A visual communication system based on weak current engineering and a method for operating the same

Abstract

The application relates to the technical field of intelligent industry, and discloses a visual communication system based on weak current engineering and a running method thereof. The system comprises a multi-modal signal acquisition module, an intelligent data processing module, a double-channel communication transmission module, an interactive visualization module and an autonomous control system module. The method comprises the following steps: synchronously acquiring video, audio and environmental parameters through a distributed sensor array; eliminating multi-source signal deviation by adopting a space-time alignment algorithm; realizing feature fusion and quality evaluation based on a deep neural network; dynamically selecting a wired and wireless hybrid transmission mode; constructing an AR-enhanced visual human-computer interface; and implementing fault self-diagnosis and fault-tolerant recovery mechanism. The application reduces the error code rate of the communication system, guarantees the continuity and reliability of the communication process, and improves the human-computer interaction efficiency and emergency response speed.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent industry, in particular to a visual communication system based on weak current engineering and a running method thereof. BACKGROUND

[0002] Intelligent industry is a new type of industrial system, which integrates various terminals with environmental perception ability, computing mode based on ubiquitous technology, mobile communication and other technologies into various links of industrial production, improves manufacturing efficiency, improves product quality, reduces product cost and resource consumption, and promotes traditional industry to a new stage of intelligentization.

[0003] At present, due to the characteristics of strong electromagnetic interference and variable environment of weak current engineering application scene, when real-time transmission of multi-modal data is carried out, the deployed communication equipment is affected by environmental electromagnetic noise, and cannot dynamically adjust the signal processing strategy, if the signal distortion exceeds the tolerance range, it may cause video lag and audio interruption, and cannot guarantee the stability of communication quality.

[0004] Therefore, the present application provides a visual communication system based on weak current engineering and a running method thereof to solve the above problems. SUMMARY

[0005] In view of the deficiencies of the prior art, the present application provides a visual communication system based on weak current engineering and a running method thereof, which solves the problem of being unable to dynamically adjust the signal processing strategy and being unable to guarantee the stability of communication quality as described in the background.

[0006] To achieve the above purpose, the present application provides the following technical scheme: a visual communication system based on weak current engineering and a running method thereof, the system comprises:

[0007] A signal acquisition module synchronously acquires video, audio and infrared signals of a target area by using a multi-modal sensor unit, acquires environmental light and electromagnetic parameters by an environmental parameter detection unit, and outputs standardized acquisition data by a signal preprocessing unit;

[0008] A data processing module receives the standardized acquisition data, analyzes multi-modal signal features by a feature extraction unit, performs time-space alignment processing by a data fusion unit, and outputs fusion data by system performance evaluation;

[0009] The communication transmission module receives the fusion data, distributes the data through a double-channel transmission network composed of a wired transmission unit and a wireless transmission unit, and sets three working modes of a standard mode, an anti-interference mode and an emergency mode. In the standard mode, when it is detected that the electromagnetic interference intensity is lower than the electromagnetic interference intensity threshold value and the transmission efficiency is greater than 0.75, wireless transmission is performed in the 5GHz frequency band. In the anti-interference mode, when it is detected that the electromagnetic interference intensity exceeds the electromagnetic interference intensity threshold value and the transmission efficiency is less than 0.75, the wired transmission is automatically switched to and channel encryption is enabled. In the emergency mode, when the main transmission link is interrupted, multi-hop relay transmission is enabled through a Mesh network. The transmission efficiency is calculated by using a normalization formula:

[0010] ;

[0011] wherein, is the transmission efficiency, the transmission efficiency threshold value is 0.75, the standard mode is enabled when the transmission efficiency is greater than 0.75, and the anti-interference mode is switched to when the electromagnetic interference intensity exceeds the electromagnetic interference intensity threshold value and the transmission efficiency is less than 0.75, is a bit error rate correction term, is a measured bit error rate, is a bandwidth utilization rate, is a throughput, is a bandwidth, in Mbps, is a protocol overhead time ratio, and the default is =0.1;

[0012] The visual presentation module receives transmission state data, realizes information partition presentation through a multi-screen display unit, responds to user operation instructions through an interactive control unit, and dynamically adjusts display parameters through a layout optimization unit.

[0013] The system control module coordinates the running processes of the modules, realizes task scheduling through a central control unit, monitors the system health status through a self-checking and maintenance unit, and outputs control instructions through an exception handling unit.

[0014] Preferably, the multi-modal sensor unit comprises:

[0015] A high-definition camera array is arranged at a key position of the target area, and adopts H.265 encoding format to collect video streams with a resolution of 1080P or above in real time;

[0016] A directional microphone array adopts beamforming technology to realize audio collection in a specific direction and has an echo cancellation function;

[0017] An infrared thermal imager has a thermal imaging capability with a temperature sensitivity of 0.05℃, and generates a thermal map with 256×256 pixels;

[0018] The system performance evaluation adopts a normalization formula:

[0019] ;

[0020] wherein, is a normalized quality coefficient, the threshold value ≥0.6 is determined to be qualified, is a normalized signal-to-noise ratio, is a signal-to-noise ratio measured value, in dB, is a normalized bandwidth, is a bandwidth measured value, in MHz, is an environmental correction factor, by default =1.0, when the temperature and humidity are out of standard, it is calculated according to =0.9.

[0021] Preferably, the signal preprocessing unit performs operations including:

[0022] Performing adaptive filtering processing based on space-time domain on the video signal to eliminate motion blur and light sudden change effects;

[0023] Performing spectrum analysis and background noise suppression on the audio signal to retain the 200Hz-8kHz frequency band;

[0024] Performing non-uniformity correction and temperature calibration on the infrared signal to eliminate the effects of sensor self-thermal noise.

[0025] Preferably, the visualization presentation module further includes:

[0026] An augmented reality superimposition unit for superimposing device state information in AR form onto a real-time picture;

[0027] A multi-view synchronization unit for ensuring that the content of different display terminals maintains consistent time stamps;

[0028] An intelligent layout unit for automatically adjusting information arrangement priority and display proportion according to user roles.

[0029] Preferably, the method includes the following steps:

[0030] S1, synchronously collecting video, audio and infrared signals of a target region through a multi-modal sensor array, while detecting environmental light and electromagnetic parameters;

[0031] S2, performing time domain alignment and format standardization processing on the collected original signals to eliminate time deviations of different sensors;

[0032] S3, extracting feature vectors of each modal signal, including motion features in the video, voiceprint features in the audio and temperature distribution features in the infrared signal;

[0033] S4, fusing the multi-modal features based on a deep neural network to generate a comprehensive feature representation with spatio-temporal consistency;

[0034] S5, dynamically calculating transmission efficiency according to current network conditions , the transmission efficiency threshold is 0.75, when the transmission efficiency is greater than 0.75 and the electromagnetic interference intensity is lower than the electromagnetic interference intensity threshold, the standard mode is enabled, when the electromagnetic interference intensity exceeds the electromagnetic interference intensity threshold and less than 0.75, switch to the anti-interference mode, when the main link is interrupted, the emergency mode is enabled, the standard mode, when the electromagnetic interference intensity is detected to be lower than the electromagnetic interference intensity threshold and the transmission efficiency is greater than 0.75, wireless transmission is carried out in the 5GHz frequency band, the anti-interference mode, when the electromagnetic interference intensity is detected to be greater than the electromagnetic interference intensity threshold and the transmission efficiency is less than 0.75, automatically switch to wired transmission and enable channel encryption, the emergency mode, when the main transmission link is interrupted, enable Mesh network for multi-hop relay transmission;

[0035] wherein the transmission efficiency calculation uses a normalization formula:

[0036] ;

[0037] wherein, is the transmission efficiency, the transmission efficiency threshold is 0.75, when the transmission efficiency is greater than 0.75, the standard mode is enabled, when the electromagnetic interference intensity exceeds the electromagnetic interference intensity threshold and is the error rate correction term, is the measured error rate, is the bandwidth utilization rate, is the throughput, is the bandwidth, unit: Mbps, is the protocol overhead time ratio, default =0.1;

[0038] S6, intelligently layout multi-modal information in the visualization interface, support user interactive operation and display parameter adjustment;

[0039] S7, real-time monitoring of the running state of each component of the system, when an exception is detected, automatically start the fault tolerance mechanism.

[0040] Preferably, the S1 synchronously collects video, audio and infrared signals of the target area through a multi-modal sensor array, and simultaneously detects environmental light and electromagnetic parameters.

[0041] Synchronously trigger all kinds of sensors according to the preset sampling frequency to ensure the time consistency of data collection;

[0042] The camera gain and infrared instrument sensitivity are automatically adjusted according to the intensity of ambient light;

[0043] The original data packet with a timestamp is established to record the collection parameters and state information of each sensor.

[0044] Preferably, the process of the S3 extracting the feature vectors of each modal signal, including the motion features in the video, the voiceprint features in the audio and the temperature distribution features in the infrared signal, comprises:

[0045] Video feature extraction: 3D convolutional neural network is used to extract spatio-temporal features from video sequences, including motion trajectories and object shapes;

[0046] Audio feature extraction: Mel frequency cepstral coefficient is used to analyze speech content, and abnormal sound is detected at the same time;

[0047] Infrared signal feature extraction: regional growth algorithm is used to identify temperature abnormal areas, and the thermal radiation distribution gradient is calculated.

[0048] Preferably, the process of the S4 based on a deep neural network to fuse multi-modal features to generate a comprehensive feature representation with spatio-temporal consistency comprises:

[0049] A multi-modal feature correlation matrix is established to calculate the similarity weight between different modal features;

[0050] The fusion proportion of each modal feature is dynamically adjusted through an attention mechanism;

[0051] A unified time-space-feature three-dimensional tensor representation is generated to preserve the hierarchical relationship of the original signal.

[0052] Preferably, the fault-tolerant mechanism in the step S7 comprises:

[0053] Sensor failure processing: when the multi-modal sensor fails, a virtual signal generation based on a generative adversarial network is enabled;

[0054] Network interruption processing: during communication interruption, key data is locally cached, and non-key data is compressed and stored;

[0055] Display abnormality processing: when the main display unit fails, the content is automatically migrated to the backup display terminal.

[0056] Compared with the prior art, the present application provides a visual communication system based on weak current engineering and its operation method, which has the following beneficial effects:

[0057] 1. In the present application, by synchronously collecting video, audio and infrared signals, and detecting environmental electromagnetic parameters in real time, the integrity and time synchronization of different signal sources are ensured, and the adaptive filtering technology is used to process the original signals, which can reduce the signal distortion caused by motion blur and electromagnetic interference, ensure the collection quality of multi-modal data, and further reduce the error rate of the communication system.

[0058] 2. In the present application, when performing real-time data transmission in a weak current engineering environment, the wired and wireless transmission modes are dynamically switched according to the electromagnetic interference intensity, the channel quality is evaluated in real time, and the optimal transmission strategy is selected, so that the system can maintain a stable communication link in a strong interference environment, and automatically enable the Mesh network for multi-hop relay when the transmission link is interrupted, ensuring the continuity and reliability of the communication process.

[0059] 3. In the present application, when performing multi-source information comprehensive display, the information layout priority is automatically adjusted according to the user role and operation demand, and the device state information is superimposed using augmented reality technology, so that the system can display key monitoring data, reduce the cognitive load of the operator, and further improve the human-computer interaction efficiency and emergency response speed. BRIEF DESCRIPTION OF DRAWINGS

[0060] Fig. 1 is a schematic diagram of the framework of a visual communication system based on weak current engineering of the present application;

[0061] Fig. 2 is a step flow chart of a visual communication running method based on weak current engineering of the present application. DETAILED DESCRIPTION

[0062] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the accompanying drawings of the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0063] Please refer to Figs. 1-2 The visual communication system based on weak current engineering and the running method thereof, the system comprises:

[0064] The signal acquisition module synchronously acquires video, audio and infrared signals of the target area using a multi-modal sensor unit, obtains environmental light and electromagnetic parameters through an environmental parameter detection unit, and outputs standardized acquisition data from a signal preprocessing unit;

[0065] A data processing module receives standardized acquisition data, analyzes multi-modal signal features through a feature extraction unit, performs spatio-temporal alignment processing using a data fusion unit, and outputs fused data through system performance evaluation;

[0066] A communication transmission module receives fused data, performs data distribution through a dual-channel transmission network composed of a wired transmission unit and a wireless transmission unit, and sets three working modes of a standard mode, an anti-interference mode, and an emergency mode. In the standard mode, when the electromagnetic interference intensity is detected to be lower than the electromagnetic interference intensity threshold and the transmission efficiency is greater than 0.75, wireless transmission is performed using a 5 GHz frequency band. In the anti-interference mode, when the electromagnetic interference intensity is detected to be higher than the electromagnetic interference intensity threshold and the transmission efficiency is less than 0.75, automatic switching to wired transmission and enabling channel encryption are performed. In the emergency mode, when the main transmission link is interrupted, multi-hop relay transmission is enabled using a Mesh network. The transmission efficiency is calculated using a normalization formula:

[0067] ;

[0068] wherein, is the transmission efficiency, the transmission efficiency threshold is 0.75, the standard mode is enabled when the transmission efficiency is greater than 0.75, and the anti-interference mode is switched to when the electromagnetic interference intensity exceeds the electromagnetic interference intensity threshold and the transmission efficiency is less than 0.75, is a bit error rate correction term, is the measured bit error rate, is the bandwidth utilization rate, is the throughput, is the bandwidth, with a unit of Mbps, is the protocol overhead time ratio, by default = 0.1;

[0069] A visual presentation module receives transmission state data, realizes information partition presentation through a multi-screen display unit, responds to user operation instructions using an interactive control unit, and dynamically adjusts display parameters through a layout optimization unit.

[0070] A system control module coordinates the running processes of each module, realizes task scheduling through a central control unit, monitors the system health status using a self-checking and maintenance unit, and outputs control instructions through an exception handling unit.

[0071] The multi-modal sensor unit includes:

[0072] A high-definition camera array is arranged at key positions of the target area and uses H.265 encoding format to collect video streams with a resolution of 1080P or above in real time;

[0073] A directional microphone array uses beamforming technology to realize audio collection in a specific direction and has an echo cancellation function.

[0074] Infrared thermal imager, with thermal imaging capability of temperature sensitivity 0.05℃, generating a 256x256 pixel thermal map;

[0075] Wherein the system performance evaluation adopts the normalization formula:

[0076] ;

[0077] Wherein, is the normalized quality coefficient, the threshold ≥0.6 is determined to be qualified, is the normalized signal-to-noise ratio, is the measured value of the signal-to-noise ratio, in dB, is the normalized bandwidth, is the measured value of the bandwidth, in MHz, is the environmental correction factor, default =1.0, when the temperature and humidity are out of standard, it is calculated according to =0.9;

[0078] The operations performed by the signal preprocessing unit include:

[0079] Adaptive filtering processing based on space-time domain is performed on the video signal to eliminate motion blur and light sudden change effects;

[0080] Spectrum analysis and background noise suppression are performed on the audio signal to retain the 200Hz-8kHz frequency band;

[0081] Non-uniformity correction and temperature calibration are performed on the infrared signal to eliminate the effects of sensor self-thermal noise;

[0082] Wherein the signal processing efficiency evaluation adopts the normalization formula:

[0083] ;

[0084] Wherein, is the normalized fidelity, the threshold ≥0.92 is determined to be qualified, is the normalized processed video frame, the pixel value is scaled to [0,1], is the normalized original video frame, the pixel value is scaled to [0,1], is the L2 norm calculation;

[0085] The visualization presentation module further includes:

[0086] An augmented reality superposition unit for superimposing device status information in AR form on the real-time picture;

[0087] A multi-view synchronization unit to ensure that the content of different display terminals maintains the same timestamp.

