CN113807268B - Cylindrical object vibration detection method, device and computer storage medium - Google Patents
Cylindrical object vibration detection method, device and computer storage medium Download PDFInfo
- Publication number
- CN113807268B CN113807268B CN202111105615.7A CN202111105615A CN113807268B CN 113807268 B CN113807268 B CN 113807268B CN 202111105615 A CN202111105615 A CN 202111105615A CN 113807268 B CN113807268 B CN 113807268B
- Authority
- CN
- China
- Prior art keywords
- cylindrical object
- position change
- frequency
- vibration
- sequence
- 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.)
- Active
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computational Linguistics (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Health & Medical Sciences (AREA)
- Signal Processing (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Probability & Statistics with Applications (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The application provides a cylindrical object vibration detection method, a device and a computer storage medium, wherein the method comprises the following steps: performing time domain to frequency domain conversion on the position change time sequence of the cylindrical object to obtain a position change frequency spectrum signal; removing zero value peak positions on the position change spectrum number; calculating a series of maxima of the residual spectrum on the position change spectrum number; determining candidate peak positions on the position change spectrum number according to the maximum value of the residual spectrum on the position change spectrum number; determining a base frequency sequence according to the candidate peak position on the position change spectrum number and the corresponding frequency multiplication; and determining the vibration detection result of the cylindrical object according to the fundamental frequency sequence. The embodiment of the application can accurately and rapidly locate the reliable peak position under complex background and numerous mixed frequency conditions, and give out possible base frequency sequences according to frequency multiplication and peak values.
Description
Technical Field
The embodiment of the application relates to the field of computer vision, in particular to a cylindrical object vibration detection method, a cylindrical object vibration detection device and a computer storage medium.
Background
Vibration detection methods for cylindrical objects (e.g., such as guys on guys, high voltage transmission lines, posts, wires, ropes, etc.) are generally of the contact type and the non-contact type.
Taking a guy cable as an example, the contact type vibration detection technology uses a certain acceleration sensor to be attached to a target object to measure a vibration signal. The conventional contact vibration detection techniques include a vibration frequency method represented by an acceleration sensor, a magnetic flux method represented by an electromagnetic sensor, and a pressure method represented by a through-hole test ring, and these sensors are all required to contact a cable for measurement. The contact technology is costly, complex to implement in the field, difficult to operate, and most field conditions do not allow for the placement of sensors on the target object.
Non-contact vibration detection techniques include the use of both microwave radar techniques and computer vision techniques. Vibration detection technology based on microwave radar often needs to strike a stay rope with a hammer to increase the vibration amplitude of the stay rope, so that the sensitivity of the vibration detection technology is generally lower than that of the vibration detection technology based on computer vision. Computer vision-based vibration detection techniques basically use feature matching methods to track the change in position of the target at each frame. In order to improve the cable recognition accuracy, an image segmentation method based on a convolutional neural network is applied to cable recognition, and then the position of each frame of cable is tracked using a gravity center method.
However, in the prior art, under the complex background and numerous mixed frequency conditions, the reliable peak position cannot be accurately and rapidly positioned, and possible base frequency sequences are given according to frequency multiplication and peak values.
Disclosure of Invention
In view of the above, one of the technical problems to be solved by the embodiments of the present application is to provide a cylindrical object vibration detection scheme for overcoming or alleviating the above-mentioned drawbacks in the prior art.
A vibration detection method of a cylindrical object, comprising:
performing time domain to frequency domain conversion on the position change time sequence of the cylindrical object to obtain a position change frequency spectrum signal;
removing zero value peak positions on the position change spectrum number;
calculating a series of maxima of the residual spectrum on the position change spectrum number;
determining candidate peak positions on the position change spectrum number according to the maximum value of the residual spectrum on the position change spectrum number;
determining a base frequency sequence according to the candidate peak position on the position change spectrum number and the corresponding frequency multiplication;
and determining the vibration detection result of the cylindrical object according to the fundamental frequency sequence.
Optionally, the calculating the maximum value of the residual spectrum on the position change spectrum number includes: calculating a peak height-half width ratio corresponding to the maximum value of the residual spectrum on the position change spectrum number; identifying high-frequency background information from the residual frequency spectrum on the position change spectrum number, and calculating a corresponding high-frequency background value;
correspondingly, the determining the candidate peak position on the position change spectrum number according to the maximum value of the residual spectrum on the position change spectrum number comprises the following steps: and determining peaks in the residual spectrum on the position change spectrum number, which are larger than the high-frequency background value and larger than a relatively loose peak height-half-width ratio, and taking the corresponding peak positions as the candidate peak positions.
