An explosion-proof intelligent video monitoring control system

Abstract

The application discloses an explosion-proof intelligent video monitoring control system and relates to the technical field of intelligent security and protection.The system comprises a collection module, a feature extraction module, a linkage factor construction module, an abnormality discrimination module, a path modeling module and a risk early warning module.The collection module is used for collecting multispectral video data and generating partitioned time sequence original data sets.The feature extraction module is used for extracting pixel intensity changes and constructing original fluctuation spectra.The linkage factor construction module is used for constructing a multi-channel disturbance parameter matrix based on the same region multi-channel original fluctuation spectrum and extracting multi-dimensional eddy current spectrum linkage factors.The abnormality discrimination module is used for performing high-dimensional density peak clustering, identifying an abnormal high condensation area and calculating a spectrum convergence factor.The path modeling module is used for performing time sequence tracking on the abnormal high condensation area and establishing a multi-dimensional diffusion path model.The risk early warning module is used for setting multi-level response thresholds and realizing graded risk early warning of abnormal areas.Through multi-channel disturbance modeling, abnormal aggregation area discrimination and diffusion tracking, the sensitive perception of early risks in complex scenes and the dynamic evolution monitoring capability are improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent security and protection, in particular to an explosion-proof intelligent video monitoring control system. BACKGROUND

[0002] With the continuous growth of industrial automation and intelligent security needs, the real-time monitoring and risk warning requirements of explosion-proof key places such as petrochemical industry, natural gas, coal mine, military industry and dangerous goods storage are increasingly stringent. The traditional explosion-proof video monitoring system mainly uses single sensor or visible light video monitoring method, which is difficult to realize dynamic capture of multi-modal, multi-region and rapid change situation in complex environment, and has limited intelligent identification ability for small disturbance, dangerous signs and early abnormal diffusion.

[0003] The existing explosion-proof monitoring technology mainly has the following shortcomings: only relying on single channel or low frame rate video information, lacking stereoscopic identification ability for early risk characteristics of abnormal disturbance, smoke, flame, gas leakage and other multi-source; it is difficult to effectively analyze the synergistic anomaly and spatial energy agglomeration effect of multiple points and multiple channels in the monitoring area, which is easy to cause missed detection, false alarm and delay of early warning of abnormal events; it is unable to combine multi-parameters such as time sequence dynamic characteristics, regional multi-dimensional linkage factors and spatial diffusion path to carry out graded early warning and response linkage of risk, and lacks closed-loop control and intelligent adjustment mechanism.

[0004] Therefore, an explosion-proof intelligent video monitoring control system combining multi-spectral video data, spatio-temporal feature extraction, multi-channel linkage factor analysis and spatial path dynamic modeling is needed, which can realize early identification, spatial evolution tracking and graded risk warning of abnormal high condensation area, so as to significantly improve the intrinsic safety level and emergency response ability of explosion-proof places. SUMMARY

[0005] Based on the above-mentioned shortcomings of the prior art, the purpose of the present application is to provide an explosion-proof intelligent video monitoring control system to solve the above technical problems.

[0006] In order to achieve the above-mentioned purpose, the present application provides the following technical scheme: an explosion-proof intelligent video monitoring control system, comprising:

[0007] The acquisition module is used for real-time acquisition of multi-spectral video data in the explosion-proof monitoring area, multi-channel high-frequency synchronous sampling of the sub-region based on the preset spatial grid division, and generation of partitioned time sequence original data set;

[0008] The feature extraction module is used for extracting the continuous frame pixel intensity change in the time window of the partitioned time sequence original data set, and obtaining the light field disturbance feature and constructing the original wave spectrum through time domain transformation and frequency domain transformation processing;

[0009] Linkage factor construction module: for constructing a multi-channel disturbance parameter joint matrix based on the native fluctuation spectrum of different channels in the same region, obtaining a multi-dimensional eddy current spectrum linkage factor, and the multi-dimensional eddy current spectrum linkage factor includes vortex density index, spectrum aggregation flow index and perturbation field change rate;

[0010] Abnormality discrimination module: for high-dimensional density peak clustering of multi-dimensional eddy current spectrum linkage factors of all spatial partitions, identifying a region with significant vortex disturbance aggregation trend, calculating a spectrum aggregation factor, and identifying and marking an abnormal high condensation area in the space according to the spectrum aggregation factor;

[0011] Path modeling module: for time series tracking of the abnormal high condensation area, extracting spatial diffusion speed, displacement direction and duration, and constructing a multi-dimensional diffusion path model;

[0012] Risk early warning module: for setting multi-level response thresholds based on the multi-dimensional diffusion path model and performing risk early warning.

[0013] The application further provides that the generation logic of the partition time sequence original data set comprises:

[0014] The explosion-proof monitoring area is spatially divided according to preset spatial grid parameters, and the area is discretized into a plurality of sub-areas;

[0015] An explosion-proof intelligent video monitoring device is arranged corresponding to each sub-area, and the explosion-proof intelligent video monitoring device is configured with a multi-spectral video acquisition unit for multi-channel synchronous acquisition control;

[0016] Through a unified synchronous clock and a sampling trigger mechanism, high-frequency data acquisition is performed in the same time window to obtain video frame data containing infrared, visible light and ultraviolet;

[0017] The multi-channel video frames collected at all sampling moments of each sub-area are structured and arranged according to the partition number, time sequence and channel type to generate a partition time sequence original data set.

[0018] The application further provides that the construction logic of the native fluctuation spectrum comprises:

[0019] The partition time sequence original data set is segmented according to a set time window, and continuous multi-frame video images of each partition and each channel are segmented;

[0020] For each continuous video frame in the time window, a pixel intensity sequence in the time window is extracted according to the pixel coordinates to generate a time sequence pixel intensity matrix;

[0021] The time sequence pixel intensity matrix is subjected to first-order differentiation to obtain an inter-frame pixel intensity change sequence;

[0022] Temporal statistical features are extracted from the pixel intensity variation sequence, including sliding mean, variance, kurtosis and skewness;

[0023] Fast Fourier transform is performed on each pixel intensity variation sequence to extract frequency domain statistical features, including main frequency component, energy spectrum distribution, local amplitude peak value and instantaneous phase parameter;

[0024] The temporal statistical features and the frequency domain statistical features of all pixels and all channels in each time window are summarized according to the spatial partition to obtain a partitioned light field disturbance feature group;

[0025] The partitioned light field disturbance feature group is combined with the three indexes of partition, channel and time window to generate a native fluctuation spectrum data group.

