A method for monitoring gas density changes using an event camera

Abstract

The present invention relates to the technical field of gas density change monitoring, and specifically, to a method for monitoring gas density changes using an event camera. The method comprises the following steps: setting a light source and installing a background plate; setting a trigger threshold of the event camera; combining backlight schlieren technology to capture the refractive index change caused by hydrogen leakage, combining the environmental factors of temperature, humidity and air pressure, and using a multi-dimensional environment to correct the event rate change rate algorithm, and calculating the event generation rate change rate; monitoring the event generation rate change rate threshold, and automatically triggering an alarm signal when the event generation rate change rate exceeds the threshold. The method for monitoring gas density changes using an event camera can achieve accurate detection and effective control of gas leaks in low-light and complex environments by combining the high-resolution characteristics of the event camera and the backlight schlieren technology, as well as using a multi-dimensional environment correction algorithm and a frequency domain analysis method.

Description

Technical Field

[0001] The present invention relates to the technical field of gas density change monitoring, and in particular to a method for monitoring gas density change using an event camera. Background Art

[0002] A method for monitoring gas density changes using an event camera aims to accurately capture tiny refractive index changes during gas leaks and improve detection accuracy in complex environments. By combining backlight schlieren technology with a multi-dimensional environment-corrected event rate change rate algorithm, the interference of environmental factors is controlled to achieve low-power, high-sensitivity real-time hydrogen leak monitoring.

[0003] The existing methods for monitoring hydrogen density changes mainly use two methods: sensor detection and infrared thermal imager detection. The implementation premise of sensor detection requires that the hydrogen concentration diffuses to the vicinity of the sensor before the sensor can detect and alarm. Since the density of hydrogen is extremely low and it hardly diffuses in a static environment, the sensor is often unable to be accurately arranged in the hydrogen leakage path, resulting in the inability to detect hydrogen leakage in time, a very long response time, and poor adaptability to environmental changes. Therefore, a method for monitoring gas density changes using an event camera is provided. Summary of the invention

[0004] The purpose of the present invention is to provide a method for monitoring gas density changes using an event camera, so as to solve the problem raised in the above background technology that since hydrogen density is extremely low and it hardly diffuses in a static environment, the sensor is often unable to be accurately arranged in the hydrogen leakage path, resulting in the inability to detect hydrogen leakage in time, very long response time and poor adaptability to environmental changes.

[0005] To achieve the above object, the present invention provides a method for monitoring gas density changes using an event camera, comprising the following steps:

[0006] S1. Set a light source at one end of the monitored area and install a background plate opposite the light source to capture the changes in the schlieren after the light penetrates the monitored area;

[0007] S2, fix the event camera and adjust the focal length of the event camera according to the distance between the event camera and the background plate, and set the trigger threshold of the event camera;

[0008] S3. Use event cameras for real-time monitoring, combine backlight schlieren technology to capture the refractive index changes caused by hydrogen leakage, generate event streams related to brightness changes, combine the environmental factors of temperature, humidity and air pressure, use multi-dimensional environment to correct the event rate change rate algorithm, and calculate the event generation rate change rate;

[0009] S4. Set and monitor the event generation rate change rate threshold. When the event generation rate change rate exceeds the threshold, an alarm signal is automatically triggered to indicate that a hydrogen leak is detected.

[0010] As a further improvement of the technical solution, in S1, the light source is selected as a highly focused LED light source, which is arranged at one end of the monitored area, and the light intensity of the highly focused LED light source is adjusted;

[0011] The background was chosen to be a black acrylic board, which was installed directly opposite the highly focused LED light source, and a sunshade was installed around the background board.

[0012] As a further improvement of the technical solution, in S2, setting the trigger threshold of the event camera refers to the minimum value of the brightness change in the camera pixel. When the change exceeds this threshold, the camera will generate an event. The specific steps of setting the trigger threshold of the event camera are as follows:

[0013] S2.1. Use the default values ​​of the event camera as a starting point and perform preliminary threshold settings based on the light intensity and changes in the monitoring environment.

[0014] S2.2. Under the condition that the noise in the environment can be eliminated, gradually adjust the trigger threshold until the event camera can sensitively capture the light and shadow changes caused by the gas leakage, and determine the trigger threshold;

[0015] S2.3. Simulate gas leakage of different degrees in the monitoring area to observe whether the event camera can capture the changes in the schlieren on the background plate and generate events. Based on the test results, further optimize the camera position, focal length and threshold settings.