Optionally, the removing the zero-valued peak on the position-change spectrum number includes, before: and carrying out logarithmic transformation on the position change spectrum number, and determining zero value peak positions on the position change spectrum number according to the result of the logarithmic transformation.
Optionally, the determining the base frequency sequence according to the candidate peak position on the position change spectrum number and the corresponding frequency multiplication thereof includes: and selecting and calculating the frequency multiplication corresponding to the candidate peak position on the position change spectrum number, and determining a base frequency sequence.
Optionally, the converting the time domain to the frequency domain of the position change time sequence of the cylindrical object to obtain a position change spectrum signal includes:
cutting the image of the cylindrical object to obtain a plurality of frames of partial images of the cylindrical object;
and generating a position change time sequence of the cylindrical object according to the gravity centers of the cylindrical object corresponding to the partial images of the two adjacent frames.
Optionally, the slicing the image of the cylindrical object to obtain multiple frames of partial images of the cylindrical object includes: and cutting the image of the cylindrical object by using a multi-sliding window to obtain a plurality of frames of partial images of the cylindrical object.
Optionally, after the slicing the image of the cylindrical object to obtain multiple frames of partial images of the cylindrical object, generating a position change time sequence of the cylindrical object according to the center of gravity of the cylindrical object corresponding to the partial images of the cylindrical object in two adjacent frames includes: and calculating the gravity center of the cylindrical object corresponding to the partial images of the cylindrical object in two frames.
Optionally, the generating a position change time sequence of the cylindrical object according to the center of gravity of the cylindrical object corresponding to the partial images of the two adjacent frames of the cylindrical object includes:
generating a position change time source sequence of the cylindrical object according to the gravity centers of the cylindrical object corresponding to the partial images of the two adjacent frames of the cylindrical object;
and performing cluster analysis on the position change time source sequence of the cylindrical object to generate the position change time sequence of the cylindrical object.
Optionally, before the slicing the image of the cylindrical object to obtain multiple frames of partial images of the cylindrical object, the method includes: based on the depth convolution neural network, an image of the cylindrical object is determined and separated from the target video.
Optionally, the determining and separating the image of the cylindrical object from the target video based on the depth convolutional neural network includes: and carrying out image enhancement processing on the source video to obtain the target video, wherein the image enhancement comprises at least one of brightness enhancement, motion enhancement and vibration compensation.
Optionally, the determining the vibration detection result of the cylindrical object according to the fundamental frequency sequence includes:
based on the historical vibration detection result of the cylindrical object, predicting the vibration health condition of the cylindrical object according to the fundamental frequency sequence to obtain a vibration health condition prediction result;
and determining the vibration detection result of the cylindrical object according to the vibration health condition prediction result.
Optionally, the predicting, based on the historical vibration detection result of the cylindrical object, the vibration health condition of the cylindrical object according to the fundamental frequency sequence, to obtain a vibration health condition prediction result, includes:
if the historical vibration detection result of the cylindrical object shows that the frequency value sequence curve is simpler or tends to be obvious, a data fitting and extrapolation algorithm is adopted to predict the vibration health condition of the cylindrical object according to the fundamental frequency sequence so as to obtain a vibration health condition prediction result;
and if the frequency value sequence curve is complex and tends to be unobvious, predicting the vibration health condition of the cylindrical object according to the fundamental frequency sequence by adopting a deep neural network time sequence prediction algorithm to obtain a vibration health condition prediction result.
A vibration detection apparatus for a cylindrical object, comprising:
the conversion unit is used for performing time domain to frequency domain conversion on the position change time sequence of the cylindrical object to obtain a position change frequency spectrum signal;
the peak searching unit is used for removing zero value peak positions on the position change spectrum number; calculating a series of maxima of the residual spectrum on the position change spectrum number; determining candidate peak positions on the position change spectrum number according to the maximum value of the residual spectrum on the position change spectrum number;
the sequence determining unit is used for determining a base frequency sequence according to the candidate peak position on the position change spectrum number and the corresponding frequency multiplication;
and the vibration detection unit is used for determining a vibration detection result of the cylindrical object according to the fundamental frequency sequence.
A computer storage medium having a computer-executable program stored thereon that operates to implement a method according to any of the embodiments of the present application.
In the embodiment of the application, the position change frequency spectrum signal is obtained by performing time domain to frequency domain conversion on the position change time sequence of the cylindrical object; determining a base frequency sequence according to the candidate peak position on the position change spectrum number and the corresponding frequency multiplication; and determining the vibration detection result of the cylindrical object according to the base frequency sequence, so that the reliable peak position can be accurately and rapidly positioned under the complex background and numerous mixed frequency conditions, and possible base frequency sequences can be given according to frequency multiplication and peak values.
Drawings
Some specific embodiments of the application will be described in detail hereinafter by way of example and not by way of limitation with reference to the accompanying drawings. The same reference numbers will be used throughout the drawings to refer to the same or like parts or portions. Those skilled in the art will appreciate that the drawings are not necessarily drawn to scale. In the accompanying drawings:
FIG. 1 is a schematic diagram of a system architecture for implementing a vibration detection method for a cylindrical object according to the present application;
FIG. 2 is a flow chart of a method for detecting vibration of a cylindrical object according to an embodiment of the present application;
FIG. 3 is a schematic diagram of generating a position change time sequence through a computing center according to an embodiment of the present application;
FIG. 4 is a schematic diagram of the application of the scheme of the present application to a detection scenario of a bridge cable;
fig. 5 is a schematic flow chart of a vibration detecting device for a cylindrical object according to an embodiment of the application.
Detailed Description
It is not necessary for any of the embodiments of the application to be practiced with all of the advantages described above.
In order to better understand the technical solutions in the embodiments of the present application, the following description will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application, and it is obvious that the described embodiments are only some embodiments of the present application, not all embodiments. All other embodiments, which are derived by a person skilled in the art based on the embodiments of the present application, shall fall within the scope of protection of the embodiments of the present application.
FIG. 1 is a schematic diagram of a system architecture for implementing a vibration detection method for a cylindrical object according to the present application; the system photographs and detects a cylindrical object 100 at a distance. Of course, multiple cylindrical objects may be in the same field of view of the camera. As shown in fig. 1, it includes: the portable box 101, the camera device 102, the camera lens 103, the data transmission and control line 104, the computing unit 105, the user interface 106, and the database or server 107, the camera device 102, the data transmission and control line 104, the computing unit 105 are housed in the portable box 101, and the computing unit 105 is, for example, a microcomputer, a microprocessor, an FPGA-based or GPU-based processor. In the present application, the camera lens 103 is provided in the apparatus box 101, thereby providing an integrated system; the camera lens 103 and the camera apparatus 102 may also be separated from the portable box 101 as a separate system. The camera device 102 is connected to the computing unit 105 via a data transmission and control line 104 (including a USB line or a network line or a wireless network, etc.). The captured images or videos may be uploaded to a database or server 107 for storage or further analysis. The user may control the overall system, including displaying or manipulating images or video, through a user interface 106 (including a screen/mouse/keyboard, etc.).
Further, in order to reduce the vibration of the cylindrical object 100 from the vibration influencing system in the surrounding environment, a vibration-proof pad or spring structure 108 is added to the base of the camera 102 device. In the case of a compact system, a shock pad or shock spring structure 109 is also added to the base or support of the portable case 101.
Further, the acceleration sensor 110 may be provided on the camera apparatus 102 to measure the vibration frequency of the camera apparatus 102 itself, so as to remove the frequency caused by the self vibration of the camera apparatus, i.e., remove the impurity frequency, in performing the following steps S205-207, thereby ensuring the accuracy of the detection result.
FIG. 2 is a flow chart of a method for detecting vibration of a cylindrical object according to an embodiment of the present application; as shown in fig. 2, it includes:
s201, segmenting the image of the cylindrical object to obtain a plurality of frames of partial images of the cylindrical object;
in the step S201, the image of the cylindrical object is segmented to obtain a plurality of frames of partial images of the cylindrical object, including: and cutting the image of the cylindrical object by using a plurality of sliding windows to obtain a plurality of frames of partial images of the cylindrical object, thereby reducing separation errors caused by severe vibration of the cylindrical object.
S202, generating a position change time sequence of the cylindrical object according to the gravity centers of the cylindrical object corresponding to the partial images of the two adjacent frames.
In step S202, first, the center of gravity of the cylindrical object corresponding to the partial images of the cylindrical object in two frames is calculated, and a time sequence of the position change of the cylindrical object is generated according to the center of gravity of the cylindrical object corresponding to the partial images of the cylindrical object in two adjacent frames.
In the step S202, a time sequence of the position change of the cylindrical object is generated according to the center of gravity of the cylindrical object corresponding to the partial images of the two adjacent frames, and the time sequence includes:
generating a position change time source sequence of the cylindrical object according to the gravity centers of the cylindrical object corresponding to the partial images of the two adjacent frames of the cylindrical object;
and performing cluster analysis on the position change time source sequence of the cylindrical object to generate the position change time sequence of the cylindrical object.