[0088] The intelligent layout unit automatically adjusts the information arrangement priority and display ratio according to the user role;

[0089] The display quality index adopts a normalization model:

[0090]

[0091] wherein, is a normalized quality index, and the threshold value ≥3.2 is excellent, is the normalized weight of the kth parameter, =1, is the normalized measurement value of the kth parameter, is a normalized response delay, in seconds, is a normalized resolution, in megapixels;

[0092] The method comprises the following steps:

[0093] S1, synchronously collecting video, audio and infrared signals of a target area through a multi-modal sensor array, while detecting environmental illumination and electromagnetic parameters;

[0094] S2, performing time domain alignment and format standardization processing on the collected original signals to eliminate time deviations of different sensors;

[0095] S3, extracting feature vectors of each modal signal, including motion features in the video, voiceprint features in the audio and temperature distribution features in the infrared signal;

[0096] S4, fusing multi-modal features based on a deep neural network to generate a comprehensive feature representation with spatiotemporal consistency;

[0097] S5, dynamically calculating a transmission efficiency , the transmission efficiency threshold is 0.75, when the transmission efficiency is greater than 0.75 and the electromagnetic interference intensity is lower than the electromagnetic interference intensity threshold, the standard mode is enabled, when the electromagnetic interference intensity exceeds the electromagnetic interference intensity threshold and is less than 0.75, the anti-interference mode is switched to, when the main link is interrupted, the emergency mode is enabled, the standard mode, when it is detected that the electromagnetic interference intensity is lower than the electromagnetic interference intensity threshold and the transmission efficiency is greater than 0.75, wireless transmission is performed in the 5GHz frequency band, the anti-interference mode, when it is detected that the electromagnetic interference intensity exceeds the electromagnetic interference intensity threshold and the transmission efficiency is less than 0.75, automatic switching to wired transmission and enabling channel encryption, the emergency mode, when the main transmission link is interrupted, enabling Mesh network for multi-hop relay transmission; ​

[0098] wherein the transmission efficiency calculation adopts a normalization formula:

[0099] ;

[0100] wherein, is the transmission efficiency, the transmission efficiency threshold is 0.75, the standard mode is enabled when the transmission efficiency is greater than 0.75, and the anti-interference mode is switched to when the electromagnetic interference intensity exceeds the electromagnetic interference intensity threshold and the transmission efficiency is less than 0.75, is a bit error rate correction term, is the measured bit error rate, is the bandwidth utilization rate, is the throughput, is the bandwidth, in Mbps, is the protocol overhead time ratio, by default =0.1;

[0101] S6, intelligently layout multi-modal information in the visualization interface, support user interactive operation and display parameter adjustment;

[0102] S7, real-time monitoring of the running state of each component of the system, automatically starting the fault tolerance mechanism when an exception is detected;

[0103] wherein the feature fusion efficiency normalization model is:

[0104] ;

[0105] wherein, ∈(0,1) is the normalized fusion efficiency, the threshold ≥0.88, is the normalized weight of the mth mode, , is the normalized confidence of the mth mode;

[0106] S1, the process of synchronously collecting video, audio and infrared signals of the target area through a multi-modal sensor array, while detecting environmental light and electromagnetic parameters, includes:

[0107] Synchronously trigger various sensors according to the preset sampling frequency to ensure the time consistency of data collection;

[0108] Automatically adjust the camera gain and infrared instrument sensitivity according to the environmental light intensity;

[0109] Establish a raw data packet with a timestamp to record the collection parameters and state information of each sensor;

[0110] S3, the process of extracting feature vectors of each modal signal, including motion features in video, voiceprint features in audio and temperature distribution features in infrared signals, includes:

[0111] Video feature extraction: 3D convolutional neural network is used to extract spatio-temporal features from video sequences, including motion trajectories and object shapes;

[0112] Audio feature extraction: Mel-frequency cepstral coefficient is used to analyze speech content, while abnormal sound is detected;

[0113] Infrared signal feature extraction: region growing algorithm is used to identify temperature anomaly regions, and thermal radiation distribution gradient is calculated;

[0114] where the normalized feature completeness index is:

[0115] ;

[0116] where, is the normalized completeness, the threshold ≥ 0.95, is the normalized weight of the i-th feature, is the normalized mutual information;

[0117] S4. The process of fusing multi-modal features based on deep neural network to generate comprehensive feature representation with spatio-temporal consistency includes:

[0118] Establish a multi-modal feature correlation matrix to calculate the similarity weight between different modal features;

[0119] Adjust the fusion ratio of each modal feature dynamically through attention mechanism;

[0120] Generate a unified time-space-feature three-dimensional tensor representation to preserve the hierarchical relationship of the original signal;

[0121] The fault-tolerant mechanism in step S7 includes:

[0122] Sensor failure handling: when multi-modal sensors fail, enable virtual signal generation based on generative adversarial network;

[0123] Network interruption handling: locally cache critical data during communication interruption, and compress non-critical data;

[0124] Display exception handling: when the main display unit fails, automatically migrate the content to the backup display terminal;

[0125] where the fault-tolerant performance normalization model is:

[0126] ;

[0127] where, is the normalized fault-tolerant performance, the threshold ≥ 0.94, normalized recovery time, in seconds, normalized recovery success rate of the jth fault, normalized complexity factor, normalized redundant resource proportion.

[0128] The operation steps of a visual communication system based on weak current engineering and its operation method are as follows:

[0129] Step one: environmental perception and data acquisition

[0130] After the system is started, each sensor node synchronously acquires environmental data according to a preset sampling frequency, a high-definition camera captures a video stream of a monitoring area in real time, a directional microphone array focuses on a specific sound source direction by using a beamforming technology, an infrared thermal imager continuously scans a temperature distribution, and meanwhile, an environmental detection unit measures illumination intensity, temperature and humidity, and electromagnetic interference parameters in real time, thereby providing an environmental benchmark for subsequent signal processing.

[0131] Step two: multi-source signal preprocessing

[0132] The acquired original data are first subjected to time domain alignment processing, the millisecond-level time deviation between different sensors is eliminated through a hardware timestamp and a software interpolation algorithm, a video signal is subjected to time-space domain adaptive filtering to reduce picture blur caused by device movement and illumination mutation, an audio signal is subjected to spectrum analysis and noise suppression to retain a human voice frequency band, and infrared signal data are subjected to non-uniformity correction and temperature calibration to ensure the accuracy of a thermal map.

[0133] Step three: intelligent feature extraction

[0134] The preprocessed data are sent to a feature extraction engine. A video analysis module extracts motion trajectory and abnormal behavior features from a continuous frame sequence by using a 3D convolutional neural network, an audio processing module identifies specific voice prints and abnormal sound by using a mel-frequency cepstral coefficient analysis, and infrared signals mark temperature abnormal areas and calculate thermal radiation gradients based on a region growing algorithm, and the feature vectors output by each module are all attached with confidence scores.

[0135] Step four: dynamic data fusion

[0136] After the feature data enter a fusion center, the system first evaluates channel quality under the current environment, wireless transmission mode is enabled when electromagnetic interference is low, and wired channels are automatically switched to when transmission efficiency exceeds a threshold, a fusion algorithm dynamically adjusts weights according to the confidence of each modal feature, time-space alignment is realized through an attention mechanism, and finally, a comprehensive situation representation containing multi-dimensional information such as video, audio, and temperature is generated.

[0137] Step five: adaptive visual presentation

[0138] The fused data is intelligently laid out according to user roles and task requirements, and key monitoring information is superimposed on the real-time picture in the form of augmented reality. Multiple display screens keep the content updated synchronously, and the system continuously monitors the operator's interaction behavior. When it detects that a certain type of data is frequently accessed, it automatically promotes the display priority and detail level of that information.

[0139] Step six: autonomous fault tolerance and optimization

[0140] The system diagnoses the status of each component in real time through the health monitoring module. When a sensor fails, it immediately enables virtual signal generation based on a generative adversarial network. When the network is interrupted, it starts the local cache and compression transmission mechanism. When an anomaly is displayed, it automatically migrates the content to a backup terminal. At the same time, the system records performance indicators during operation and continuously optimizes parameter configurations through machine learning algorithms.

[0141] It should be noted that, in this text, relational terms such as first and second are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "contain" or any other variant thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed or inherent to such process, method, article or device. Without further limitation, the element defined by the statement "includes a" does not exclude the presence of another identical element in the process, method, article or device that includes the element.

[0142] Although embodiments of the present application have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and alterations can be made thereto without departing from the principles and spirit of the present application, and the scope of the present application is defined by the appended claims and their equivalents.

Claims

1. A visual communication system based on low voltage engineering, characterized by: The method comprises the following steps: S1, synchronously collecting video, audio and infrared signals of a target area by a multi-modal sensor array, and detecting environmental light and electromagnetic parameters; S2, performing time domain alignment and format standardization processing on the collected original signals to eliminate time deviation of different sensors; S3, receiving the standardized collection data, analyzing multi-modal signal characteristics by a feature extraction unit, performing space-time alignment processing by a data fusion unit, and outputting fusion data by system performance evaluation; ； Wherein, For transmission efficiency, the transmission efficiency threshold is 0.75, the standard mode is enabled when the transmission efficiency is greater than 0.75, and the anti-interference mode is switched to when the electromagnetic interference intensity exceeds the electromagnetic interference intensity threshold and the transmission efficiency is less than 0.75, For the error code rate correction term, For the measured error code rate, For bandwidth utilization, For throughput, For bandwidth, unit: Mbps, For protocol overhead time ratio, default =0.1; S4, receiving the fusion data, performing data distribution by a double-channel transmission network composed of a wired transmission unit and a wireless transmission unit, and setting three working modes of a standard mode, an anti-interference mode and an emergency mode; S5, receiving transmission state data, realizing information partition presentation by a multi-screen display unit, responding to user operation instructions by an interactive control unit, and dynamically adjusting display parameters by a layout optimization unit; 2. A visual communication system based on low current engineering according to claim 1, characterized in that: S6, coordinating module operation processes, realizing task scheduling by a central control unit, monitoring system health status by a self-checking and maintenance unit, and outputting control instructions by an exception handling unit. The multi-modal sensor unit comprises: a high-definition camera array arranged at key positions of the target area, which adopts H.265 encoding format to collect video streams with a resolution of 1080P or above in real time; a directional microphone array which adopts beamforming technology to realize audio collection in a specific direction and has echo cancellation function; an infrared thermal imager which has a thermal imaging capability with a temperature sensitivity of 0.05℃ and generates a thermal map with 256x256 pixels; ； wherein, is a normalized quality coefficient, threshold ≥ 0.6 is determined to be qualified, is a normalized signal-to-noise ratio, is a signal-to-noise ratio measured value, unit: dB, is a normalized bandwidth, is a bandwidth measured value, unit: MHz, is an environmental correction factor, default = 1.0, when the temperature and humidity are out of standard, according to = 0.9 is calculated.

3. The visual communication system based on low current engineering according to claim 1, characterized in that: wherein the system performance evaluation adopts a normalization formula: The signal preprocessing unit performs operations including: performing adaptive filtering processing on video signals based on space-time domain to eliminate motion blur and light sudden change effects; performing spectrum analysis and background noise suppression on audio signals to retain a frequency band of 200Hz-8kHz; 4. The visual communication system based on low current engineering according to claim 1, characterized in that: performing non-uniformity correction and temperature calibration on infrared signals to eliminate the influence of sensor self-thermal noise. The visualization presentation module further comprises: an augmented reality superposition unit for superimposing device state information in AR form on a real-time picture; a multi-view synchronization unit for ensuring that the content of different display terminals maintains the same timestamp; 5. A visual communication operation method based on weak current engineering, characterized in that: an intelligent layout unit for automatically adjusting information arrangement priority and display proportion according to user roles. The method comprises the following steps: S1, synchronously collecting video, audio and infrared signals of a target area by a multi-modal sensor array, and detecting environmental light and electromagnetic parameters; S2, performing time domain alignment and format standardization processing on the collected original signals to eliminate time deviation of different sensors; S3, extract the feature vector of each modal signal, including the motion feature in the video, the voiceprint feature in the audio and the temperature distribution feature in the infrared signal; S4, based on the deep neural network, the multi-modal features are fused to generate a comprehensive feature representation with spatio-temporal consistency; S5, dynamically calculating transmission efficiency according to current network status , the transmission efficiency threshold is 0.75, the standard mode is enabled when the transmission efficiency is greater than 0.75 and the electromagnetic interference intensity is lower than the electromagnetic interference intensity threshold, the anti-interference mode is switched to when the electromagnetic interference intensity exceeds the electromagnetic interference intensity threshold and the transmission efficiency is less than 0.75, the emergency mode is enabled when the main link is interrupted, the standard mode, when it is detected that the electromagnetic interference intensity is lower than the electromagnetic interference intensity threshold and the transmission efficiency is greater than 0.75, wireless transmission is performed in the 5GHz frequency band, the anti-interference mode, when it is detected that the electromagnetic interference intensity exceeds the electromagnetic interference intensity threshold and the transmission efficiency is less than 0.75, automatic switching to wired transmission and enabling channel encryption, the emergency mode, when the main transmission link is interrupted, enabling Mesh network for multi-hop relay transmission; The transmission efficiency calculation adopts a normalization formula: ； Wherein, For transmission efficiency, the transmission efficiency threshold is 0.75, the standard mode is enabled when the transmission efficiency is greater than 0.75, and the anti-interference mode is switched to when the electromagnetic interference intensity exceeds the electromagnetic interference intensity threshold and the transmission efficiency is less than 0.75, For the error code rate correction term, For the measured error code rate, For bandwidth utilization, For throughput, For bandwidth, unit: Mbps, For protocol overhead time ratio, default =0.1; S6, intelligently layout multi-modal information in the visualization interface, support user interactive operation and display parameter adjustment; S7, real-time monitoring of the running state of each component of the system, when an abnormality is detected, automatically start the fault-tolerant mechanism.

6. The visual communication operation method based on the weak current engineering according to claim 5, characterized in that: The S1 includes the process of synchronously collecting video, audio and infrared signals of the target area by a multi-modal sensor array, while detecting environmental light and electromagnetic parameters, which includes: Synchronously trigger various sensors according to the preset sampling frequency to ensure the time consistency of data collection; Automatically adjust the camera gain and infrared instrument sensitivity according to the environmental light intensity; Establish a timestamped raw data packet to record the collection parameters and state information of each sensor.

7. The visual communication operation method based on the weak current engineering according to claim 5, characterized in that: The S3 includes the process of extracting the feature vector of each modal signal, including the motion feature in the video, the voiceprint feature in the audio and the temperature distribution feature in the infrared signal, which includes: Video feature extraction: 3D convolutional neural network is used to extract spatio-temporal features from video sequences, including motion trajectory and object shape; Audio feature extraction: Mel frequency cepstral coefficient is used to analyze speech content while detecting abnormal sound; Infrared signal feature extraction: based on region growing algorithm, temperature abnormal area is identified and heat radiation distribution gradient is calculated.

8. The visual communication operation method based on weak current engineering according to claim 5, characterized in that: The S4 includes the process of fusing multi-modal features based on deep neural network to generate a comprehensive feature representation with spatio-temporal consistency, which includes: Establish a multi-modal feature correlation matrix to calculate the similarity weight between different modal features; Through attention mechanism, dynamically adjust the fusion ratio of each modal feature; Generate a unified time-space-feature three-dimensional tensor representation to preserve the hierarchical relationship of the original signal.

9. The visual communication operation method based on the weak current engineering according to claim 5, characterized in that: The fault-tolerant mechanism in S7 includes: Sensor failure handling: when the multi-modal sensor fails, enable the virtual signal generation based on the generative adversarial network; Network interruption handling: during communication interruption, cache key data locally and compress non-key data; Display abnormality handling: when the main display unit fails, automatically migrate the content to the backup display terminal.

Images