[0016] As a further improvement of the technical solution, in S3, the multi-dimensional environment correction event rate change rate algorithm is implemented based on multi-scale analysis and frequency domain analysis, and the environmental correction coefficient matrix is ​​used to monitor hydrogen leakage in real time and calculate the event rate change rate during hydrogen leakage according to the dynamic changes of temperature, humidity and air pressure environmental parameters;

[0017] In S31, the event generation rate change rate is calculated by combining the environmental factors of temperature, humidity and air pressure and using a multi-dimensional environment correction event rate change rate algorithm. The specific method is as follows:

[0018] S3.1. Real-time collection of environmental parameters in the monitoring area , including temperature ,humidity and air pressure , and determine the temperature, humidity and pressure reference values ​​of environmental parameters under no-leak conditions , ;

[0019] ;

[0020] S3.2 Environmental parameters Perform wavelet transform to obtain features at different scales;

[0021] S3.3 Environmental parameters Perform discrete Fourier transform and calculate the frequency domain representation , identify the main frequency components in the environmental parameters through frequency domain analysis, divide the frequency components into different frequency intervals based on the discrete Fourier transform results, and assign different weights to different frequency intervals , and finally obtain the weighted frequency domain representation ;

[0022] S3.4. Define the correction coefficient matrix To characterize the effects of temperature, humidity and air pressure on the rate at which events occur;

[0023] S3.5. Correction coefficient matrix The correction coefficients in are nonlinearly combined, and the correction coefficient matrix Perform point-by-point integration correction;

[0024] S3.6. Combined correction coefficient matrix And the weighted frequency domain representation Calculate the rate of change of complex event generation rate .

[0025] As a further improvement of this technical solution, in S3.2, the environmental parameters Perform wavelet transform to obtain multi-scale features. The specific method is as follows:

[0026] ;

[0027] in, is the result of wavelet transform, which means that and scale Multi-scale features on is the time parameter; is the scale parameter, which is used to control the degree of expansion and contraction of the wavelet function; is the integration variable; It is the complex conjugate of the mother wavelet function and is used to construct the wavelet transform; are the translation and scaling parameters of the mother wavelet function.

[0028] As a further improvement of this technical solution, in S3.3, the environmental parameters Perform discrete Fourier transform and calculate the frequency domain representation , identify the main frequency components in the environmental parameters through frequency domain analysis, divide the frequency components into different frequency intervals based on the discrete Fourier transform results, and assign different weights to different frequency intervals , and finally obtain the weighted frequency domain representation , the specific method is as follows:

[0029] S3.3.1. Using Discrete Fourier Transform to Calculate Environmental Parameters Frequency domain representation of ;

[0030] S3.3.2. Calculate the power spectrum after discrete Fourier transform , identify the main frequency components in the frequency domain and divide the frequency components into three intervals: low frequency, medium frequency and high frequency;

[0031] S3.3.3. Assign different frequency weights to different frequency intervals ;

[0032] S3.3.4. Representing the frequency domain Multiply by the corresponding frequency weight , and get the weighted frequency domain representation .

[0033] As a further improvement of the technical solution, in S3.3.1, discrete Fourier transform is used to calculate the environmental parameters Frequency domain representation of , the specific method is as follows:

[0034] ;

[0035] in, Environmental parameters At the point in time The value at It is the result of discrete Fourier transform, which indicates that the signal is in frequency The frequency domain component at ; is the frequency index, indicating the frequency components; is the time index, the range is ; is the total number of sampling points, indicating environmental parameters The number of samples after discretization; is the complex exponential form of the Fourier basis function;

[0036] In S3.3.3, different frequency weights are assigned to different frequency intervals. , the specific weight distribution criteria are as follows:

[0037] Low-frequency components: usually reflect long-term trends and stable changes in the environment and are given higher weights;

[0038] Medium-frequency components: reflect medium-term changes and cyclical fluctuations, and the weights are adjusted appropriately according to the actual application scenarios;

[0039] High-frequency components: usually correspond to short-term fluctuations and noise, and are given lower weights;

[0040] In S3.3.4, the frequency domain is represented Multiply by the corresponding frequency weight , and get the weighted frequency domain representation , the specific method is as follows:

[0041] ;

[0042] in, To calculate environmental parameters using discrete Fourier transform Frequency domain representation of ; For frequency The corresponding weight coefficient; is the weighted frequency domain representation, indicating that the signal is at frequency The frequency domain components at .

[0043] As a further improvement of the technical solution, in S3.4, the correction coefficient matrix is ​​defined as To characterize the effects of temperature, humidity and air pressure on the rate of event generation, the specific method is as follows:

[0044] S3.4.1. Calculate the current time Deviation of environmental parameters from the reference value :

[0045] ;

[0046] S3.4.2. Define the correction factor matrix To characterize the combined effect of temperature, humidity and air pressure on the rate at which events occur:

[0047] ;

[0048] in, is the identity matrix; is the weight matrix; is a vector containing the deviations of the environmental parameters of temperature, humidity and pressure;

[0049] S3.4.3. Constructing the weight matrix ,according to Each element in the weight matrix is ​​calculated by the weighted sum of :

[0050] ;

[0051] in, and Environmental parameters and In frequency The frequency domain amplitude at ;

[0052] S3.4.4. Finally, the correction coefficient matrix is ​​obtained .

[0053] As a further improvement of the technical solution, in S3.5, the correction coefficient matrix The correction coefficients in are nonlinearly combined, and the correction coefficient matrix Perform point-by-point integral correction. The specific method is as follows:

[0054] ;

[0055] ;

[0056] in, is the correction coefficient matrix Middle Row, No. Elements of a column; is the coefficient of the nonlinear combination, representing the environmental parameters The sensitivity of changes in the rate at which events occur to the event; Environmental parameters In time The deviation of a moment from its reference value; is a nonlinear index, which controls the nonlinearity of the correction coefficient; is the weight matrix Middle Row, No. Elements of a column; is the time when monitoring starts, which is the initial time of integration;

[0057] In S3.6, combined with the correction coefficient matrix And the weighted frequency domain representation Calculate the rate of change of complex event generation rate , the specific method is as follows:

[0058]

[0059] in, The rate change rate of the event generation is the rate change of hydrogen leakage; For the time camera at time The rate at which events are generated; For time interval The change in the rate at which internal events occur; For time The relevant weight coefficient; Environmental parameters The wavelet transform result.

[0060] As a further improvement of the technical solution, in S4, a threshold value of the rate of change of the event generation rate is set and monitored. When the rate of change of the event generation rate exceeds the threshold, the system will automatically trigger an alarm signal to indicate that a hydrogen leak has been detected. The specific operations are as follows:

[0061] S4.1. Calculate the mean of the rate of change of the event generation rate under normal operating conditions and standard deviation , set the threshold based on the mean and standard deviation of the rate of change of event generation rate :

[0062] ;

[0063] in, is the mean of the rate of change of event generation rate; is the standard deviation of the rate of change of event generation rate; is the threshold value of the rate of change of event generation rate; is a constant, and ;

[0064] S4.2. Continuously calculate the current moment in real time The rate of change of event generation rate and with the preset threshold Make comparisons;

[0065] S4.3. When detected The alarm signal is triggered when a hydrogen leak is detected.

[0066] Compared with the prior art, the present invention has the following beneficial effects:

[0067] 1. In the method of monitoring gas density changes using an event camera, based on the combination of the event camera and the backlit schlieren technology, it is possible to keenly capture tiny refractive index changes caused by gas leakage in a low-light and high dynamic range environment, and significantly amplify these changes, thereby improving the early detection capability of gas leakage.

[0068] 2. In the method of monitoring gas density changes using an event camera, the signal noise caused by ambient temperature, humidity and air pressure is effectively filtered through a multi-dimensional environment-corrected event rate change algorithm and a weight matrix based on frequency domain analysis, thereby significantly improving the accuracy and stability of gas leak detection under complex and changeable environmental conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 The figure is a flow chart of the overall method of the present invention. DETAILED DESCRIPTION

[0070] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Example

[0071] See also Figure 1 As shown, this embodiment provides a method for monitoring gas density changes using an event camera, comprising the following steps:

[0072] S1. Set a light source at one end of the monitored area and install a background plate opposite the light source to capture the changes in the schlieren after the light penetrates the monitored area;

[0073] The light source is selected as a highly focused LED light source, which is set at one end of the monitored area, and the light intensity of the highly focused LED light source is adjusted;

[0074] The purpose of selecting a highly focused LED light source is to produce a uniform, strong and directional light beam, which helps to form a clear contrast between light and shadow. The purpose of setting the light source at one end of the monitored area is to ensure that the light beam emitted by the light source can completely cover the monitored area, so that the refractive index change caused by hydrogen leakage in each area can affect the propagation path of the light beam. The light intensity of the highly focused LED light source is adjusted to ensure that the light can form sufficient brightness contrast on the background board after passing through the monitoring area. Appropriately reducing the brightness of the light source can reduce the power consumption of the event camera, while ensuring that the contrast of the background image is sufficient to reflect the refractive index change caused by gas leakage.

[0075] The background is a black acrylic board, which is installed directly opposite the highly focused LED light source, and a light shield is installed around the background board;

[0076] The black acrylic board as the background can form a strong light contrast under high-brightness light sources, which helps to enhance the visual effect brought about by the change of refractive index; a sunshade is installed around the background board to reduce the interference of ambient light;

[0077] Since light propagates in different paths in media with different refractive indices, hydrogen leakage will cause the light and shadow on the background plate to be distorted or the brightness to change. This is called Schlieren change. By analyzing the Schlieren change on the background plate, the system can sensitively detect the occurrence of hydrogen leakage and calculate its change rate.

[0078] S2, fix the event camera and adjust the focal length of the event camera according to the distance between the event camera and the background plate, and set the trigger threshold of the event camera;

[0079] Setting the trigger threshold of the event camera refers to the minimum value of the brightness change in the camera pixel. When the change exceeds this threshold, the camera will generate an event. The specific steps to set the trigger threshold of the event camera are as follows:

[0080] S2.1. Use the default values ​​of the event camera as a starting point and perform preliminary threshold settings based on the light intensity and changes in the monitoring environment.

[0081] S2.2. Under the condition that the noise in the environment can be eliminated, gradually adjust the trigger threshold until the event camera can sensitively capture the light and shadow changes caused by the gas leakage, and determine the trigger threshold;

[0082] S2.3. Simulate gas leakage of different degrees in the monitoring area to observe whether the event camera can capture the schlieren changes on the background plate and generate events. Based on the test results, further optimize the camera position, focal length and threshold settings.

[0083] The setting of the trigger threshold of the event camera directly affects the sensitivity of the camera: the lower the threshold, the easier it is for the camera to capture tiny changes in brightness; but too low a threshold may result in too many noise events; the event camera may capture some noise when no actual event occurs, for example, brightness changes caused by tiny light flickering or temperature fluctuations. By adjusting the camera's trigger threshold and image processing algorithm, these noises can be effectively suppressed and the possibility of false alarms can be reduced.

[0084] S3. Use event cameras for real-time monitoring, combine backlight schlieren technology to capture the refractive index changes caused by hydrogen leakage, generate event streams related to brightness changes, combine the environmental factors of temperature, humidity and air pressure, use multi-dimensional environment to correct the event rate change rate algorithm, and calculate the event generation rate change rate;

[0085] The multi-dimensional environment correction event rate change rate algorithm is based on multi-scale analysis and frequency domain analysis. It monitors hydrogen leakage in real time and calculates the event rate change rate during hydrogen leakage according to the dynamic changes of temperature, humidity and air pressure environmental parameters through the environmental correction coefficient matrix;

[0086] In S31, the event generation rate change rate is calculated by combining the environmental factors of temperature, humidity and air pressure and using a multi-dimensional environment correction event rate change rate algorithm. The specific method is as follows:

[0087] S3.1. Real-time collection of environmental parameters in the monitoring area , including temperature ,humidity and air pressure , and determine the temperature, humidity and pressure reference values ​​of environmental parameters under no-leak conditions , ;

[0088] ;

[0089] S3.2 Environmental parameters Perform wavelet transform to obtain features at different scales;

[0090] S3.3 Environmental parameters Perform discrete Fourier transform and calculate the frequency domain representation , identify the main frequency components in the environmental parameters through frequency domain analysis, divide the frequency components into different frequency intervals based on the discrete Fourier transform results, and assign different weights to different frequency intervals , and finally obtain the weighted frequency domain representation ;

[0091] Among them, the discrete Fourier transform of the environmental parameters is selected The purpose of processing is because the discrete Fourier transform is suitable for processing discrete time series data and can efficiently implement frequency domain analysis in digital systems, which makes it more direct and practical in analyzing the frequency components of sampled data;

[0092] S3.4. Define the correction coefficient matrix To characterize the effects of temperature, humidity and air pressure on the rate at which events occur;

[0093] Here we introduce the correction coefficient matrix The purpose is to quantitatively characterize the impact of environmental parameters such as temperature, humidity and air pressure on the event generation rate, and comprehensively consider the interaction of these parameters in the form of a matrix, so as to dynamically correct and adjust the calculation results of the event generation rate change rate;

[0094] S3.5. Correction coefficient matrix The correction coefficients in are nonlinearly combined, and the correction coefficient matrix Perform point-by-point integration correction;

[0095] S3.6. Combined correction coefficient matrix And the weighted frequency domain representation Calculate the rate of change of complex event generation rate .

[0096] In S3.2, the environmental parameters Perform wavelet transform to obtain multi-scale features. The specific method is as follows:

[0097] ;

[0098] in, is the result of wavelet transform, which means that and scale Multi-scale features on is the time parameter; is the scale parameter, which is used to control the degree of expansion and contraction of the wavelet function; is the integration variable; It is the complex conjugate of the mother wavelet function and is used to construct the wavelet transform; are the translation and scaling parameters of the mother wavelet function.

[0099] In S3.3, the environmental parameters Perform discrete Fourier transform and calculate the frequency domain representation , identify the main frequency components in the environmental parameters through frequency domain analysis, divide the frequency components into different frequency intervals based on the discrete Fourier transform results, and assign different weights to different frequency intervals , and finally obtain the weighted frequency domain representation , the specific method is as follows:

[0100] S3.3.1. Using Discrete Fourier Transform to Calculate Environmental Parameters Frequency domain representation of ;

[0101] S3.3.2. Calculate the power spectrum after discrete Fourier transform , identify the main frequency components in the frequency domain and divide the frequency components into three intervals: low frequency, medium frequency and high frequency;

[0102] S3.3.3. Assign different frequency weights to different frequency intervals ;

[0103] S3.3.4. Representing the frequency domain Multiply by the corresponding frequency weight , and get the weighted frequency domain representation .

[0104] In S3.3.1, use discrete Fourier transform to calculate environmental parameters Frequency domain representation of , the specific method is as follows:

[0105] ;

[0106] in, Environmental parameters At the point in time The value at It is the result of discrete Fourier transform, which indicates that the signal is in frequency The frequency domain component at ; is the frequency index, indicating the frequency components; is the time index, the range is ; is the total number of sampling points, indicating environmental parameters The number of samples after discretization; is the complex exponential form of the Fourier basis function;

[0107] In S3.3.3, different frequency weights are assigned to different frequency intervals. , the specific weight distribution criteria are as follows:

[0108] Low-frequency components: usually reflect long-term trends and stable changes in the environment and are given higher weights;

[0109] Medium-frequency components: reflect medium-term changes and cyclical fluctuations, and the weights are adjusted appropriately according to the actual application scenarios;

[0110] High-frequency components: usually correspond to short-term fluctuations and noise, and are given lower weights;

[0111] In S3.3.4, the frequency domain representation Multiply by the corresponding frequency weight , and get the weighted frequency domain representation , the specific method is as follows:

[0112] ;

[0113] in, To calculate environmental parameters using discrete Fourier transform Frequency domain representation of ; For frequency The corresponding weight coefficient; is the weighted frequency domain representation, indicating that the signal is at frequency The frequency domain components at .

[0114] In S3.4, the correction coefficient matrix is ​​defined To characterize the effects of temperature, humidity and air pressure on the rate of event generation, the specific method is as follows:

[0115] S3.4.1. Calculate the current time Deviation of environmental parameters from the reference value :

[0116] ;

[0117] S3.4.2. Define the correction factor matrix To characterize the combined effect of temperature, humidity and air pressure on the rate at which events occur:

[0118] ;

[0119] in, is the identity matrix; is the weight matrix; is a vector containing the deviations of the environmental parameters of temperature, humidity and pressure;

[0120] S3.4.3. Constructing the weight matrix ,according to Each element in the weight matrix is ​​calculated by the weighted sum of :

[0121] ;

[0122] in, and Environmental parameters and In frequency The frequency domain amplitude at ;

[0123] according to Each element in the weight matrix is ​​calculated by the weighted sum of The reason is that each frequency weight represents the contribution of a specific frequency component to the event generation rate, and the elements in the weight matrix It is necessary to reflect the mutual influence between environmental parameters at different frequencies by weighting these frequencies. By performing weighted sum, we can comprehensively consider the impact of each frequency component on different environmental parameters, thus constructing a weight matrix that accurately reflects these interactions;

[0124] S3.4.4. Finally, the correction coefficient matrix is ​​obtained .

[0125] Correction coefficient matrix The correction coefficients in are nonlinearly combined, and the correction coefficient matrix Perform point-by-point integral correction. The specific method is as follows:

[0126] ;

[0127] ;

[0128] in, is the correction coefficient matrix Middle Row, No. Elements of a column; is the coefficient of the nonlinear combination, representing the environmental parameters The sensitivity of changes in the rate at which events occur to the event; Environmental parameters In time The deviation of a moment from its reference value; is a nonlinear index, which controls the nonlinearity of the correction coefficient; is the weight matrix Middle Row, No. Elements of a column; is the time when monitoring starts, which is the initial time of integration;

[0129] In S3.6, combined with the correction coefficient matrix And the weighted frequency domain representation Calculate the rate of change of complex event generation rate , the specific method is as follows:

[0130]

[0131] in, The rate change rate of the event generation is the rate change of hydrogen leakage; For the time camera at time The rate at which events are generated; For time interval The change in the rate at which internal events occur; For time The relevant weight coefficient; Environmental parameters The wavelet transform result of

[0132] With time The relevant weight coefficient Directly set according to industry standards;

[0133] S4. Setting and monitoring the event generation rate change rate threshold, when the event generation rate change rate exceeds the threshold, automatically triggering an alarm signal, indicating that a hydrogen leak is detected;

[0134] Set and monitor the event generation rate change rate threshold. When the event generation rate change rate exceeds the threshold, the system will automatically trigger an alarm signal to indicate that a hydrogen leak has been detected. The specific operations are as follows:

[0135] S4.1. Calculate the mean of the rate of change of the event generation rate under normal operating conditions and standard deviation , set the threshold based on the mean and standard deviation of the rate of change of event generation rate :

[0136] ;

[0137] in, is the mean of the rate of change of event generation rate; is the standard deviation of the rate of change of event generation rate; is the threshold value of the rate of change of event generation rate; is a constant, and ; The specific value of depends on the sensitivity requirements of the application and determines the sensitivity of the alarm trigger; when When is smaller, a lower threshold means that the system is more sensitive to abnormal fluctuations in the rate of change of event generation; even with smaller fluctuations, the system may consider that there is a potential hydrogen leak, thereby triggering an alarm; when When it is larger, a higher threshold is obtained, which means that the system is less sensitive to abnormal fluctuations in the rate of change of event generation; the system will trigger an alarm only when a larger fluctuation is detected;

[0138] S4.2. Continuously calculate the current moment in real time The rate of change of event generation rate and with the preset threshold Make comparisons;

[0139] S4.3. When detected The alarm signal is triggered when a hydrogen leak is detected.

[0140] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments, and the above embodiments and descriptions are only preferred examples of the present invention, and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, and these changes and improvements all fall within the scope of the present invention to be protected.

Claims

1. A method for monitoring gas density changes using an event camera, characterized in that: The following steps are involved: S1. Set a light source at one end of the monitored area and install a background plate opposite the light source to capture the changes in the schlieren after the light penetrates the monitored area; S2, fix the event camera and adjust the focal length of the event camera according to the distance between the event camera and the background plate, and set the trigger threshold of the event camera; S3. Use event cameras for real-time monitoring, combine backlight schlieren technology to capture the refractive index changes caused by hydrogen leakage, generate event streams related to brightness changes, combine the environmental factors of temperature, humidity and air pressure, use multi-dimensional environment to correct the event rate change rate algorithm, and calculate the event generation rate change rate; S4. Setting and monitoring the event generation rate change rate threshold, when the event generation rate change rate exceeds the threshold, automatically triggering an alarm signal, indicating that a hydrogen leak is detected; In S3, the multi-dimensional environment correction event rate change rate algorithm is implemented based on multi-scale analysis and frequency domain analysis, and monitors hydrogen leakage in real time and calculates the event rate change rate during hydrogen leakage according to the dynamic changes of temperature, humidity and air pressure environmental parameters through the environmental correction coefficient matrix; In S3, the event generation rate change rate is calculated by combining the environmental factors of temperature, humidity and air pressure and using a multi-dimensional environment correction event rate change rate algorithm. The specific method is as follows: S3.

1. Real-time collection of environmental parameters in the monitoring area , including temperature ,humidity and air pressure , and determine the temperature, humidity and pressure reference values ​​of environmental parameters under no-leak conditions , ; ; S3.2 Environmental parameters Perform wavelet transform to obtain features at different scales; S3.3 Environmental parameters Perform discrete Fourier transform and calculate the frequency domain representation , identify the main frequency components in the environmental parameters through frequency domain analysis, divide the frequency components into different frequency intervals based on the discrete Fourier transform results, and assign different weights to different frequency intervals , and finally obtain the weighted frequency domain representation ; S3.

4. Define the correction coefficient matrix To characterize the effects of temperature, humidity and air pressure on the rate at which events occur; S3.

5. Correction coefficient matrix The correction coefficients in are nonlinearly combined, and the correction coefficient matrix Perform point-by-point integration correction; S3.

6. Combined correction coefficient matrix And the weighted frequency domain representation Calculate the rate of change of complex event generation rate , For the time camera at time The rate at which events are generated; In S3.6, combined with the correction coefficient matrix And the weighted frequency domain representation Calculate the rate of change of complex event generation rate , the specific method is as follows: ; in, The rate change rate of the event generation is the rate change of hydrogen leakage; For the time camera at time The rate at which events are generated; For time interval The change in the rate at which internal events occur; For time The relevant weight coefficient; Environmental parameters The wavelet transform result.

2. The method for monitoring gas density changes using an event camera according to claim 1, characterized in that: In S1, the light source is selected as a highly focused LED light source, which is arranged at one end of the monitored area, and the light intensity of the highly focused LED light source is adjusted; The background was chosen to be a black acrylic board, which was installed directly opposite the highly focused LED light source, and a sunshade was installed around the background board.

3. The method for monitoring gas density changes using an event camera according to claim 2, characterized in that: In S2, the specific steps of setting the trigger threshold of the event camera are as follows: S2.

1. Use the default values ​​of the event camera as a starting point and perform preliminary threshold settings based on the light intensity and changes in the monitoring environment. S2.

2. Under the condition that the noise in the environment can be eliminated, gradually adjust the trigger threshold until the event camera can sensitively capture the light and shadow changes caused by the gas leakage, and determine the trigger threshold; S2.

3. Simulate gas leakage of different degrees in the monitoring area to observe whether the event camera can capture the changes in the schlieren on the background plate and generate events. Based on the test results, further optimize the camera position, focal length and threshold settings.

4. The method for monitoring gas density changes using an event camera according to claim 3, characterized in that: In S3.2, the environmental parameters Perform wavelet transform to obtain multi-scale features. The specific method is as follows: ; in, is the result of wavelet transform, which means that and scale Multi-scale features on is the time parameter; is the scale parameter, which is used to control the degree of expansion and contraction of the wavelet function; is the integration variable; It is the complex conjugate of the mother wavelet function and is used to construct the wavelet transform; are the translation and scaling parameters of the mother wavelet function.

5. The method for monitoring gas density changes using an event camera according to claim 4, characterized in that: In S3.3, the environmental parameters Perform discrete Fourier transform and calculate the frequency domain representation , identify the main frequency components in the environmental parameters through frequency domain analysis, divide the frequency components into different frequency intervals based on the discrete Fourier transform results, and assign different weights to different frequency intervals , and finally obtain the weighted frequency domain representation , the specific method is as follows: S3.3.

1. Using Discrete Fourier Transform to Calculate Environmental Parameters Frequency domain representation of ; S3.3.

2. Calculate the power spectrum after discrete Fourier transform , identify the main frequency components in the frequency domain and divide the frequency components into three intervals: low frequency, medium frequency and high frequency; S3.3.

3. Assign different frequency weights to different frequency intervals ; S3.3.

4. Representing the frequency domain Multiply by the corresponding frequency weight , and get the weighted frequency domain representation .

6. The method for monitoring gas density changes using an event camera according to claim 5, characterized in that: In S3.3.1, the environmental parameters are calculated using discrete Fourier transform Frequency domain representation of , the specific method is as follows: ; in, Environmental parameters At the point in time The value at It is the result of discrete Fourier transform, which indicates that the signal is in frequency The frequency domain component at ; is the frequency index, indicating the frequency components; is the time index, the range is ; is the total number of sampling points, indicating environmental parameters The number of samples after discretization; is the complex exponential form of the Fourier basis function; In S3.3.3, different frequency weights are assigned to different frequency intervals. , the specific weight distribution criteria are as follows: Low-frequency components: usually reflect long-term trends and stable changes in the environment and are given higher weights; Medium-frequency components: reflect medium-term changes and cyclical fluctuations, and the weights are adjusted appropriately according to the actual application scenarios; High-frequency components: usually correspond to short-term fluctuations and noise, and are given lower weights; In S3.3.4, the frequency domain is represented Multiply by the corresponding frequency weight , and get the weighted frequency domain representation , the specific method is as follows: ; in, To calculate environmental parameters using discrete Fourier transform Frequency domain representation of ; For frequency The corresponding weight coefficient; is the weighted frequency domain representation, indicating that the signal is at frequency The frequency domain components at .

7. The method for monitoring gas density changes using an event camera according to claim 6, characterized in that: In S3.4, the correction coefficient matrix is ​​defined To characterize the effects of temperature, humidity and air pressure on the rate of event generation, the specific method is as follows: S3.4.

1. Calculate the current time Deviation of environmental parameters from the reference value : ; S3.4.

2. Define the correction factor matrix To characterize the combined effect of temperature, humidity and air pressure on the rate at which events occur: ; in, is the identity matrix; is the weight matrix; is a vector containing the deviations of the environmental parameters of temperature, humidity and pressure; S3.4.

3. Constructing the weight matrix ,according to Each element in the weight matrix is ​​calculated by the weighted sum of : ; in, and Environmental parameters and In frequency The frequency domain amplitude at ; S3.4.

4. Finally, the correction coefficient matrix is ​​obtained .

8. The method for monitoring gas density changes using an event camera according to claim 7, characterized in that: In S3.5, the correction coefficient matrix The correction coefficients in are nonlinearly combined, and the correction coefficient matrix Perform point-by-point integral correction. The specific method is as follows: ; ; in, is the correction coefficient matrix Middle Row, No. Elements of a column; is the coefficient of the nonlinear combination, representing the environmental parameters The sensitivity of changes in the rate at which events occur to the event; Environmental parameters In time The deviation of a moment from its reference value; is a nonlinear index, which controls the nonlinearity of the correction coefficient; is the weight matrix Middle Row, No. Elements of a column; The time when monitoring starts is the initial time of integration.

9. The method for monitoring gas density changes using an event camera according to claim 8, characterized in that: In S4, a threshold value of the rate of change of the event generation rate is set and monitored. When the rate of change of the event generation rate exceeds the threshold value, the system will automatically trigger an alarm signal to indicate that a hydrogen leak has been detected. The specific operations are as follows: S4.

1. Calculate the mean of the rate of change of the event generation rate under normal operating conditions and standard deviation , set the threshold based on the mean and standard deviation of the rate of change of event generation rate : ; in, is the mean of the rate of change of event generation rate; is the standard deviation of the rate of change of event generation rate; is the threshold value of the rate of change of event generation rate; is a constant, and ; S4.

2. Continuously calculate the current moment in real time The rate of change of event generation rate and with the preset threshold Make comparisons; S4.

3. When detected The alarm signal is triggered when a hydrogen leak is detected.

Images