Trajectory Clustering and Clustering Section 630: Movement Pattern Characteristic Analysis and Classification Unit 640: Abnormal movement form and abnormal pattern detection unit 700 : Integration and Visualization Section 710: Data Integration and Matching Unit 720: Visualization Interface Design and Implementation Section 730: Real-time data feedback and notification integration unit 740 : Report Generation Section

Claims

Claim 1 A density detection system using CCTV, comprising: a shooting unit that photographs an area set by a camera; an object recognition unit that recognizes stationary or moving objects in the video captured by the shooting unit and identifies the objects based on deep learning or machine learning; a density analysis unit that uses the location information of the objects detected by the object recognition unit to divide the shooting area into a plurality of grid cells, calculates a density index per cell by reflecting the number of objects detected within each grid cell and weights according to object characteristics, and quantitatively analyzes the spatial density state of the entire shooting area based on the spatial distribution of the density index per cell; and a storage unit that stores the object identification result output from the object recognition unit, the density index per cell and risk classification result output from the density analysis unit, and the original video acquired from the shooting unit. and a control unit connected to the above-mentioned shooting unit, the above-mentioned object recognition unit, the above-mentioned density analysis unit, and the above-mentioned storage unit, respectively, to control the operation of the above-mentioned shooting unit, the above-mentioned object recognition unit, the above-mentioned density analysis unit, and the above-mentioned storage unit; wherein the density analysis unit comprises: a spatial division and weight setting unit that divides the shooting area into multiple grid cells or spatial blocks serving as units of analysis and assigns weights to each grid cell; an object density calculation unit that calculates a density index by reflecting the number, size, and weight of objects detected within each cell after the spatial division and weight setting unit has completed the spatial division and weighting; a density threshold classification and risk evaluation unit that classifies the density status of the entire shooting area based on the cell-specific density values ​​calculated through the object density calculation unit and evaluates the risk by considering the size, continuity, and change trends of the dense area; and a density change analysis unit that analyzes density changes over time through the density threshold classification and risk evaluation unit.and a density judgment result integration and multidimensional interpretation unit that integrates the density values ​​per individual cell, risk classification, and time series change analysis results calculated through the density change analysis unit above to interpret the density situation of the entire space in a multidimensional manner;The spatial partitioning and weight setting unit includes a basic spatial partitioning unit, a weight application criteria establishment unit, an object-based weight mapping unit, and a cell weight integration and normalization unit; the basic spatial partitioning unit divides the entire observation area into grid cells or spatial blocks of a set size based on the frames of the captured images of the capturing unit; the weight application criteria establishment unit defines evaluation criteria for setting weights to be assigned to cells through the basic spatial partitioning unit; the object-based weight mapping unit analyzes the characteristics of each object detected by the object recognition unit according to the evaluation criteria set through the weight application criteria establishment unit, assigns a corresponding weight, and reflects it in the cell to which the object belongs; the cell weight integration and normalization unit quantifies and makes comparable the density distribution across the entire space based on the weights accumulated in each cell through the object-based weight mapping unit; and the object density calculation unit includes an object quantity identification unit, an object characteristic-based correction unit, a cell density index calculation unit, and a relative comparison between cells and outlier processing unit; the object quantity identification unit [uses] the center coordinates of each object obtained through the object recognition unit A density detection system using a CCTV, characterized by identifying the cell to which the object belongs and calculating the cumulative number of objects per cell based on the object, wherein the object characteristic-based correction unit corrects the density within the cell by reflecting at least one of the object's size, movement type, dwell time, and directionality through the object quantity identification unit, wherein the cell density index calculation unit calculates the final density value of each cell based on the number of objects collected through the object characteristic-based correction unit and the correction weight, and wherein the cell-to-cell relative comparison and outlier processing unit sets a reference value or comparison cell group for each space type through the cell density index calculation unit and converts it into a relative value within the group. Claim 2 A density detection system using a CCTV according to claim 1, wherein the object recognition unit comprises: an image preprocessing unit that converts an original image input from the shooting unit into a form suitable for a deep learning or machine learning-based object detection model; a feature extraction unit that extracts meaningful set visual features from an image preprocessed through the image preprocessing unit; an object candidate region generation unit that sets a location where an object is likely to exist as a candidate region based on a feature map extracted through the feature extraction unit; and an object classification and bounding box adjustment unit that identifies what each object is in the candidate region through the object candidate region generation unit and accurately adjusts its location. Claim 3 A density detection system using a CCTV according to claim 2, wherein the object recognition unit further comprises an ID assignment and inter-frame connection unit that assigns a unique ID to each object and connects them to track the same object in consecutive frames after the object is identified through the object classification and bounding box adjustment unit. Claim 4 delete Claim 5 A density detection system using a CCTV according to claim 1, further comprising a movement path analysis unit that is controlled by the control unit and uses information obtained through the object recognition unit to record the location data of an object in chronological order and clusters the movement trajectory data to derive a pattern. Claim 6 A density detection system using a CCTV according to claim 5, further comprising an integration and visualization unit controlled by the control unit and integrating the density value, number of objects, and movement path analyzed by the movement path analysis unit to provide to the user in an intuitive form. Claim 7 A density detection system using a CCTV according to claim 5, wherein the movement path analysis unit comprises: a movement path data preprocessing and matching unit that detects, removes, or corrects outliers in position data and applies interpolation to connect trajectories when continuity is broken; a movement path clustering and clustering unit that performs clustering operations by grouping objects with similar movement paths based on movement path data preprocessed through the movement path data preprocessing and matching unit; a movement pattern characteristic analysis and classification unit that analyzes and classifies movement path clusters classified by clustering after passing through the movement path clustering and clustering unit in detail according to their respective characteristics; and an abnormal movement behavior and abnormal pattern detection unit that detects behaviors deviating from a standard set through the movement pattern characteristic analysis and classification unit in real time. Claim 8 A density detection system using a CCTV according to claim 6, wherein the integration and visualization unit comprises: a data integration and matching unit that consistently combines various forms of data provided from the object recognition unit, the density analysis unit, and the movement path analysis unit; a visualization interface design and implementation unit that visualizes based on the data integrated through the data integration and matching unit; a real-time data feedback and notification linkage unit that rapidly transmits density and movement pattern information changing in real time to a user through the visualization interface design and implementation unit; and a report generation unit that generates various customized reports based on the analysis data through the real-time data feedback and notification linkage unit.

Images