[0026] The application further provides that the acquisition logic of the multi-dimensional eddy current spectrum linkage factor includes:

[0027] The native fluctuation spectrum data groups of different channels under the same partition and the same time window are aligned according to the main disturbance parameter to construct a multi-channel disturbance parameter joint matrix;

[0028] Based on the energy spectrum distribution, the spatial aggregation degree of disturbance energy in the partition is calculated to extract a vortex density index;

[0029] According to the main frequency component and the local amplitude peak value, the energy synchronous change rate and the peak energy synthesis intensity of different channels in the main frequency band are counted to extract a spectrum aggregation flow index;

[0030] According to the change gradient of the energy spectrum with time in the energy spectrum distribution, the rate of disturbance energy transfer from low frequency to high frequency is calculated to extract a micro-disturbance field transfer rate;

[0031] The vortex density index, the spectrum aggregation flow index and the micro-disturbance field transfer rate are set as the multi-dimensional eddy current spectrum linkage factor.

[0032] The application further provides that the alignment according to the main disturbance parameter to construct the multi-channel disturbance parameter joint matrix includes:

[0033] The main frequency component, the local amplitude peak value and the instantaneous phase parameter are extracted from the native fluctuation spectrum data group under each channel and are set as the main disturbance parameter;

[0034] The main frequency components of all channels in the same partition and the same time window form a main frequency index set;

[0035] If the main frequency components of different channels exist slight offset, the parameter values of each channel are uniformly mapped to the frequency points of the main frequency index set through linear interpolation or nearest neighbor matching algorithm to ensure that the main disturbance parameters of each channel correspond to each other in the frequency dimension;

[0036] The aligned channel main frequency components, local amplitude peak values and instantaneous phase parameters are arranged according to the main frequency indexes to form a multi-channel disturbance parameter joint matrix, wherein each row of the matrix corresponds to a channel, and each column corresponds to a disturbance parameter at a main frequency point.

[0037] The application further provides that the extraction logic of the vortex density index, the spectral aggregation flow index and the perturbation field transition rate comprises:

[0038] Based on the energy spectrum distribution of each partition in the original fluctuation spectrum of each channel, a high-frequency energy threshold is set, high-frequency disturbance energy points with energy greater than the high-frequency energy threshold are screened, the density of the high-frequency disturbance energy points in the partition space is calculated, and the density is set as the vortex density index.

[0039] The main frequency components and the corresponding local amplitude peak values of each channel in each partition are aligned, the synchronous change rate of the main frequency amplitude of each channel is counted, and the synchronous change rate is set as the spectral aggregation flow index.

[0040] According to the energy spectrum distribution of each channel, a main energy frequency trajectory sequence is established, the time change rate of the main energy frequency is calculated, the rate of disturbance energy transfer from low frequency to high frequency is extracted, and the time change rate is set as the perturbation field transition rate.

[0041] The application further provides that the logic of the anomaly discrimination module comprises:

[0042] Based on the multi-dimensional vortex spectrum linkage factor of all spatial partitions under the same time window, a multi-dimensional feature vector with the vortex density index, the spectral aggregation flow index and the perturbation field transition rate as components is constructed, and the corresponding feature vectors of each partition form a multi-dimensional feature space.

[0043] The feature vectors of each partition in the multi-dimensional feature space are clustered, the local density of each partition and the distance between the partition and a higher density partition are calculated, the partition with high local density and far distance from other high density partitions is identified as a density peak, and a preset number of high density peak clustering centers are determined.

[0044] According to the high density clustering center, the remaining partitions are attributed to the nearest clustering center according to the feature distance, and a high condensation area candidate set is formed.

[0045] For each high condensation area candidate set, based on the spectral aggregation flow index in the same region, the consistency of energy distribution, the peak synthesis intensity and the correlation between channels in the region are calculated, and are set as the spectral convergence factor.

[0046] A spectral convergence factor threshold is set, the high condensation area candidate set with the spectral convergence factor higher than the corresponding threshold is determined as an abnormal high condensation area, the region is marked for spatial partition, and an abnormal partition index is output.

[0047] The application is further configured that the calculation logic of the consistency of the inter-channel energy distribution in the region, the peak synthesis intensity and the correlation includes:

[0048] The energy data extracted at the main frequency component of all channels in the spatial partition is obtained, the mean and standard deviation of the main frequency energy of all channels are obtained, and the ratio of the value and the standard deviation is set as the energy distribution consistency index;

[0049] The main frequency energy of all channels is weighted and summed to obtain the synthesis intensity of the peak energy in the partition, which is set as the peak synthesis intensity, and is used to measure the overall aggregation level of multi-channel energy;

[0050] The correlation coefficient between any two channels is calculated for the main frequency energy data of each channel, and the absolute value mean of all correlation coefficients is set as the correlation index of the inter-channel energy distribution.

[0051] The application is further configured that the construction logic of the multi-dimensional diffusion path model includes:

[0052] For the identified abnormal high condensation area, in the continuous time window, based on the time partition space distribution and the region label result, the time sequence trajectory tracking is carried out, the abnormal high condensation areas with adjacent spatial positions in adjacent time windows are determined as the same object, and the continuous spatial trajectory of the abnormal area is generated;

[0053] For each abnormal high condensation area, the spatial center point coordinates in each time window are extracted, and based on the position change of the center point in the time sequence, the diffusion speed and the main diffusion direction of the abnormal area in the space are quantified, the diffusion speed is calculated by the center point displacement and the time interval in adjacent time windows, and the main diffusion direction is determined by the time sequence position change trend;

[0054] The number of continuous time windows experienced by each abnormal high condensation area from the first identification to the disappearance is counted, and the time length corresponding to the number of continuous time windows is defined as the duration of the abnormal high condensation area;

[0055] The spatial diffusion speed, displacement direction and duration of the abnormal high condensation area are integrated to establish a multi-dimensional diffusion path model.

[0056] The application is further configured that based on the multi-dimensional diffusion path model, multi-level response thresholds are set for risk warning, including:

[0057] When the spatial diffusion speed is less than the first spatial diffusion speed threshold and the duration is less than the first duration threshold, it is determined as a low risk state, and an abnormal state record is made;

[0058] When the spatial diffusion speed is greater than or equal to the first spatial diffusion speed threshold value and less than the second spatial diffusion speed threshold value and the duration is greater than or equal to the first duration threshold value and less than the second duration threshold value, a medium-risk state is determined, and a local risk prompt is triggered;

[0059] When the spatial diffusion speed is greater than or equal to the second spatial diffusion speed threshold value and the duration is greater than or equal to the second duration threshold value, a high-risk state is determined, and a remote alarm is performed.

[0060] The application provides an explosion-proof intelligent video monitoring control system, comprising a collection module: for real-time collection of multispectral video data in an explosion-proof monitoring area, based on a preset spatial grid division, multi-channel high-frequency synchronous sampling of a sub-area is performed to generate a partitioned time sequence original data set; a feature extraction module: for extraction of continuous frame pixel intensity changes in a time window from the partitioned time sequence original data set, through time domain transformation and frequency domain transformation processing, light field disturbance features are obtained and a native fluctuation spectrum is constructed; a linkage factor construction module: for construction of a multi-channel disturbance parameter joint matrix based on native fluctuation spectra of different channels of the same area, multi-dimensional vortex spectrum linkage factors are obtained, the multi-dimensional vortex spectrum linkage factors include a vortex density index, a spectrum aggregation flow index and a perturbation field change rate; an anomaly discrimination module: for high-dimensional density peak clustering of multi-dimensional vortex spectrum linkage factors of all spatial partitions, a vortex disturbance aggregation trend significant area is identified, a spectrum convergence factor is calculated, and an abnormal high condensation area in the space is identified and marked according to the spectrum convergence factor; a path modeling module: for time sequence tracking of the abnormal high condensation area, extraction of a spatial diffusion speed, a displacement direction and a duration, and construction of a multi-dimensional diffusion path model; a risk warning module: for setting of multi-level response threshold values based on the multi-dimensional diffusion path model, risk warning and production of beneficial effects, including:

[0061] Pixel-level anomaly early identification: the pixel intensity change analysis in the time window and the native fluctuation spectrum construction can sensitively capture small disturbances and early anomaly signals, and effectively enhance the identification capability of early risks in complex scenes such as smoke, flame and gas leakage;

[0062] Multi-channel disturbance collaborative modeling: through extraction of multi-dimensional linkage factors such as the vortex density index, the spectrum aggregation flow index and the perturbation field change rate, the multi-channel, cross-area disturbance collaboration and spatial energy aggregation phenomenon can be quantitatively reflected, and the deficiency of traditional systems in multi-source linkage anomaly analysis is made up;

[0063] Intelligent anomaly aggregation area discrimination: through high-dimensional density peak clustering and spectrum convergence factor statistical methods, automatic identification and partition marking of the abnormal high condensation area in the space are realized, the risk of anomaly missed detection and false alarm is greatly reduced, and the accuracy and intelligent level of anomaly detection are improved;

[0064] Dynamic path and diffusion process tracking: through the modeling of multi-dimensional parameters such as trajectory association, diffusion speed, displacement direction and duration of the abnormally high condensation area, the dynamic evolution of the abnormal area can be tracked in real time, providing data support for subsequent risk analysis and event tracing.

[0065] The above description is only a summary of the technical solutions of the present application. In order to more clearly understand the technical means of the present application, the following specific embodiments of the present application can be implemented according to the content of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the following specific embodiments of the present application are described. BRIEF DESCRIPTION OF DRAWINGS

[0066] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the following will briefly introduce the drawings needed in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and those skilled in the art can obtain other drawings according to these drawings without creating any creative labor. In the drawings:

[0067] Figure 1 A flow chart of an explosion-proof intelligent video monitoring control system is shown for an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0068] The embodiments of the present application will be described below with reference to the drawings and preferred embodiments, and those skilled in the art can easily understand other advantages and effects of the present application from the content disclosed in the specification. The present application can also be implemented or applied by different specific embodiments, and the details in the specification can be modified or changed based on different views and applications without departing from the spirit of the present application. It should be understood that the preferred embodiments are only for illustrating the present application, but not for limiting the protection scope of the present application.

[0069] It should be noted that the diagrams provided in the following embodiments only illustrate the basic concept of the present application in a schematic manner, and the drawings only show the components related to the present application, but not the number, shape and size of the components when actually implemented. The actual implementation of each component may be arbitrarily changed in type, number and proportion, and the layout pattern of the components may also be more complex.

[0070] In the following description, a large number of details are discussed to provide a more thorough explanation of the embodiments of the present application, however, it is obvious to those skilled in the art that the embodiments of the present application can be implemented without these specific details, and in other embodiments, the known structures and devices are shown in the form of block diagrams rather than in the form of details, to avoid making the embodiments of the present application difficult to understand.

[0071] An explosion-proof intelligent video monitoring control system, as shown in Figure 1 includes:

[0072] The acquisition module is configured to acquire multi-spectral video data in the explosion-proof monitoring area in real time, perform multi-channel high-frequency synchronous sampling on the sub-regions based on preset spatial grid division, and generate a partitioned time-series original data set.

[0073] The feature extraction module is configured to extract the changes in the pixel intensity of the continuous frames within the time window of the partitioned time-series original data set, process the time-domain transformation and the frequency-domain transformation, obtain the light field disturbance features, and construct the original wave fluctuation spectrum.

[0074] The linkage factor construction module is configured to construct a multi-channel disturbance parameter joint matrix based on the original wave fluctuation spectrum of different channels in the same region, obtain a multi-dimensional vortex spectrum linkage factor, and the multi-dimensional vortex spectrum linkage factor includes a vortex density index, a spectrum aggregation flow index, and a perturbation field change rate.

[0075] The anomaly discrimination module is configured to perform high-dimensional density peak clustering on the multi-dimensional vortex spectrum linkage factor of all spatial partitions, identify a region with a significant vortex disturbance aggregation trend, calculate a spectrum aggregation factor, and identify and label an abnormal high-aggregation region in the space according to the spectrum aggregation factor.

[0076] The path modeling module is configured to track the abnormal high-aggregation region in the time series, extract the spatial diffusion speed, displacement direction, and duration, and construct a multi-dimensional diffusion path model.

[0077] The risk early warning module is configured to set a multi-level response threshold based on the multi-dimensional diffusion path model and perform risk early warning.

[0078] The application further provides that the generation logic of the partitioned time-series original data set includes:

[0079] The explosion-proof monitoring area is spatially divided according to the preset spatial grid parameters, and the region is discretized into a plurality of sub-regions; specifically, according to the actual range and monitoring requirements of the explosion-proof monitoring area, the spatial grid parameters are preset, and the overall region is evenly or as needed divided into a plurality of sub-regions, and each sub-region is taken as an independent video monitoring acquisition unit.

[0080] The explosion-proof intelligent video monitoring is arranged corresponding to each sub-region, and the explosion-proof intelligent video monitoring is configured with a multi-spectral video acquisition unit for multi-channel synchronous acquisition control; specifically, the explosion-proof intelligent video monitoring device is arranged in each sub-region, and the multi-spectral video acquisition unit is configured for the explosion-proof intelligent video monitoring device, which can synchronously acquire infrared, visible light, and ultraviolet channel data.

[0081] Through unified synchronous clock and sampling trigger mechanism, high-frequency data acquisition is carried out in the same time window, and video frame data containing infrared, visible light and ultraviolet are obtained; specifically, through the global synchronous clock and the unified sampling trigger mechanism, the multispectral video acquisition units of all sub-regions are ensured to carry out high-frequency and synchronous data acquisition in the same time window, and the timing alignment between different regions and different channels is ensured; at each sampling time, multispectral video frames of each sub-region are collected, such as infrared, visible light and ultraviolet, to form a data stream classified by channel;

[0082] The multispectral video frames collected by each sub-region at all sampling times are structured and arranged according to the partition number, time sequence and channel type to generate a partition time sequence original data set; specifically, the multispectral video frames of all sub-regions and all sampling times are structured and arranged according to the partition number, time sequence and channel type to generate a partition time sequence original data set, which is uploaded to a data processing center.

[0083] The application further provides that the construction logic of the original fluctuation spectrum comprises:

[0084] The partition time sequence original data set is segmented according to a set time window, and continuous multi-frame video images of each partition and each channel are segmented; specifically, the continuous multi-frame video images of each partition and each channel in the partition time sequence original data set are time segmented according to the set time window, and the image frames in each time window are taken as a group for subsequent processing;

[0085] For the continuous video frames in each time window, the pixel intensity sequence in the time window is extracted according to the pixel coordinates to generate a time sequence pixel intensity matrix; specifically, for the video frame sequence in each time window, all pixel coordinates are traversed, and the pixel intensity value of the pixel point in the time window is extracted to form a time sequence pixel intensity matrix arranged in time;

[0086] The time sequence pixel intensity matrix is subjected to first-order difference to obtain an inter-frame pixel intensity change sequence; specifically, the time sequence pixel intensity matrix is subjected to first-order difference operation to obtain the intensity change sequence of the pixel between continuous frames, reflecting the slight disturbance of the local light field;

[0087] Temporal statistical features are extracted from the pixel intensity change sequence, including sliding mean, variance, kurtosis and skewness; specifically, for each pixel intensity change sequence, the sliding mean, variance, kurtosis and skewness and other temporal statistical features are calculated to reflect the change trend, fluctuation amplitude and non-Gaussian characteristics;

[0088] The Fourier transform is performed on each pixel intensity variation sequence to extract frequency domain statistical features, including main frequency component, energy spectrum distribution, local amplitude peak value and instantaneous phase parameter; specifically, the Fourier transform is performed on each pixel intensity variation sequence to extract frequency domain feature parameters such as main frequency component, energy spectrum distribution, local amplitude peak value and instantaneous phase, which are used to reflect the dominant period and energy distribution characteristics of the disturbance;

[0089] The time domain statistical features and the frequency domain statistical features of all pixels and all channels in each time window are summarized according to the spatial partition to obtain a partitioned light field disturbance feature group; specifically, the time domain and frequency domain feature parameters of all pixels and all channels in each time window are statistically reduced according to the spatial partition to form a partitioned light field disturbance feature group;

[0090] The partitioned light field disturbance feature group is combined to generate a native fluctuation spectrum data group with the partition, channel and time window as the three indexes; specifically, the disturbance feature groups under each partition, each channel and each time window are combined and stored according to the three index structure to generate a native fluctuation spectrum data group.

[0091] The application further provides that the acquisition logic of the multi-dimensional eddy current spectrum linkage factor comprises:

[0092] The native fluctuation spectrum data groups of different channels under the same partition and the same time window are aligned according to the main disturbance parameter to construct a multi-channel disturbance parameter joint matrix; specifically, the main frequency component and the corresponding local amplitude peak value of each channel are extracted from the native fluctuation spectrum data of different channels under the same partition and the same time window, and the main disturbance parameters of different channels are aligned according to the unified main frequency index through interpolation or frequency point matching to generate a multi-channel disturbance parameter joint matrix;

[0093] Based on the energy spectrum distribution, the spatial aggregation degree of the disturbance energy in the partition is calculated to extract a vortex density index; specifically, based on the energy spectrum distribution in the native fluctuation spectrum, a high-frequency energy threshold is set, high-frequency disturbance energy points with energy exceeding the threshold are screened, and the density of these high-frequency energy points in the partition space is counted, and the density value is taken as the vortex density index to reflect the energy aggregation characteristics;

[0094] According to the main frequency component and the local amplitude peak value, the energy synchronous change rate and the peak energy synthesis intensity of different channels in the main frequency band are counted to extract a spectrum aggregation flow index; specifically, for the main frequency component and the local amplitude peak value, the energy synchronous change rate of each channel in the main frequency band and the weighted synthesis intensity of the main frequency amplitude of each channel are counted in the aligned multi-channel parameter matrix, and the synthesis result is defined as the spectrum aggregation flow index to reflect the energy aggregation and synchronous change degree;

[0095] According to the gradient of the change of the energy spectrum with time in the energy spectrum distribution, the rate of the transfer of the perturbation energy from low frequency to high frequency is calculated to extract the perturbation field rate; specifically, the trajectory of the change of the energy spectrum with time in the original wave spectrum is analyzed, the rate of the change of the main energy frequency with time is tracked, the speed of the energy migration from low frequency to high frequency in the frequency spectrum is calculated, and the rate is taken as the perturbation field rate for measuring the energy migration and dynamic activity.

[0096] The vortex density index, the spectrum aggregation flow index and the perturbation field rate are set as the multi-dimensional vortex spectrum linkage factors; specifically, the vortex density index, the spectrum aggregation flow index and the perturbation field rate obtained above are taken as the multi-dimensional vortex spectrum linkage factors together as the input features for subsequent high-dimensional clustering, anomaly discrimination and region marking.

[0097] The application is further provided to align according to the main disturbance parameters to construct a multi-channel disturbance parameter joint matrix, including:

[0098] Under each channel, the main frequency component, the local amplitude peak value and the instantaneous phase parameter are extracted from the original wave spectrum data set and set as the main disturbance parameters; specifically, in each partition and each time window, the original wave spectrum data set of each channel (including infrared, visible light and ultraviolet) is analyzed respectively, the main frequency component, the local amplitude peak value and the instantaneous phase parameter of the channel in the time window are extracted, and the main disturbance parameter set of the channel is taken as the main disturbance parameter set.

[0099] The main frequency components of all channels in the same partition and the same time window form a main frequency index set; specifically, the main frequency components extracted by all channels in the current partition and the current time window are summarized to form the main frequency index set. The index set provides a unified frequency reference for the alignment of all channel parameters;

[0100] For the main frequency index set, if there is a slight shift in the main frequency components of different channels, the parameter values of each channel are uniformly mapped to the frequency points of the main frequency index set through linear interpolation or nearest neighbor matching algorithm to ensure that the main disturbance parameters of each channel correspond to each other in the frequency dimension; specifically, for the possible slight shift of the main frequency components between different channels, the linear interpolation method or the nearest neighbor frequency matching algorithm is used to map the main disturbance parameter values of each channel to the frequency points of the main frequency index set to realize the one-to-one correspondence of the main disturbance parameters of each channel in the frequency dimension.

[0101] The aligned channel main frequency components, local amplitude peak values and instantaneous phase parameters are arranged according to the main frequency indexes to form a multi-channel disturbance parameter joint matrix, wherein each row of the matrix corresponds to a channel, and each column corresponds to a disturbance parameter at a main frequency point; specifically, the aligned channel main frequency components, local amplitude peak values and instantaneous phase parameters are combined into a multi-channel disturbance parameter joint matrix according to the arrangement order of the main frequency index set. Each row of the matrix represents a channel, each column represents a main frequency point in the main frequency index set, and each element is a disturbance parameter of the corresponding channel at the main frequency point.

[0102] The application is further provided that the extraction logic of the vortex density index, the spectral aggregation flow index and the perturbation field rate includes:

[0103] Based on the energy spectrum distribution of each partition in the original fluctuation spectrum of each channel, a high-frequency energy threshold is set, high-frequency disturbance energy points with energy greater than the high-frequency energy threshold are screened, the density of the high-frequency disturbance energy points in the partition space is calculated, and the density is set as the vortex density index; specifically, based on the original fluctuation spectrum of each channel, a high-frequency energy threshold (for example, the mean value or the mean value plus double standard deviation of the energy spectrum distribution of each channel) is set in each partition, and high-frequency disturbance energy points with energy higher than the threshold are screened. The distribution of all high-frequency disturbance energy points in space is counted, the spatial density (i.e. the proportion of high-frequency point number to total pixel point number) of the high-frequency disturbance energy points in the partition is calculated, and the density is set as the vortex density index of the partition, which is used to represent the agglomeration phenomenon of high-frequency disturbance energy;

[0104] The main frequency components and corresponding local amplitude peak values of each channel in each partition are aligned, the synchronous change rate of the main frequency amplitude of each channel is counted, and the synchronous change rate is set as the spectral aggregation flow index; specifically, the main frequency index of the main frequency components and the local amplitude peak values of different channels in each partition is aligned, the amplitude sequence of all channels at each main frequency point is counted, the synchronous change rate (for example, the correlation coefficient, covariance or the reciprocal of standard deviation) of the main frequency amplitude of each channel at the same main frequency point is calculated, the mean value or weighted synthesis of the synchronous change rates of all main frequency points is taken as the spectral aggregation flow index of the partition, which is used to represent the cooperative strength of energy disturbance of different channels;

[0105] According to the energy spectrum distribution of each channel, a main energy frequency track sequence is established, the time variation rate of the main energy frequency is calculated, and the time variation rate is set as the perturbation field rate; specifically, for each partition and each channel, the evolution process of the energy spectrum distribution on the time axis is analyzed, the main energy frequency position (i.e. the frequency point of the maximum energy spectrum) in each time window is recorded, the time sequence track of the main energy frequency is established, the variation rate of the frequency with time (such as the ratio of the difference between the main energy frequencies of adjacent time windows to the time interval) is calculated, and the average variation rate is taken as the perturbation field rate, which is used to reflect the activity degree of the disturbance energy transfer from low frequency to high frequency.

[0106] The application further provides that the logic of the anomaly discrimination module comprises:

[0107] Based on the multi-dimensional eddy current spectrum linkage factors of all spatial partitions in the same time window, a multi-dimensional feature vector with vortex density index, spectrum aggregation flow index and perturbation field rate as components is constructed, and the feature vectors corresponding to the partitions form a multi-dimensional feature space; specifically, for all spatial partitions, the vortex density index, spectrum aggregation flow index and perturbation field rate of each partition are collected in the same time window, a multi-dimensional feature vector of a stereoscopic coordinate is formed, and a high-dimensional feature space is formed;

[0108] The feature vectors of the partitions in the multi-dimensional feature space are clustered, the local density of each partition and the distance between the partition and the higher density partition are calculated, the partition with high local density and far away from other high density partitions is identified as a density peak, and a preset number of high density peak clustering centers are determined; specifically, in the high-dimensional feature space, density peak clustering is performed on all partition feature vectors, and the local density (such as the number of partitions within a certain distance radius) of each partition in the feature space is calculated. Then, the distance between each partition and all higher density partitions is calculated, and the point with high local density and far away from other high density partitions is selected as the density peak (clustering center). According to the preset number of clustering centers, the high density peaks are finally selected as the high condensation zone centers;

[0109] According to the high density clustering centers, the remaining partitions are attributed to the nearest clustering center according to the feature distance, and a high condensation zone candidate set is formed; specifically, for the remaining partitions, the distance between the feature vector and each clustering center is calculated, and the partition is attributed to the nearest center, thereby forming several high condensation zone candidate sets;

[0110] For each high condensation region candidate set, based on the spectral aggregation flow index in the same region, the consistency of the inter-channel energy distribution in the region, the peak synthesis intensity and the correlation are calculated, and the spectral convergence factor is set; Specifically, for each high condensation region candidate set, based to the spectral aggregation flow index of each sub-region in the region, the consistency of the inter-channel energy distribution in the region, the peak synthesis intensity and the correlation (such as mean, standard deviation, correlation coefficient and other statistical quantities) are calculated. The spectral convergence factor of the above statistical quantity synthesis region is used to reflect the synchronization and cooperation characteristics of the energy distribution between the sub-regions.

[0111] The spectral convergence factor threshold is set, and for the high condensation region candidate set whose spectral convergence factor is higher than the corresponding threshold, the abnormal high condensation region is determined, the spatial partition marking is performed on the region, and the abnormal region index is output; Specifically, the discrimination threshold of the spectral convergence factor is set, and the high condensation region candidate set higher than the threshold is determined as the abnormal high condensation region, and the spatial partition marking and the abnormal region index output are completed.

[0112] The application further sets that the calculation logic of the consistency of the inter-channel energy distribution in the region, the peak synthesis intensity and the correlation is:

[0113] The energy data extracted at the main frequency component of all channels in the spatial partition is obtained, the mean and standard deviation of the main frequency energy of all channels are obtained, and the ratio of the value and the standard deviation is set as the energy distribution consistency index; Specifically, for all channels in the same region, the energy data is extracted at the main frequency component, and the mean and standard deviation of the main frequency energy of all channels are calculated; The ratio of the main frequency energy standard deviation to the mean is defined as the energy distribution consistency index of the region. The smaller the ratio is, the more uniform the main frequency energy distribution of each channel is, and the higher the consistency is;

[0114] The main frequency energy of all channels is weighted and summed, the synthesis intensity of the peak energy in the partition is obtained, and the peak synthesis intensity is set, which is used to measure the overall aggregation level of the multi-channel energy; Specifically, the main frequency energy of all channels in the region is weighted and summed according to the set weight (such as equal weight or weight according to the importance of the channel), and the total energy is obtained as the peak energy synthesis intensity; The index reflects the overall aggregation effect of the multi-channel energy, and the larger the value is, the higher the overall level of the energy between the channels is;

[0115] The correlation coefficient between any two channels is calculated for the main frequency energy data of each channel, and the absolute value mean of all correlation coefficients is set as the correlation index of the inter-channel energy distribution; Specifically, the main frequency energy data of each channel is combined in pairs, and the Pearson correlation coefficient between the main frequency energy of the channels is calculated; The absolute value of all correlation coefficients is taken and the mean is calculated, and the correlation index of the inter-channel energy distribution is obtained. The larger the value is, the more synchronous and cooperative the energy change between the channels is.

[0116] The application is further configured that the construction logic of the multi-dimensional diffusion path model comprises:

[0117] For the identified abnormal high condensation area, in a continuous time window, based on the time partition space distribution and the area marking result, time sequence trajectory tracking is performed, the abnormal high condensation areas with adjacent spatial positions in adjacent time windows are determined as the same object, and a continuous spatial trajectory of the abnormal area is generated.

[0118] For each abnormal high condensation area, the spatial center point coordinates in each time window are extracted, and based on the position change of the center point in the time sequence, the diffusion speed and the main diffusion direction of the abnormal area in the space are quantified.

[0119] The number of continuous time windows experienced by each abnormal high condensation area from being identified for the first time to disappearing is counted, and the time length corresponding to the number of continuous time windows is defined as the duration of the abnormal high condensation area.

[0120] The spatial diffusion speed, displacement direction and duration of the abnormal high condensation area are integrated to establish a multi-dimensional diffusion path model.

[0121] The application is further configured that multi-level response thresholds are set based on the multi-dimensional diffusion path model to perform risk early warning, comprising:

[0122] When the spatial diffusion speed is less than the first spatial diffusion speed threshold and the duration is less than the first duration threshold, it is determined as a low-risk state, and an abnormal state record is performed.

[0123] When the spatial diffusion speed is greater than or equal to the first spatial diffusion speed threshold value and less than the second spatial diffusion speed threshold value and the duration is greater than or equal to the first duration threshold value and less than the second duration threshold value, it is determined as a medium risk state, and a local risk prompt is triggered;

[0124] When the spatial diffusion speed is greater than or equal to the second spatial diffusion speed threshold value and the duration is greater than or equal to the second duration threshold value, it is determined as a high risk state, and a remote alarm is performed; specifically, for each abnormal high condensation area, the spatial diffusion speed and the duration existing time of the area are continuously collected in a continuous monitoring period. The system sets the first and second threshold values of the spatial diffusion speed and the duration in advance according to actual application requirements, as the grading basis for low risk, medium risk and high risk.

[0125] When the spatial diffusion speed of the abnormal high condensation area is lower than the first threshold value and the duration is lower than the first threshold value, it is determined as a low risk state, and only an abnormal state record is made;

[0126] When the spatial diffusion speed is between the first threshold value and the second threshold value and the duration is between the first threshold value and the second threshold value, it is determined as a medium risk state, and the system automatically triggers a local risk prompt;

[0127] When the spatial diffusion speed is greater than or equal to the second threshold value and the duration is greater than or equal to the second threshold value, it is determined as a high risk state, and a remote alarm and emergency response measures are immediately started.

[0128] According to the risk level, the system respectively takes grading disposal such as abnormal archiving, local prompting, remote alarming and the like, and realizes automatic dynamic risk management.

[0129] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited to this, any person skilled in the art can easily think of changes or replacements within the technical range disclosed in the present application, which should be covered in the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. An explosion-proof intelligent video monitoring control system, characterized in that, The method comprises the following steps: The acquisition module is used for real-time acquisition of multi-spectral video data in the explosion-proof monitoring area, multi-channel high-frequency synchronous sampling of sub-regions based on a preset spatial grid division, and generation of partitioned time sequence original data sets; The feature extraction module is used for extracting the pixel intensity change of the continuous frames in the time window, obtaining the light field disturbance features through time domain transformation and frequency domain transformation processing, and constructing the original wave spectrum, wherein the original wave spectrum is a three-index combination of the partition, channel and time window, and the data group obtained by processing the light field disturbance features through time domain transformation and frequency domain transformation; The linkage factor construction module is used for constructing a multi-channel disturbance parameter joint matrix based on the original wave spectrum of different channels in the same region, obtaining a multi-dimensional vortex spectrum linkage factor, and the multi-dimensional vortex spectrum linkage factor comprises a vortex density index, a spectrum aggregation flow index and a perturbation field change rate, wherein the acquisition logic of the multi-dimensional vortex spectrum linkage factor comprises the following steps: aligning the original wave spectrum data groups of different channels in the same partition and the same time window according to the main disturbance parameter, constructing a multi-channel disturbance parameter joint matrix, calculating the spatial aggregation degree of the disturbance energy in the partition based on the energy spectrum distribution, extracting the vortex density index, according to the main frequency component and the local amplitude peak value, calculating the energy synchronous change rate and the peak energy synthesis intensity of different channels in the main frequency band, extracting the spectrum aggregation flow index, calculating the rate of disturbance energy transfer from low frequency to high frequency according to the energy spectrum change gradient in the energy spectrum distribution, extracting the perturbation field change rate, and setting the vortex density index, the spectrum aggregation flow index and the perturbation field change rate as the multi-dimensional vortex spectrum linkage factor; The anomaly discrimination module is used for high-dimensional density peak clustering of the multi-dimensional vortex spectrum linkage factor of all spatial partitions, identifying a vortex disturbance aggregation trend significant region, calculating a spectrum convergence factor, and identifying and marking an abnormal high condensation area in the space according to the spectrum convergence factor, wherein the spectrum convergence factor is composed of an energy distribution consistency index obtained based on the ratio of the main frequency energy standard deviation to the mean value, a peak energy synthesis intensity obtained by weighted summation of the main frequency energy of all channels, and a correlation index represented by the mean value of the absolute values of the correlation coefficients of the main frequency energy of two channels; The path modeling module is used for time sequence tracking of the abnormal high condensation area, extracting the spatial diffusion speed, displacement direction and duration, and constructing a multi-dimensional diffusion path model; The risk early warning module is used for setting multi-level response thresholds based on the multi-dimensional diffusion path model and performing risk early warning.

2. The explosion-proof intelligent video monitoring control system according to claim 1, characterized in that, The generation logic of the partitioned time sequence original data set comprises the following steps: The explosion-proof monitoring area is spatially divided according to the preset spatial grid parameters, and the area is discretized into multiple sub-regions; An explosion-proof intelligent video monitoring device is arranged corresponding to each sub-region, and the explosion-proof intelligent video monitoring device is configured with a multi-spectral video acquisition unit for multi-channel synchronous acquisition control; High-frequency data acquisition is performed in the same time window through a unified synchronous clock and a sampling trigger mechanism to obtain video frame data containing infrared, visible light and ultraviolet light; The multi-channel video frames collected by each sub-region at all sampling time points are structured and arranged according to the partition number, time sequence and channel type to generate a partition time sequence original data set.

3. The explosion-proof intelligent video monitoring control system according to claim 1, characterized in that, The construction logic of the original wave fluctuation spectrum includes: The partition time sequence original data set is segmented according to the set time window, and the continuous multi-frame video images of each partition and each channel are processed; For the continuous video frames in each time window, the pixel intensity sequence in the time window is extracted according to the pixel coordinates to generate a time sequence pixel intensity matrix; The time sequence pixel intensity matrix is first-order differentiated to obtain an inter-frame pixel intensity change sequence; The time domain statistical features of the pixel intensity change sequence are extracted, including sliding mean, variance, kurtosis and skewness; The frequency domain statistical features of each pixel intensity change sequence are extracted by fast Fourier transform, including the main frequency component, energy spectrum distribution, local amplitude peak value and instantaneous phase parameter; The time domain statistical features and the frequency domain statistical features of all pixels and all channels in each time window are summarized according to the spatial partition to obtain a partition light field disturbance feature group; The partition light field disturbance feature group is combined with the partition, channel and time window as a three-index to generate an original wave fluctuation spectrum data group.

4. The explosion-proof intelligent video monitoring control system according to claim 3, characterized in that, According to the alignment of the main disturbance parameters, a multi-channel disturbance parameter joint matrix is constructed, including: In each channel, the main frequency component, local amplitude peak value and instantaneous phase parameter are extracted from the original wave fluctuation spectrum data group and set as the main disturbance parameters; The main frequency components of all channels in the same partition and the same time window form a main frequency index set; For the main frequency index set, if there is a slight shift in the main frequency components of different channels, the parameter values of each channel are uniformly mapped to the frequency points of the main frequency index set through linear interpolation or nearest neighbor matching algorithm to ensure that the main disturbance parameters of each channel correspond one by one in the frequency dimension; The aligned main frequency components, local amplitude peak values and instantaneous phase parameters of each channel are arranged according to the main frequency index to form a multi-channel disturbance parameter joint matrix, wherein each row of the matrix corresponds to a channel and each column corresponds to a disturbance parameter at a main frequency point.

5. The explosion-proof intelligent video monitoring control system according to claim 3, characterized in that, The extraction logic of the vortex density index, spectral aggregation flow index and perturbation field transition rate includes: Based on the energy spectrum distribution of each partition in the original wave fluctuation spectrum of each channel, a high-frequency energy threshold is set to screen high-frequency disturbance energy points with energy greater than the high-frequency energy threshold, and the density of the high-frequency disturbance energy points in the partition space is calculated to set the density as the vortex density index; The main frequency components and the corresponding local amplitude peak values of each channel in each partition are aligned, and the synchronous change rate of the main frequency amplitude of each channel is calculated to set the synchronous change rate as the spectral aggregation flow index; According to the energy spectrum distribution of each channel, a main energy frequency trajectory sequence is established, the time change rate of the main energy frequency is calculated, the rate of disturbance energy transfer from low frequency to high frequency is extracted, and the time change rate is set as the perturbation field transition rate.

6. The explosion-proof intelligent video monitoring control system according to claim 1, characterized in that, The logic of the abnormality discrimination module includes: Based on the multi-dimensional vortex spectrum linkage factor of all spatial partitions under the same time window, a multi-dimensional feature vector with the vortex density index, the spectral aggregation flow index and the perturbation field transition rate as components is constructed, and the feature vectors of each partition form a multi-dimensional feature space. Clustering each partition feature vector in the multi-dimensional feature space, calculating the local density of each partition and the distance between partitions with higher density, identifying the partition with high local density and far away from other high-density partitions as a density peak, and determining a preset number of high-density peak clustering centers; According to the high-density clustering center, the remaining partitions are attributed to the nearest clustering center according to the feature distance, forming a high condensation area candidate set; For each high condensation area candidate set, based on the spectrum aggregation flow index in the same region, the consistency of the energy distribution between channels, the peak synthesis intensity and the correlation in the region are calculated, and the spectral aggregation factor is set; Set the spectral aggregation factor threshold, for the high condensation area candidate set with the spectral aggregation factor higher than the corresponding threshold, determine the abnormal high condensation area, mark the region for spatial partition, and output the abnormal partition index.

7. The explosion-proof intelligent video monitoring control system according to claim 6, characterized in that, The calculation logic of the consistency of the energy distribution between channels, the peak synthesis intensity and the correlation in the region is: For the energy data extracted at the main frequency component of all channels in the spatial partition, the mean and standard deviation of the main frequency energy of all channels are obtained, and the ratio of the value and the standard deviation is set as the energy distribution consistency index; The weighted sum of the main frequency energy of all channels is obtained, and the synthesis intensity of the peak energy in the partition is set as the peak synthesis intensity, which is used to measure the overall aggregation level of multi-channel energy; For each channel main frequency energy data, the correlation coefficient between any two channels is calculated, and the absolute value mean of all correlation coefficients is set as the correlation index of the energy distribution between channels.

8. The explosion-proof intelligent video monitoring control system according to claim 1, characterized in that, The construction logic of the multi-dimensional diffusion path model includes: For the identified abnormal high condensation area, in the continuous time window, based on the time partition space distribution and the region marking result, the time sequence trajectory tracking is carried out, the abnormal high condensation areas with adjacent spatial positions in adjacent time windows are determined as the same object, and the continuous spatial trajectory of the abnormal area is generated; For each abnormal high condensation area, the spatial center point coordinates in each time window are extracted, and based on the position change of the center point in the time sequence, the diffusion speed and the main diffusion direction of the abnormal area in space are quantified, the diffusion speed is calculated by the center point displacement and the time interval in adjacent time windows, and the main diffusion direction is determined by the trend of the time sequence position change; The number of continuous time windows experienced by each abnormal high condensation area from the first identification to the disappearance is counted, and the time length corresponding to the number of continuous time windows is defined as the duration of the abnormal high condensation area; The spatial diffusion speed, displacement direction and duration of the abnormal high condensation area are integrated to establish a multi-dimensional diffusion path model.

9. The explosion-proof intelligent video monitoring control system according to claim 1, wherein, Based on the multi-dimensional diffusion path model, set multi-level response thresholds for risk warning, including: When the spatial diffusion speed is less than the first spatial diffusion speed threshold and the duration is less than the first duration threshold, it is determined as a low risk state, and an abnormal state record is made; When the spatial diffusion speed is greater than or equal to the first spatial diffusion speed threshold and less than the second spatial diffusion speed threshold, and the duration is greater than or equal to the first duration threshold and less than the second duration threshold, it is determined as a medium risk state, and a local risk prompt is triggered. When the spatial diffusion speed is greater than or equal to the second spatial diffusion speed threshold value and the duration is greater than or equal to the second duration threshold value, a high-risk state is determined, and remote alarm is performed.

Images