FIG. 3 is a schematic diagram of generating a position change time sequence through a computing center according to an embodiment of the present application; as shown in fig. 3, a plurality of sliding windows are set, each sliding window cuts the image of the cylindrical object to obtain a frame of partial image of the cylindrical object, the plurality of sliding windows together obtain a plurality of frames of partial images of the cylindrical object, the center of gravity of the cylindrical object corresponding to the frame of partial image of the cylindrical object is calculated, the centers of gravity of the cylindrical object corresponding to the partial images of two adjacent frames of partial images of the cylindrical object are compared to obtain a position change time source sequence, and the position change time source sequence is subjected to cluster analysis, so that the final position change time sequence is obtained, namely the position change time sequence in the step S202, thereby reducing the center of gravity calculation error caused by severe vibration. The clustering algorithm may be, for example, the k-means algorithm, or, in other embodiments, the clustering algorithm may be replaced directly by an average or median.
If in fig. 3, for the starting time t0, a gravity center position value (x 0, y 0) can be calculated, and for a sub-image corresponding to a certain time ti, the position change is dxi =xi-x 0, dyi=yi-y 0, and for a partial image of the cylindrical object having N frames, the position change source time series is (dx 0, dy 0), (dx 1, dy 1), …, (dxi, dyi), …, (dxN, dyN); and clustering the position change source time sequence to obtain a position change time sequence.
S203, performing time domain to frequency domain conversion on the position change time sequence of the cylindrical object to obtain a position change frequency spectrum signal;
s204, carrying out logarithmic transformation on the position change spectrum number, and determining zero value peak positions on the position change spectrum number according to the result of the logarithmic transformation.
S205, removing the zero value peak position on the position change spectrum number;
s206, calculating a series of maximum values of the residual frequency spectrum on the position change spectrum number;
s207, calculating a peak height-half-width ratio corresponding to the maximum value of the residual spectrum on the position change spectrum number; identifying high-frequency background information from the residual frequency spectrum on the position change spectrum number, and calculating a corresponding high-frequency background value;
s208, determining candidate peak positions on the position change spectrum number according to the maximum value of the residual spectrum on the position change spectrum number.
In the step S208, determining candidate peak positions on the position-change spectrum number according to the maximum value of the residual spectrum on the position-change spectrum number includes: and determining a peak value which is larger than the high-frequency background value and larger than a relatively loose peak height-half-width ratio in the residual frequency spectrum on the position change spectrum number, and taking the corresponding peak position as the candidate peak position, thereby realizing accurate and rapid positioning of a reliable peak position under complex background and numerous mixed frequency conditions.
S209, determining a base frequency sequence according to the candidate peak position on the position change spectrum number and the corresponding frequency multiplication;
in step S209, determining a base frequency sequence according to the candidate peak position on the position-changing spectrum number and the corresponding frequency multiplication thereof, including: and selecting and calculating the corresponding frequency multiplication of the candidate peak positions on the position change spectrum number, and determining a base frequency sequence, thereby determining possible multiple relations of the candidate peak positions, such as 1.2Hz,2.4Hz and 3.6Hz ….
S210, determining a vibration detection result of the cylindrical object according to the fundamental frequency sequence.
In the step S210, determining a vibration detection result of the cylindrical object according to the fundamental frequency sequence includes:
based on the historical vibration detection result of the cylindrical object, predicting the vibration health condition of the cylindrical object according to the fundamental frequency sequence to obtain a vibration health condition prediction result;
and determining the vibration detection result of the cylindrical object according to the vibration health condition prediction result.
Optionally, the predicting, based on the historical vibration detection result of the cylindrical object, the vibration health condition of the cylindrical object according to the fundamental frequency sequence to obtain a vibration health condition prediction result includes:
if the historical vibration detection result of the cylindrical object shows that the frequency value sequence curve is simpler or tends to be obvious, a data fitting and extrapolation algorithm is adopted to predict the vibration health condition of the cylindrical object according to the fundamental frequency sequence so as to obtain a vibration health condition prediction result; data fitting and extrapolation algorithms such as linear interpolation, polynomial interpolation, etc.
And if the frequency value sequence curve is complex and tends to be unobvious, predicting the vibration health condition of the cylindrical object according to the fundamental frequency sequence by adopting a deep neural network time sequence prediction algorithm to obtain a vibration health condition prediction result. The deep neural network time series prediction algorithm adopts a basic LSTM (Long Short-Term Memory) network structure, for example.
In this embodiment, after the vibration health condition prediction result is obtained, a report may be created or an early warning may be sent, and the report and the video data of each inspection may be stored in a database or a server through WiFi or other internet data transmission methods. When the change or trend of the vibration (frequency, amplitude) through the cylindrical object exceeds a certain alarm threshold, an alarm is raised to the user through a user interface or server and other alarm devices. The user can extract reports and video data from the server at any time.
Specifically, in another embodiment, the slicing the image of the cylindrical object, before obtaining a plurality of frames of partial images of the cylindrical object, includes: based on the deep convolutional neural network, the image of the cylindrical object is determined and separated from the target video, so that the situation that the actual image has a complex background due to severe weather conditions and wrong shooting parameters and the cylindrical object cannot be accurately identified can be dealt with.
For example, if the input is a color image, for example, for each frame of the target image, it is first converted into a grayscale image. The full frame image size (height, width) is then changed to 512 x 512 image sizes as input to the deep convolutional neural network. Through the calculation of the deep convolutional neural network, a mask map with 512 x 512 is output, then the mask map is amplified into a mask map (height, width) and is matched with the size of the whole frame of image, and finally, the original whole frame of image is multiplied by the mask map, so that the image of the cylindrical object is separated, the calculation time is greatly reduced, and meanwhile, the segmentation precision of the cylindrical object can be ensured.
Further, the determining and separating the image of the cylindrical object from the target video based on the depth convolution neural network includes: and carrying out image enhancement processing on the source video to obtain the target video, wherein the image enhancement comprises at least one of brightness enhancement, motion enhancement and vibration compensation so as to solve the influence caused by stronger vibration (such as a cart passing or a strong wind environment) outside detection system equipment.
Specifically, for example, the Gamma enhancement technology is used to increase the brightness of the image, i.e. to realize brightness enhancement; the vibration amplitude of the cylindrical object in the image is enhanced by using a motion enhancement technology, namely, the motion enhancement is realized. If the system has stronger vibration (such as a cart passing or a strong wind environment), after each frame of image of the source video is matched with the whole image of the first image, the relative displacement of each frame of image is used for carrying out vibration compensation on each frame of image to obtain a target video, so that the situation that when cylindrical objects such as a guy cable, a bridge column or a high-voltage line are detected, for example, vibration of measuring equipment can be caused by passing vehicles, people and strong wind when the guy cable is measured, the photographed image is dithered, and a plurality of spectrum signals irrelevant to the vibration of the guy cable are mixed in the measured spectrum, so that the detection fails can be effectively caused.
The application provides a complete set of shockproof scheme, which comprises the steps of adding shockproof measures to the measuring equipment, and simultaneously measuring the vibration frequency of the equipment by using an acceleration sensor of the measuring equipment. In software, an image matching algorithm is used for calculating jitter displacement between frames for the whole frame image, and the self-oscillation displacement of the cylindrical object is compensated
FIG. 4 is a schematic diagram of the application of the scheme of the present application to a detection scenario of a bridge cable; as shown in fig. 4, a video of 2000 x 800 pixels is shot for 8 seconds by using a video camera with a frame frequency of 60f/s, and according to the data processing flow of fig. 2, the cable vibration spectrum analysis result is obtained, and the result shows that the very accurate fundamental frequency sequence sequences 3.01, 6.13, 9.14 and … 27.21.21 Hz are displayed, so that the fundamental frequency of the bridge cable is about 3.03Hz.
FIG. 5 is a flow chart of a vibration detecting apparatus for a cylindrical object according to an embodiment of the present application; as shown in fig. 5, it includes:
a conversion unit 501, configured to perform time-domain to frequency-domain conversion on the position change time sequence of the cylindrical object, so as to obtain a position change spectrum signal;
a peak searching unit 502, configured to remove a zero value peak position on the position change spectrum number; calculating a series of maxima of the residual spectrum on the position change spectrum number; determining candidate peak positions on the position change spectrum number according to the maximum value of the residual spectrum on the position change spectrum number;
a sequence determining unit 503, configured to determine a base frequency sequence according to the candidate peak positions on the position-changing spectrum number and the corresponding frequency multiples thereof;
and a vibration detection unit 504, configured to determine a vibration detection result of the cylindrical object according to the fundamental frequency sequence.
In the embodiment shown in fig. 5, the preferred explanation of each unit cell can be referred to the above method embodiment, and detailed description thereof will be omitted.
Thus, particular embodiments of the present subject matter have been described. Other embodiments are within the scope of the following claims. In some cases, the actions recited in the claims can be performed in a different order and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may be advantageous.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one …" does not exclude the presence of other like elements in a process, method, article or apparatus that comprises an element.
In this specification, each embodiment is described in a progressive manner, and identical and similar parts of each embodiment are all referred to each other, and each embodiment mainly describes differences from the other embodiments. In particular, for system embodiments, since they are substantially similar to method embodiments, the description is relatively simple, as relevant to see a section of the description of method embodiments.
The foregoing is merely exemplary of the present application and is not intended to limit the present application. Various modifications and variations of the present application will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement, etc. which come within the spirit and principles of the application are to be included in the scope of the claims of the present application.
Claims (9)
1. A vibration detection method of a cylindrical object, comprising:
cutting the image of the cylindrical object to obtain a plurality of frames of partial images of the cylindrical object;
generating a position change time sequence of the cylindrical object according to the gravity centers of the cylindrical object corresponding to the partial images of the two adjacent frames of the cylindrical object;
performing time domain to frequency domain conversion on the position change time sequence of the cylindrical object to obtain a position change frequency spectrum signal;
carrying out logarithmic transformation on the position change spectrum number, and determining zero value peak positions on the position change spectrum number according to the result of the logarithmic transformation;
removing zero-value peak positions on the position change spectrum signals;
calculating a series of maxima of a residual spectrum on the position-change spectrum signal;
calculating a peak height-half width ratio corresponding to the maximum value of the residual spectrum on the position change spectrum number; identifying high-frequency background information from the residual frequency spectrum on the position change spectrum number, and calculating a corresponding high-frequency background value;
determining a peak value which is larger than the high-frequency background value and larger than a peak height-half-width ratio in a residual frequency spectrum on the position change frequency spectrum signal, and taking a corresponding peak position as a candidate peak position;
determining a base frequency sequence according to the candidate peak position and the corresponding frequency multiplication of the position change spectrum signal;
determining a vibration detection result of the cylindrical object according to the fundamental frequency sequence;
wherein, according to the fundamental frequency sequence, determining the vibration detection result of the cylindrical object includes:
based on the historical vibration detection result of the cylindrical object, predicting the vibration health condition of the cylindrical object according to the fundamental frequency sequence to obtain a vibration health condition prediction result;
determining a vibration detection result of the cylindrical object according to the vibration health condition prediction result;
the method for predicting the vibration health condition of the cylindrical object based on the historical vibration detection result of the cylindrical object according to the fundamental frequency sequence to obtain a vibration health condition prediction result comprises the following steps:
if the historical vibration detection result of the cylindrical object shows that the frequency value sequence curve is simple or tends to be obvious, a data fitting and extrapolation algorithm is adopted to predict the vibration health condition of the cylindrical object according to the base frequency sequence so as to obtain a vibration health condition prediction result;
and if the frequency value sequence curve is complex and tends to be unobvious, predicting the vibration health condition of the cylindrical object according to the fundamental frequency sequence by adopting a deep neural network time sequence prediction algorithm to obtain a vibration health condition prediction result.
2. The method for detecting vibration of a cylindrical object according to claim 1, wherein determining a base frequency sequence according to candidate peak positions on the position-change spectrum signal and corresponding frequency multiples thereof comprises: and selecting and calculating frequency multiplication corresponding to the candidate peak position on the position change spectrum signal, and determining a base frequency sequence.
3. The vibration detecting method of claim 1, wherein the segmenting the image of the cylindrical object to obtain a plurality of frames of partial images of the cylindrical object comprises: and cutting the image of the cylindrical object by using a multi-sliding window to obtain a plurality of frames of partial images of the cylindrical object.
4. The method for detecting vibration of a cylindrical object according to claim 1, wherein after the image of the cylindrical object is segmented to obtain a plurality of frames of partial images of the cylindrical object, the generating a time sequence of position change of the cylindrical object according to the center of gravity of the corresponding cylindrical object in the partial images of the two adjacent frames of the cylindrical object includes: and calculating the gravity center of the cylindrical object corresponding to the partial images of the cylindrical object in two frames.
5. The vibration detecting method of a cylindrical object according to claim 1, wherein the generating a time series of position changes of the cylindrical object based on a center of gravity of the cylindrical object corresponding to partial images of the cylindrical object in two adjacent frames, comprises:
generating a position change time source sequence of the cylindrical object according to the gravity centers of the cylindrical object corresponding to the partial images of the two adjacent frames of the cylindrical object;
and performing cluster analysis on the position change time source sequence of the cylindrical object to generate the position change time sequence of the cylindrical object.
6. The method for detecting vibration of a cylindrical object according to claim 1, wherein the step of segmenting the image of the cylindrical object to obtain a plurality of partial images of the cylindrical object comprises: based on the depth convolution neural network, an image of the cylindrical object is determined and separated from the target video.
7. The vibration detection method of a cylindrical object according to claim 6, wherein the determining and separating an image of the cylindrical object from a target video based on a depth convolution neural network comprises: and carrying out image enhancement processing on the source video to obtain the target video, wherein the image enhancement comprises at least one of brightness enhancement, motion enhancement and vibration compensation.
8. A vibration detecting apparatus for a cylindrical object, comprising:
the conversion unit is used for performing time domain to frequency domain conversion on the position change time sequence of the cylindrical object to obtain a position change frequency spectrum signal;
a peak searching unit for removing zero value peak position on the position change frequency spectrum signal; calculating a series of maxima of a residual spectrum on the position-change spectrum signal; determining candidate peak positions on the position change spectrum signal according to the maximum value of the residual spectrum on the position change spectrum signal;
the sequence determining unit is used for determining a base frequency sequence according to the candidate peak position on the position change spectrum signal and the corresponding frequency multiplication;
the vibration detection unit is used for determining a vibration detection result of the cylindrical object according to the fundamental frequency sequence;
before the time domain to frequency domain conversion is performed on the position change time sequence of the cylindrical object to obtain a position change frequency spectrum signal, the method comprises the following steps:
cutting the image of the cylindrical object to obtain a plurality of frames of partial images of the cylindrical object;
generating a position change time sequence of the cylindrical object according to the gravity centers of the cylindrical object corresponding to the partial images of the two adjacent frames of the cylindrical object;
the removing of the zero-valued peak bits on the position-change spectrum signal is preceded by: carrying out logarithmic transformation on the position change spectrum number, and determining zero value peak positions on the position change spectrum number according to the result of the logarithmic transformation;
the calculating of the maximum value of the residual spectrum on the position change spectrum number comprises the following steps: calculating a peak height-half width ratio corresponding to the maximum value of the residual spectrum on the position change spectrum number; identifying high-frequency background information from the residual frequency spectrum on the position change spectrum number, and calculating a corresponding high-frequency background value;
correspondingly, the determining the candidate peak position on the position change spectrum number according to the maximum value of the residual spectrum on the position change spectrum number comprises the following steps: determining a peak value which is larger than the high-frequency background value and larger than a peak height-half-width ratio in a residual frequency spectrum on the position change spectrum number, and taking a corresponding peak position as the candidate peak position;
wherein, according to the fundamental frequency sequence, determining the vibration detection result of the cylindrical object includes:
based on the historical vibration detection result of the cylindrical object, predicting the vibration health condition of the cylindrical object according to the fundamental frequency sequence to obtain a vibration health condition prediction result;
determining a vibration detection result of the cylindrical object according to the vibration health condition prediction result;
the method for predicting the vibration health condition of the cylindrical object based on the historical vibration detection result of the cylindrical object according to the fundamental frequency sequence to obtain a vibration health condition prediction result comprises the following steps:
if the historical vibration detection result of the cylindrical object shows that the frequency value sequence curve is simple or tends to be obvious, a data fitting and extrapolation algorithm is adopted to predict the vibration health condition of the cylindrical object according to the base frequency sequence so as to obtain a vibration health condition prediction result;
and if the frequency value sequence curve is complex and tends to be unobvious, predicting the vibration health condition of the cylindrical object according to the fundamental frequency sequence by adopting a deep neural network time sequence prediction algorithm to obtain a vibration health condition prediction result.
9. A computer storage medium having a computer executable program stored thereon, the computer executable program being operative to implement the method of any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111105615.7A CN113807268B (en) | 2021-09-22 | 2021-09-22 | Cylindrical object vibration detection method, device and computer storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111105615.7A CN113807268B (en) | 2021-09-22 | 2021-09-22 | Cylindrical object vibration detection method, device and computer storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113807268A CN113807268A (en) | 2021-12-17 |
| CN113807268B true CN113807268B (en) | 2023-09-29 |
Family
ID=78939942
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111105615.7A Active CN113807268B (en) | 2021-09-22 | 2021-09-22 | Cylindrical object vibration detection method, device and computer storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113807268B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103994817A (en) * | 2014-05-19 | 2014-08-20 | 深圳艾瑞斯通技术有限公司 | Vibration source identification method based on long-distance optical fiber frequent occurring events |
| CN110346591A (en) * | 2018-04-05 | 2019-10-18 | 计算系统有限公司 | Machine rotational speed is determined based on rumble spectrum figure |
| WO2019220883A1 (en) * | 2018-05-18 | 2019-11-21 | 日本電信電話株式会社 | Sensing device, method for same, and program |
| CN110542474A (en) * | 2019-09-04 | 2019-12-06 | 中国科学院上海高等研究院 | Method, system, medium, and apparatus for detecting vibration signal of device |
| CN111275744A (en) * | 2020-01-20 | 2020-06-12 | 福州大学 | Non-contact vibration frequency measurement method based on deep learning and image processing |
| CN112733759A (en) * | 2021-01-15 | 2021-04-30 | 中国电力科学研究院有限公司 | Structural vibration measurement method and system based on visual image local binarization processing |
-
2021
- 2021-09-22 CN CN202111105615.7A patent/CN113807268B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103994817A (en) * | 2014-05-19 | 2014-08-20 | 深圳艾瑞斯通技术有限公司 | Vibration source identification method based on long-distance optical fiber frequent occurring events |
| CN110346591A (en) * | 2018-04-05 | 2019-10-18 | 计算系统有限公司 | Machine rotational speed is determined based on rumble spectrum figure |
| WO2019220883A1 (en) * | 2018-05-18 | 2019-11-21 | 日本電信電話株式会社 | Sensing device, method for same, and program |
| CN110542474A (en) * | 2019-09-04 | 2019-12-06 | 中国科学院上海高等研究院 | Method, system, medium, and apparatus for detecting vibration signal of device |
| CN111275744A (en) * | 2020-01-20 | 2020-06-12 | 福州大学 | Non-contact vibration frequency measurement method based on deep learning and image processing |
| CN112733759A (en) * | 2021-01-15 | 2021-04-30 | 中国电力科学研究院有限公司 | Structural vibration measurement method and system based on visual image local binarization processing |
Non-Patent Citations (1)
| Title |
|---|
| 基于激光多普勒技术的空调六维振动检测系统研究;郑文炜;中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑(第03期);1-77 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113807268A (en) | 2021-12-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR102348002B1 (en) | Pedestrian re-identification methods, devices, electronic devices and storage media | |
| US11392792B2 (en) | Method and apparatus for generating vehicle damage information | |
| JP2953712B2 (en) | Moving object detection device | |
| CN112488073A (en) | Target detection method, system, device and storage medium | |
| CN110555405B (en) | Target tracking method and device, storage medium and electronic equipment | |
| CN101888535A (en) | Motion object test equipment, motion object detection method and computer program | |
| JP2021048617A (en) | Information processing system, information processing method, and program | |
| CN113470024B (en) | Hub internal defect detection method, device, equipment, medium and program product | |
| JP4764487B2 (en) | Video surveillance system | |
| EP3410353A1 (en) | Method for estimating a timestamp in a video stream and method of augmenting a video stream with information | |
| CN110705759B (en) | Water level early warning monitoring method and device, storage medium and electronic equipment | |
| JP6944405B2 (en) | Building judgment system | |
| CN111723634B (en) | Image detection method and device, electronic equipment and storage medium | |
| EP3438875A1 (en) | Image processing apparatus and control method therefor | |
| CN113920538B (en) | Object detection method, device, equipment, storage medium and computer program product | |
| CN111353496B (en) | Real-time detection method for infrared dim targets | |
| US20170053172A1 (en) | Image processing apparatus, and image processing method | |
| US20200184622A1 (en) | Degradation detection device, degradation detection method, and computer-readable recording medium | |
| CN111815576A (en) | Method, device, equipment and storage medium for detecting corrosion condition of metal part | |
| CN113807268B (en) | Cylindrical object vibration detection method, device and computer storage medium | |
| CN112053363B (en) | Retina blood vessel segmentation method, retina blood vessel segmentation device and model construction method | |
| US20200311401A1 (en) | Analyzing apparatus, control method, and program | |
| CN117314863A (en) | Defect output method, device, equipment and storage medium | |
| JP2014074977A (en) | Image processing apparatus, image processing method, and image processing program | |
| CN111695404B (en) | Pedestrian falling detection method and device, electronic equipment and storage medium |
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 | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20230510 Address after: Room 1104, Unit 1, Building 5, Wanda Mansion, No. 5 Keyuan Road, Jinshui District, Zhengzhou City, Henan Province, 450008 Applicant after: Li Bing Address before: 101100 204-a4, block B, No. a 560, Luyuan South Street, Tongzhou District, Beijing Applicant before: Beijing quark Chuangzhi Technology Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |