CN116381334A - Event camera-based power grid frequency estimation method and system - Google Patents
Event camera-based power grid frequency estimation method and system Download PDFInfo
- Publication number
- CN116381334A CN116381334A CN202310243639.1A CN202310243639A CN116381334A CN 116381334 A CN116381334 A CN 116381334A CN 202310243639 A CN202310243639 A CN 202310243639A CN 116381334 A CN116381334 A CN 116381334A
- Authority
- CN
- China
- Prior art keywords
- event
- grid frequency
- power grid
- time
- polarity
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 27
- 238000005286 illumination Methods 0.000 claims abstract description 90
- 230000008859 change Effects 0.000 claims abstract description 40
- 238000012952 Resampling Methods 0.000 claims abstract description 28
- 238000010606 normalization Methods 0.000 claims abstract description 24
- 238000001914 filtration Methods 0.000 claims description 26
- 238000005070 sampling Methods 0.000 claims description 19
- 238000001228 spectrum Methods 0.000 claims description 7
- 230000009466 transformation Effects 0.000 claims description 3
- 230000033001 locomotion Effects 0.000 description 8
- 238000010586 diagram Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 5
- 238000003384 imaging method Methods 0.000 description 3
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 230000002207 retinal effect Effects 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R23/00—Arrangements for measuring frequencies; Arrangements for analysing frequency spectra
- G01R23/02—Arrangements for measuring frequency, e.g. pulse repetition rate; Arrangements for measuring period of current or voltage
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Remote Monitoring And Control Of Power-Distribution Networks (AREA)
Abstract
The invention belongs to the technical field of electric power, and discloses a power grid frequency estimation method and system based on an event camera. Firstly, recording an event stream corresponding to the flickering of a light source in an illumination environment through an event camera; then resampling the event stream in the time dimension to obtain a plurality of event blocks which are uniformly sampled; then, carrying out polarity normalization on all event blocks to obtain the illumination intensity change polarity of each event block in the corresponding time; combining the illumination intensity change polarities of all event blocks according to time sequence to obtain a complete illumination polarity sequence; and finally, performing time-frequency analysis on the illumination polarity sequence to obtain power grid frequency estimation information corresponding to the illumination environment. The invention can obtain a good power grid frequency estimation result.
Description
Technical Field
The invention belongs to the technical field of electric power, and particularly relates to a power grid frequency estimation method and system based on an event camera.
Background
The grid supply frequency, i.e. the grid frequency (Electric Network Frequency, ENF), is the supply frequency in the urban distribution network, and the distribution network around the world adopts a transmission system standard based on a frequency nominal value of 50 or 60Hz, i.e. the grid frequency is nominal value of 50/60Hz around the world. Since the total instantaneous power used in a city is difficult to estimate, it is difficult for the power supply to match the supply and demand of power, the grid frequency cannot always be kept at its nominal value, but fluctuations occur. However, if the power grid frequency is too large compared with the nominal value fluctuation, the working states of power supply equipment and electric appliances in the power grid may be unstable to damage, and large economic loss is caused, so that the power supply departments of all the power grids control the fluctuation of the power grid frequency to be within a small range near the nominal value as much as possible, so as to ensure the safety and stability of the power, and the power grid frequency has the characteristic of randomly fluctuating within a small range near the nominal value. Meanwhile, the illumination intensity of the light source is in direct proportion to the instantaneous size of the power supply power, so that the illumination intensity of the light source can change along with the change of the power grid frequency, and the estimation of the change of the power grid frequency can be realized by recording the illumination change of the light source by utilizing images or videos. However, due to the disadvantages of the conventional camera such as low dynamic range and insufficient sampling rate, it is often difficult for the image or video to clearly record the illumination change, resulting in poor power grid frequency result estimated therefrom or difficulty in achieving correct estimation.
Disclosure of Invention
The invention provides a power grid frequency estimation method and a system based on an event camera, which solve the problem that the effect of estimating the power grid frequency by using a traditional camera in the prior art is poor.
The invention provides a power grid frequency estimation method based on an event camera, which comprises the following steps of:
step 1, recording an event stream corresponding to the flickering of a light source in an illumination environment through an event camera;
step 2, resampling the event stream in the time dimension to obtain a plurality of event blocks which are uniformly sampled;
step 3, carrying out polarity normalization on all event blocks to obtain the illumination intensity change polarity of each event block in the corresponding time; combining the illumination intensity change polarities of all event blocks according to time sequence to obtain a complete illumination polarity sequence;
and 4, performing time-frequency analysis on the illumination polarity sequence to obtain power grid frequency estimation information corresponding to the illumination environment.
Preferably, in the step 2, the resampling is performed as follows:
e i ＝e×g τ (t 0 +i×Δt)
wherein e is the original event stream recorded by the event camera, e i Is the i-th event block, g, obtained by resampling the original event stream e τ (t) is a rectangular function, t 0 Is the time at which the first event in the event stream occurs, Δt is the sampling interval;
g τ (t 0 +i×Δt)＝u(t 0 +i×Δt)-u(t 0 +(i-1)×Δt)
where u (t) is a step function.
Preferably, in the step 3, the manner of performing polarity normalization on the event block is as follows:
wherein E (i) is the polarity of illumination intensity variation of the ith event block in its corresponding time; sgn (n) is a sign function, and returns to +1 or-1;
is the number of positive events in the ith event block,/>
Is the number of negative events in the ith event block.
Preferably, in the step 3, the illumination polarity sequence is expressed as:
E(n)＝(E(1),E(2),…,E(i),…)
where E (n) is the illumination polarity sequence and n is the total number of event blocks.
Preferably, the step 4 includes the following substeps:
step 401, performing band-pass filtering on the illumination polarity sequence at the position twice the power grid frequency nominal value to obtain a filtering sequence;
step 402, performing short-time fourier transform analysis on the filtering sequence, performing energy peak search on the estimated frequency spectrum result in each time window, taking the frequency component with the strongest energy as an instantaneous power grid frequency estimation result of the corresponding time, and thus obtaining a power grid frequency estimation result at the second harmonic in the whole event stream time period;
step 403, normalizing the power grid frequency estimation result at the second harmonic to a power grid frequency nominal value to obtain a standard power grid frequency corresponding to the harmonic, and using the standard power grid frequency as power grid frequency estimation information corresponding to the lighting environment, wherein the standard power grid frequency is expressed as:
f ENF ＝f 2ENF /2
wherein f 2ENF Is the power grid frequency estimation result at the second harmonic, f ENF Is the standard grid frequency corresponding to the harmonic.
Preferably, in the step 402, when the short-time fourier transform analysis is performed on the filtering sequence, the time window is set to be 16×1/Δt, the step is set to be 1/Δt, and Δt is the sampling time interval set when the resampling is performed in the step 2.
In another aspect, the present invention provides an event camera-based grid frequency estimation system, comprising:
the event camera is used for recording and obtaining an event stream corresponding to the flickering of the light source in the lighting environment;
the resampling unit is used for resampling the event stream in the time dimension to obtain a plurality of event blocks which are uniformly sampled;
the polarity normalization unit is used for performing polarity normalization on all the event blocks to obtain the illumination intensity change polarity of each event block in the corresponding time; and the system is used for combining the illumination intensity change polarities of all event blocks in time sequence to obtain a complete illumination polarity sequence;
and the time-frequency analysis unit is used for performing time-frequency analysis on the illumination polarity sequence to obtain power grid frequency estimation information corresponding to the illumination environment.
Preferably, the resampling unit performs resampling in the following manner:
e i ＝e×g τ (t 0 +i×Δt)
wherein e is the original event stream recorded by the event camera, e i Is the i-th event block, g, obtained by resampling the original event stream e τ (t) is a rectangular function, t 0 Is the time at which the first event in the event stream occurs, Δt is the sampling interval;
g τ (t 0 +i×Δt)＝u(t 0 +i×Δt)-u(t 0 +(i-1)×Δt)
where u (t) is a step function.
Preferably, the manner in which the polarity normalization unit performs polarity normalization on the event block is as follows:
wherein E (i) is the polarity of illumination intensity variation of the ith event block in its corresponding time; sgn (n) is a sign function, and returns to +1 or-1;
is the number of positive events in the ith event block,/>
Is the number of negative events in the ith event block.
Preferably, the time-frequency analysis unit includes:
the band-pass filtering unit is used for carrying out band-pass filtering on the illumination polarity sequence at the position twice the nominal value of the power grid frequency to obtain a filtering sequence;
the transformation unit is used for carrying out short-time Fourier transform analysis on the filtering sequence, carrying out energy peak search on the estimated frequency spectrum result in each time window, taking the frequency component with the strongest energy as the instantaneous power grid frequency estimation result of the corresponding time, and obtaining the power grid frequency estimation result at the second harmonic in the whole event stream time period;
and the normalization unit is used for normalizing the power grid frequency estimation result at the second harmonic to the power grid frequency nominal value to obtain the standard power grid frequency corresponding to the harmonic and serve as power grid frequency estimation information corresponding to the lighting environment.
One or more technical schemes provided by the invention have at least the following technical effects or advantages:
the invention aims to solve the problems that the traditional camera is easy to be influenced by the low dynamic state and the low sampling rate of a sensor when estimating the power grid frequency information in the illumination change of a light source, and meanwhile, the recorded pixel intensity change in the video is easy to be interfered by factors such as the motion content except the illumination change, and the effect of the traditional camera on the power grid frequency estimation is poor; then normalizing the event polarity in the event block to obtain a polarity sequence corresponding to the time environment illumination change, namely obtaining an illumination polarity sequence; and finally, based on the recognition that the power grid frequency only fluctuates in a small range near the nominal value and the illumination change frequency is twice as high as the power grid frequency, carrying out band-pass filtering on the illumination polarity sequence at the twice nominal value, obtaining and standardizing the power grid frequency estimation result at the second harmonic through short-time Fourier transform and frequency spectrum peak search, and realizing the power grid frequency estimation based on the event camera. The invention not only provides the use of the event camera to realize the power grid frequency estimation for the first time, but also provides a specific solution for the situation that the data mode output by the event camera is greatly different from the traditional vision sensor and the method based on the traditional camera cannot be applied to the event stream generated by the event camera.
Drawings
FIG. 1 is a comparison of data record forms of a conventional camera and an event camera;
FIG. 2 is a schematic diagram of an event camera recording illumination changes;
fig. 3 is a schematic flow chart of a method for estimating a grid frequency based on an event camera according to an embodiment of the present invention;
fig. 4 is a schematic diagram of a process of resampling an event stream in a time dimension in an event camera-based grid frequency estimation method according to an embodiment of the present invention;
fig. 5 is a schematic diagram of a process of performing polarity normalization on an event block in an event camera-based power grid frequency estimation method according to an embodiment of the present invention;
fig. 6 is a schematic diagram of grid frequency estimation information obtained by a grid frequency estimation method based on an event camera according to an embodiment of the present invention.
Detailed Description
In order to better understand the above technical solutions, the following detailed description will refer to the accompanying drawings and specific embodiments.
In urban lighting environments, some illumination sources may be invisible to the human eye, but may flash recorded by sensors such as cameras. Because of the direct proportion relation between the illumination intensity of the light source and the power supply power, the flicker frequency of the light source is twice the instantaneous value of the power grid frequency, namely the instantaneous power supply frequency in the power grid can be estimated from the illumination change recorded by the camera. However, when the traditional camera estimates the power grid frequency information in the illumination change of the light source, the influence of the low dynamic state and the low sampling rate of the sensor is easy to occur, and meanwhile, the change of the pixel intensity in the recorded video is also easy to be interfered by factors such as the motion content except the illumination change, so that the effect of estimating the power grid frequency by the traditional camera is poor.
The invention employs event cameras or dynamic vision sensors (Dynamic Vision Sensor, DVS)Is a novel sensor with a higher dynamic range and sampling rate than conventional vision sensors. Unlike conventional sensors that use a fixed sampling rate to record the absolute value of pixel intensity, event cameras simulate the principle of biological retinal imaging, producing pulses in response to changes in intensity at each pixel separately. More specifically, the event camera is a discrete asynchronous sensor whose output is a series of event stream data, not a standard image, due to its unique imaging modality that only records pixel intensity changes. If at time t j At the time, pixel position u j ＝(x j ,y j ) The intensity change at the location reaches a threshold + -C (C > 0), then the location will generate an event e j ＝(x j ,y j ,t j ,p j )，p j E { +1, -1} represents the polarity of the event, +1 represents the increase in pixel intensity, -1 represents the decrease in pixel intensity, so the event camera outputs an asynchronous event stream that records the direction of change in pixel intensity, rather than the absolute value of pixel intensity recorded in a conventional image. In addition to the fact that an event camera can capture luminance changes at an almost infinite frame rate and record the time and pixel locations at which event points are generated, compared to conventional frame-based image sensors, event cameras have a very high dynamic range that can respond to and record very small pixel changes. Because of this, event cameras have great advantages in terms of data generation principle, sampling rate and dynamic range for scenes where illumination intensity changes are weak or where pixel changes are disturbed by motion.
As shown in fig. 1, a comparison of the data recorded versions of a rotating disk with dots is captured using both a conventional camera that outputs a series of images of absolute intensity of pixels recorded at fixed time intervals and an event camera that outputs a stream of discrete events recorded at a high sampling rate. Because the data pattern output by the event camera is greatly different from that of the traditional vision sensor, many existing methods based on the traditional camera cannot be applied to the event stream generated by the event camera.
Based on the above consideration, the invention not only carries out the power grid frequency estimation based on the event camera, but also develops a specific solution, and uses the characteristics of high dynamic range and high time resolution of the event camera, and realizes the recording of the power grid frequency information by comparing the difference value of the front-back change of the pixel intensity under the influence of illumination with a threshold value and reserving the polarity of the pixel change, as shown in fig. 2. The invention utilizes the characteristic that the event camera is extremely sensitive to illumination change, and can finally obtain a good power grid frequency estimation result.
The flow diagram of the power grid frequency estimation method based on the event camera is shown in fig. 3, and mainly comprises the steps of recording the light flicker of the lighting environment to obtain an event stream, and processing the event stream to realize power grid frequency estimation.
The present invention will be specifically described with reference to examples 1 and 2.
Example 1:
embodiment 1 provides a method for estimating a grid frequency based on an event camera, comprising the steps of:
step 1, recording an event stream corresponding to flickering of a light source in an illumination environment through an event camera.
Because of the unique imaging modality of the event camera (the event camera is independent for each pixel and outputs an event only when the change in pixel intensity exceeds a threshold), the dataform of the event stream is always discrete, i.e., any pixel location at any time in the event stream may have an event generated: at any instant in time, the location of the event occurrence is not fixed, nor is the time interval at which the event occurs for one of the pixels.
And step 2, resampling the event stream in the time dimension to obtain a plurality of event blocks which are uniformly sampled.
The purpose of step 2 is to change the discrete event stream from step 1 into a uniformly sampled representation, see fig. 4, in the following way,
e i ＝e×g τ (t 0 +i×Δt)
wherein e is the original event stream recorded by the event camera, e i Is to the original event streame, the ith event block g obtained after resampling τ (t) is a rectangular function, t 0 Is the time at which the first event in the event stream occurs and Δt is the sampling interval.
For example, Δt=1/1000 s may be set.
Rectangular function g in the above procedure τ (t 0 +i×Δt) can be represented by a step function:
g τ (t 0 +i×Δt)＝u(t 0 +i×Δt)-u(t 0 +(i-1)×Δt)
where u (t) is a step function.
After the above-mentioned sampling at equal time intervals, the original event stream is divided into successive event blocks e which are sampled at a fixed frequency i Each event block contains all event points within Δt time:
step 3, carrying out polarity normalization on all event blocks to obtain the illumination intensity change polarity of each event block in the corresponding time; and combining the illumination intensity change polarities of all the event blocks according to time sequence to obtain a complete illumination polarity sequence.
And normalizing the polarity of the event in each event block to obtain the polarity of the event block corresponding to the time illumination intensity change.
After the time dimension sampling in step 2, a complete event stream is converted into continuous event blocks with Δt as time intervals, but because of the event points generated by illumination changes in the event stream, many events which are irrelevant to illumination may be generated due to noise or motion, so that the polarity of the events in each event block is not uniform.
The complete event stream after the step 2 is divided into a plurality of event blocks according to time sequence, but the discrete data form of the event stream causes uneven recording of the events in the event blocks: the event-generated pixel locations are not fixed. At the same time, the polarities of the events in the event block are not consistent: in a static scene, due to the characteristic of high sensitivity of the event, a plurality of noise events are inevitably generated, and the polarity of the noise events is unpredictable; in dynamic scenes, the event camera is also very sensitive to motion-generated pixel variations, so there are a large number of illumination-independent motion events in the event block at this time.
And normalizing the polarities of all the events in the event block in order to judge the direction of illumination change in the corresponding time of the event block.
Based on the knowledge that the number of noise events is relatively small and the number of positive and negative polarity events close to each other are detected at the same time when the motion of the closed object occurs, in the event block polarity normalization stage, the polarity of the event change in the corresponding time period of each event block is judged by adopting a mode criterion, and an illumination polarity sequence E (n) is obtained, as shown in fig. 5. The ith event block polarity normalization is expressed as follows:
wherein E (i) is the polarity of illumination intensity variation of the ith event block in its corresponding time; sgn (n) is a sign function, and returns to +1 or-1;
is the number of positive events in the ith event block,/>
Is the number of negative events in the ith event block.
The invention uses sgn (n) function to convert the polarities of a plurality of events in the event block into single illumination polarity with +1/-1 to represent, which can remove the influence of noise event and motion event polarity.
The illumination intensity variation polarities obtained by normalizing each event block are combined in time sequence, so that a sequence E (n) of complete illumination polarities is obtained:
E(n)＝(E(1),E(2),…,E(i),…)
where E (n) is the illumination polarity sequence and n is the total number of event blocks.
And 4, performing time-frequency analysis on the illumination polarity sequence to obtain power grid frequency estimation information corresponding to the illumination environment.
The event stream is processed and then converted into an illumination polarity sequence E (n) with the value of +1/-1, wherein the illumination polarity sequence E (n) contains information of flickering of a light source. And performing time-frequency analysis on the illumination polarity sequence at the position of 2 times of the nominal value of the power grid frequency to obtain the final estimation of the power grid frequency.
Specifically, the step 4 includes the following substeps:
step 401, performing band-pass filtering on the illumination polarity sequence at the position twice the nominal value of the power grid frequency to obtain a filtering sequence.
Because the flicker frequency of the illumination light source is twice the instantaneous value of the grid frequency, the illumination polarity sequence E (n) is firstly subjected to band-pass filtering at twice the nominal value of the grid frequency to obtain a corresponding filtered sequence, and the sequence is expressed as follows:
E(n)→E 2ENF (n)
wherein E is 2ENF (n) is a filtered sequence, i.e. a sequence obtained after bandpass filtering E (n) at twice the nominal value of the grid frequency.
Step 402, performing short-time fourier transform analysis on the filtering sequence, performing energy peak search on the estimated spectrum result in each time window, taking the frequency component with the strongest energy as the instantaneous grid frequency estimation result of the corresponding time, and thus obtaining the grid frequency estimation result f at the second harmonic in the whole event stream time period 2ENF 。
The short-time fourier transform formula is:
wherein w (n) is a window function; r is the step size, m is the step number of the window function, and mR represents the step number m to the right; ω is angular frequency.
For example, the time window is set to 16×1/Δt, the step is set to 1/Δt, and Δt is the sampling time interval set when the resampling is performed in step 2.
Step 403, normalizing the power grid frequency estimation result at the second harmonic to a power grid frequency nominal value to obtain a standard power grid frequency corresponding to the harmonic, and using the standard power grid frequency as power grid frequency estimation information corresponding to the lighting environment, wherein the standard power grid frequency is expressed as:
f ENF ＝f 2ENF /2
wherein f 2ENF Is the power grid frequency estimation result at the second harmonic, f ENF Is the standard grid frequency corresponding to the harmonic.
Normalizing the power grid frequency estimation result at the position of twice the nominal value to the position of the nominal value of the power grid frequency to obtain the standard power grid frequency f corresponding to the harmonic wave ENF . The second harmonic estimation result is divided by 2 for standardization due to the doubling relation between the change of the lighting flicker frequency and the change of the power grid frequency. Fig. 6 shows an example of normalized estimation results of the grid frequency (in the figure, +0.02 is an additional offset added to avoid the problem of curve overlapping, so as to more clearly and obviously compare the estimation results with the reference values), and it can be seen that the estimation results are good and conform to the fluctuation rule of the grid frequency very well.
Example 2:
embodiment 2 provides an event camera based grid frequency estimation system comprising:
the event camera is used for recording and obtaining an event stream corresponding to the flickering of the light source in the lighting environment;
the resampling unit is used for resampling the event stream in the time dimension to obtain a plurality of event blocks which are uniformly sampled;
the polarity normalization unit is used for performing polarity normalization on all the event blocks to obtain the illumination intensity change polarity of each event block in the corresponding time; and the system is used for combining the illumination intensity change polarities of all event blocks in time sequence to obtain a complete illumination polarity sequence;
and the time-frequency analysis unit is used for performing time-frequency analysis on the illumination polarity sequence to obtain power grid frequency estimation information corresponding to the illumination environment.
Specifically, the resampling unit performs resampling in the following manner:
e i ＝e×g τ (t 0 +i×Δt)
wherein e is the original event stream recorded by the event camera, e i Is the i-th event block, g, obtained by resampling the original event stream e τ (t) is a rectangular function, t 0 Is the time at which the first event in the event stream occurs, Δt is the sampling interval;
g τ (t 0 +i×Δt)＝u(t 0 +i×Δt)-u(t 0 +(i-1)×Δt)
where u (t) is a step function.
The manner in which the polarity normalization unit performs polarity normalization on the event block is as follows:
wherein E (i) is the polarity of illumination intensity variation of the ith event block in its corresponding time; sgn (n) is a sign function, and returns to +1 or-1;
is the number of positive events in the ith event block,/>
Is the number of negative events in the ith event block.
The time-frequency analysis unit includes:
the band-pass filtering unit is used for carrying out band-pass filtering on the illumination polarity sequence at the position twice the nominal value of the power grid frequency to obtain a filtering sequence;
the transformation unit is used for carrying out short-time Fourier transform analysis on the filtering sequence, carrying out energy peak search on the estimated frequency spectrum result in each time window, taking the frequency component with the strongest energy as the instantaneous power grid frequency estimation result of the corresponding time, and obtaining the power grid frequency estimation result at the second harmonic in the whole event stream time period;
and the normalization unit is used for normalizing the power grid frequency estimation result at the second harmonic to the power grid frequency nominal value to obtain the standard power grid frequency corresponding to the harmonic and serve as power grid frequency estimation information corresponding to the lighting environment.
Embodiment 2 provides a system corresponding to the method provided in embodiment 1, and the system provided in embodiment 2 is used to perform the steps in the method in embodiment 1, so that a detailed description is omitted, and the system can be understood with reference to the description of the method.
The power grid frequency estimation method and system based on the event camera provided by the embodiment of the invention at least comprise the following technical effects:
compared with the traditional camera which is easy to be interfered by scene factors such as movement when the illumination of the light source is changed, and the traditional vision sensor has the defects of insufficient sampling and the like, so that the problem that the power grid frequency is difficult to estimate correctly in video recording is solved.
Finally, it should be noted that the above-mentioned embodiments are only for illustrating the technical solution of the present invention, and not for limiting the same, and although the present invention has been described in detail with reference to examples, it should be understood by those skilled in the art that modifications and equivalents may be made to the technical solution of the present invention without departing from the spirit and scope of the technical solution of the present invention, and all such modifications and equivalents are intended to be encompassed in the scope of the claims of the present invention.
Claims (10)
1. The event camera-based power grid frequency estimation method is characterized by comprising the following steps of:
step 1, recording an event stream corresponding to the flickering of a light source in an illumination environment through an event camera;
step 2, resampling the event stream in the time dimension to obtain a plurality of event blocks which are uniformly sampled;
step 3, carrying out polarity normalization on all event blocks to obtain the illumination intensity change polarity of each event block in the corresponding time; combining the illumination intensity change polarities of all event blocks according to time sequence to obtain a complete illumination polarity sequence;
and 4, performing time-frequency analysis on the illumination polarity sequence to obtain power grid frequency estimation information corresponding to the illumination environment.
2. The event camera based grid frequency estimation method according to claim 1, wherein in step 2, the resampling is performed as follows:
e i ＝e×g τ (t 0 +i×Δt)
wherein e is the original event stream recorded by the event camera, e i Is the i-th event block, g, obtained by resampling the original event stream e τ (t) is a rectangular function, t 0 Is the time at which the first event in the event stream occurs, Δt is the sampling interval;
g τ (t 0 +i×Δt)＝u(t 0 +i×Δt)-u(t 0 +(i-1)×Δt)
where u (t) is a step function.
3. The event camera-based grid frequency estimation method according to claim 1, wherein in the step 3, the event block is subjected to polarity normalization in the following manner:
wherein E (i) is the polarity of illumination intensity variation of the ith event block in its corresponding time; sgn (n) is a sign function, and returns to +1 or-1;
is the number of positive events in the ith event block,/>
Is the number of negative events in the ith event block.
4. The event camera based grid frequency estimation method according to claim 3, wherein in step 3, the illumination polarity sequence is expressed as:
E(n)＝(E(1),E(2),...,E(i),...)
where E (n) is the illumination polarity sequence and n is the total number of event blocks.
5. The event camera based grid frequency estimation method according to claim 1, wherein the step 4 comprises the sub-steps of:
step 401, performing band-pass filtering on the illumination polarity sequence at the position twice the power grid frequency nominal value to obtain a filtering sequence;
step 402, performing short-time fourier transform analysis on the filtering sequence, performing energy peak search on the estimated frequency spectrum result in each time window, taking the frequency component with the strongest energy as an instantaneous power grid frequency estimation result of the corresponding time, and thus obtaining a power grid frequency estimation result at the second harmonic in the whole event stream time period;
step 403, normalizing the power grid frequency estimation result at the second harmonic to a power grid frequency nominal value to obtain a standard power grid frequency corresponding to the harmonic, and using the standard power grid frequency as power grid frequency estimation information corresponding to the lighting environment, wherein the standard power grid frequency is expressed as:
f ENF ＝f 2ENF /2
wherein f 2ENF Is the power grid frequency estimation result at the second harmonic, f ENF Is the standard grid frequency corresponding to the harmonic.
6. The event camera based grid frequency estimation method according to claim 5, wherein in step 402, when the filtering sequence is subjected to short-time fourier transform analysis, the time window is set to 16 x 1/Δt, the step is set to 1/Δt, and Δt is the sampling time interval set when resampling is performed in step 2.
7. A system for event camera based grid frequency estimation, comprising:
the event camera is used for recording and obtaining an event stream corresponding to the flickering of the light source in the lighting environment;
the resampling unit is used for resampling the event stream in the time dimension to obtain a plurality of event blocks which are uniformly sampled;
the polarity normalization unit is used for performing polarity normalization on all the event blocks to obtain the illumination intensity change polarity of each event block in the corresponding time; and the system is used for combining the illumination intensity change polarities of all event blocks in time sequence to obtain a complete illumination polarity sequence;
and the time-frequency analysis unit is used for performing time-frequency analysis on the illumination polarity sequence to obtain power grid frequency estimation information corresponding to the illumination environment.
8. The event camera based grid frequency estimation system of claim 7 wherein the resampling unit performs resampling in the following manner:
e i ＝e×g τ (t 0 +i×Δt)
wherein e is the original event stream recorded by the event camera, e i Is the i-th event block, g, obtained by resampling the original event stream e τ (t) is a rectangular function, t 0 Is the time at which the first event in the event stream occurs, Δt is the sampling interval;
g τ (t 0 +i×Δt)＝u(t 0 +i×Δt)-u(t 0 +(i-1)×Δt)
where u (t) is a step function.
9. The event camera based grid frequency estimation system of claim 7 wherein the polarity normalization unit performs polarity normalization of event blocks as follows:
wherein E (i) is the polarity of illumination intensity variation of the ith event block in its corresponding time; sgn (n) is a sign function, and returns to +1 or-1;
is the number of positive events in the ith event block,/>
Is the number of negative events in the ith event block.
10. The event camera based grid frequency estimation system of claim 7, wherein the time-frequency analysis unit comprises:
the band-pass filtering unit is used for carrying out band-pass filtering on the illumination polarity sequence at the position twice the nominal value of the power grid frequency to obtain a filtering sequence;
the transformation unit is used for carrying out short-time Fourier transform analysis on the filtering sequence, carrying out energy peak search on the estimated frequency spectrum result in each time window, taking the frequency component with the strongest energy as the instantaneous power grid frequency estimation result of the corresponding time, and obtaining the power grid frequency estimation result at the second harmonic in the whole event stream time period;
and the normalization unit is used for normalizing the power grid frequency estimation result at the second harmonic to the power grid frequency nominal value to obtain the standard power grid frequency corresponding to the harmonic and serve as power grid frequency estimation information corresponding to the lighting environment.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310243639.1A CN116381334A (en) | 2023-03-14 | 2023-03-14 | Event camera-based power grid frequency estimation method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310243639.1A CN116381334A (en) | 2023-03-14 | 2023-03-14 | Event camera-based power grid frequency estimation method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116381334A true CN116381334A (en) | 2023-07-04 |
Family
ID=86968508
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310243639.1A Pending CN116381334A (en) | 2023-03-14 | 2023-03-14 | Event camera-based power grid frequency estimation method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116381334A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116546339A (en) * | 2023-07-07 | 2023-08-04 | 深圳时识科技有限公司 | Noise reduction device and method for distinguishing event polarity, chip and electronic equipment |
-
2023
- 2023-03-14 CN CN202310243639.1A patent/CN116381334A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116546339A (en) * | 2023-07-07 | 2023-08-04 | 深圳时识科技有限公司 | Noise reduction device and method for distinguishing event polarity, chip and electronic equipment |
| CN116546339B (en) * | 2023-07-07 | 2023-09-08 | 深圳时识科技有限公司 | Noise reduction device and method for distinguishing event polarity, chip and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11856290B2 (en) | Method and apparatus of processing a signal from an event-based sensor | |
| JP5866200B2 (en) | Flicker detection circuit for image sensor using rolling shutter | |
| CN109167997B (en) | Video quality diagnosis system and method | |
| TWI419082B (en) | Moving object detection method and image processing system for moving object detection | |
| CN102469302B (en) | Background model learning system for lighting change adaptation utilized for video surveillance | |
| CN104023166B (en) | A kind of environment self-adaption video image noise reducing method and device | |
| CN116381334A (en) | Event camera-based power grid frequency estimation method and system | |
| WO2014199542A1 (en) | Imaging device, integrated circuit, and flicker reduction method | |
| CN111724317A (en) | Method for constructing Raw domain video denoising supervision data set | |
| CN101411190B (en) | Spurious motion filter | |
| Yuan et al. | Single image dehazing via NIN-DehazeNet | |
| CN110414558A (en) | Characteristic point matching method based on event camera | |
| Wang et al. | A linear comb filter for event flicker removal | |
| CN111899193A (en) | Criminal investigation photography system and method based on low-illumination image enhancement algorithm | |
| CN111460964A (en) | Moving target detection method under low-illumination condition of radio and television transmission machine room | |
| JP3347817B2 (en) | Image recognition device | |
| TWI425446B (en) | A method for object detection system in day-and-night environment | |
| WO2013114803A1 (en) | Image processing device, image processing method therefor, computer program, and image processing system | |
| CN110321378A (en) | A kind of mobile monitor image identification system and method | |
| US20200184606A1 (en) | Processing video frames via convolutional neural network using previous frame statistics | |
| CN102572374A (en) | Demisting image collection device | |
| CN111541877A (en) | Automatic monitoring system for substation equipment | |
| US8064655B2 (en) | Face image detection device, face image detection method and imaging apparatus | |
| WO2023001110A1 (en) | Neural network training method and apparatus, and electronic device | |
| CN201898578U (en) | Demisting image acquisition device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |