CN214796752U - Engineering vehicle identification and positioning device based on voiceprint identification and sound source positioning - Google Patents
Engineering vehicle identification and positioning device based on voiceprint identification and sound source positioning Download PDFInfo
- Publication number
- CN214796752U CN214796752U CN202022756413.6U CN202022756413U CN214796752U CN 214796752 U CN214796752 U CN 214796752U CN 202022756413 U CN202022756413 U CN 202022756413U CN 214796752 U CN214796752 U CN 214796752U
- Authority
- CN
- China
- Prior art keywords
- model
- sound source
- shaped plate
- voiceprint
- identification
- 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
Images
Landscapes
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The utility model discloses an engineering vehicle discernment positioner based on voiceprint discernment and sound source location relates to signal processing technology field. The utility model discloses a training module, test module, base plate, training module, test module all include: preprocessing, wherein the preprocessing is connected with feature extraction, and the two feature extractions are respectively connected with a training model and comparison judgment; the training model is connected with a model base, the comparison judgment is connected with an output result, the model base is connected with the comparison judgment through mode matching, and the preprocessing is connected with the input end of the DFT; the base plate is internally provided with a channel. The utility model discloses a L shaped plate that sets up, when rotating the threaded rod, the threaded rod drives L shaped plate and rotates to fix the device, made things convenient for and installed and dismantle positioner, improved the efficiency of installation and dismantlement, and the U-shaped plate of setting has reduced the condition emergence of L shaped plate slope when sliding, thereby has improved the stability of device.
Description
Technical Field
The utility model belongs to the technical field of signal processing, especially, relate to an engineering vehicle discernment positioner based on voiceprint discernment and source location.
Background
Aiming at the technical problems in the prior art, the technology provides that the special vehicle on the construction site is identified and positioned by using a voiceprint identification technology and a sound source positioning technology, and is made into a device, the device can identify and position the special vehicle on the construction site in real time in complex environments such as the construction site, and when the special vehicle on the construction site works, the device gives an early warning to a management department and suspends or stops construction operation. The device is divided into two steps, wherein the first step is voiceprint recognition, and the second step is sound source positioning. The basic principle of voiceprint recognition is that a model base which belongs to the characteristics of a special vehicle on a construction site is constructed through audio, the model base mainly comprises the voiceprint characteristics of an engineering vehicle, then the voiceprint characteristics are compared with collected audio, and whether the voice is the audio of the model base is judged through mode matching so as to achieve the purpose of recognition. The basic principle of sound source positioning is to determine the position of the engineering vehicle through the geometrical relationship by the time difference of the audio signal of the engineering vehicle reaching each microphone sensor.
There are many problems in real-world operation using video image-based vehicle detection and vehicle type recognition or distributed optical fiber vibration sensors. In reality, we need carry out real-time identification and location to building site special type vehicle, the utility model provides a device of engineering vehicle discernment and location based on voiceprint discernment and sound source location. The recognition and positioning method and device based on the sound characteristics and the sound arrival difference can collect the sound through the sound collection equipment, then recognize whether the engineering vehicle exists in the sound, and determine the position of the engineering vehicle through the geometrical relation according to different arrival time of the sound of the engineering vehicle at each microphone sensor, and the video image device is troublesome to install, so that the maintenance of a maintainer can be influenced.
SUMMERY OF THE UTILITY MODEL
An object of the utility model is to provide an engineering vehicle discernment positioner based on voiceprint discernment and sound source location, through the L shaped plate that sets up, when rotating the threaded rod, the threaded rod drives the L shaped plate and rotates to fix the device, made things convenient for and installed and dismantle positioner, improved the efficiency of installation and dismantlement, and the U-shaped plate of setting has reduced the condition emergence that the L shaped plate inclines when sliding, thereby improved the stability of device, solved the problem that exists among the above-mentioned prior art.
In order to achieve the purpose, the utility model is realized by the following technical proposal:
the utility model provides an engineering vehicle discernment positioner based on voiceprint discernment and sound source location, includes training module, test module, base plate, and training module, test module all include: preprocessing, wherein the preprocessing is connected with feature extraction, and the two feature extractions are respectively connected with a training model and comparison judgment; the training model is connected with a model base, the comparison judgment is connected with an output result, the model base is connected with the comparison judgment through mode matching, and the preprocessing is connected with the input end of the DFT; the channel has been seted up to the inside of base plate, and the inside of channel is provided with two fixed subassemblies, and the inside normal running fit threaded rod of channel, the all sides screw-thread fit of threaded rod have two U-shaped boards, and fixed subassembly includes: an L-shaped plate, a rotary cylinder; the fixed ring is installed on the periphery of the rotating cylinder, the sliding plates are installed on the outer side of the fixed ring and are respectively in sliding fit with the two U-shaped plates, and the limiting blocks are installed on the periphery of the upper end of the rotating cylinder.
Optionally, one side of the third preprocessing is connected with a voice input, a voice to be detected and an audio signal respectively.
Optionally, the DFT is connected to an input of a Mel filter bank, the Mel filter bank is connected to an input of logarithmic energy, and the logarithmic energy is connected to an input of the DCT.
Optionally, a baffle is arranged inside the channel, a connecting plate is installed on one side of the baffle, a fixing rod is installed on one side of the connecting plate, and the fixing rod is matched with the L-shaped plate in a rotating mode.
Optionally, a spring is installed on one side of the L-shaped plate, and the spring is connected with the baffle.
Optionally, a rotating plate is mounted at the end of the threaded rod.
The embodiment of the utility model has the following beneficial effect:
the utility model discloses an embodiment is through the L shaped plate that sets up, and when rotating the threaded rod, the threaded rod drives L shaped plate and rotates to fix the device, made things convenient for and installed and dismantle positioner, improved the efficiency of installation and dismantlement, and the U-shaped plate of setting has reduced the condition emergence of L shaped plate slope when sliding, thereby has improved the stability of device.
Of course, it is not necessary for any particular product to achieve all of the above-described advantages at the same time.
Drawings
The accompanying drawings, which form a part of the present application, are included to provide a further understanding of the invention, and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the invention and not to limit the invention. In the drawings:
fig. 1 is a schematic structural diagram of a training module according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of a training module according to an embodiment of the present invention;
fig. 3 is a schematic diagram of a TDOA structure according to an embodiment of the present invention;
fig. 4 is a schematic diagram of a TDOA positioning structure according to an embodiment of the present invention;
fig. 5 is a schematic structural diagram of a GCC according to an embodiment of the present invention;
fig. 6 is a schematic view of a substrate structure according to an embodiment of the present invention;
fig. 7 is a schematic diagram of a channel structure according to an embodiment of the present invention;
fig. 8 is a schematic view of the structure at a in fig. 7.
Wherein the figures include the following reference numerals:
the structure comprises a base plate 1, a channel 2, a threaded rod 3, a fixing plate 4, a U-shaped plate 5, a rotating plate 6, a baffle 7, a connecting plate 8, a fixing rod 9, an L-shaped plate 10, a spring 11, a rotating cylinder 12, a fixing ring 13, a limiting block 14 and a sliding plate 15.
Detailed Description
The technical solutions in the embodiments of the present invention will be described clearly and completely with reference to the accompanying drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only some embodiments of the present invention, not all embodiments. The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the invention, its application, or uses.
To maintain the following description of the embodiments of the present invention clear and concise, detailed descriptions of well-known functions and components may be omitted.
Referring to fig. 1 to 8, in the present embodiment, an engineering vehicle identification and positioning device based on voiceprint identification and sound source positioning is provided, which includes: training module, test module, base plate 1, training module, test module all include: preprocessing, wherein the preprocessing is connected with feature extraction, and the two feature extractions are respectively connected with a training model and comparison judgment; the training model is connected with a model base, the comparison judgment is connected with an output result, the model base is connected with the comparison judgment through mode matching, and the preprocessing is connected with the input end of the DFT; channel 2 has been seted up to base plate 1's inside, and the inside of channel 2 is provided with two fixed subassemblies, and the inside normal running fit threaded rod 3 of channel 2, the week side screw-thread fit of threaded rod 3 have two U-shaped boards 5, and fixed subassembly includes: an L-shaped plate 10, a rotary cylinder 12; a fixed ring 13 is arranged on the periphery of the rotating cylinder 12, a sliding plate 15 is arranged on the outer side of the fixed ring 13, the two sliding plates 15 are respectively in sliding fit in the two U-shaped plates 5, and a limit block 14 is arranged on the periphery of the upper end of the rotating cylinder 12.
The application of one aspect of the embodiment is as follows:
firstly, collecting voice of a special vehicle on a construction site, preprocessing the voice, extracting characteristic parameters, establishing a model database of the special vehicle, collecting audio of an engineering vehicle to be identified by using voice sensing on a complex scene of the construction site in a test module, and comparing the audio with the parameters in the model database through the same preprocessing, endpoint detection and characteristic extraction modules to judge whether the voice contains the engineering vehicle or not;
when needing installation device, at first rotating threaded rod 3, threaded rod 3 drives two sliding plates 15 respectively through two U-shaped plate 5 and slides, and two sliding plates 15 drive two through two solid fixed ring 13 respectively and rotate a section of thick bamboo 12, and a section of thick bamboo 12 drives L shaped plate 10 through stopper 14 and rotates to fix the device, when needs dismounting device, the same reason direction is rotated threaded rod 3 and can be accomplished and is dismantled. It should be noted that the electric devices referred to in this application may be powered by a storage battery or an external power source.
The specific extraction process is as follows:
(1) pretreatment of
Sampling and quantizing the collected audio signal, pre-emphasis processing, framing and windowing processing and end point detection processing to obtain a time domain signal x (n).
A. Pre-emphasis is performed. The pre-emphasis is to eliminate the influence of the lips during the sound production process and to flatten the signal spectrum. Meanwhile, the high-frequency part of the voice signal is compensated, so that the subsequent spectrum analysis is facilitated.
y(n)＝s(n)-a*s(n-1)
Wherein a is a pre-emphasis coefficient and has a value range of 0.9-1.0.
B. And (5) framing. Although the audio signal is a non-stationary time-varying signal, the audio signal still has short-time stationarity for a short period of time, and the speech is subjected to framing by utilizing the short-time stationarity. Generally, the duration of 0.02-0.05 ms is set as a frame, and the frame shift is set as a moving step length of 0.01ms, so that the two frames are partially overlapped, and the two frames are prevented from being changed too much.
C. And (5) windowing. In the MFCC feature extraction process, a commonly used window function is a hamming window. Since each frame is then fourier transformed, it is assumed that each frame is a periodic signal. The hamming window function is added, namely, the smoothing processing is carried out on the segment of the continuous signal, and the waveform is similar to a periodic function.
x'(n)＝x(n)*ω(n)
Wherein the content of the first and second substances,
D. and detecting an end point. The end point detection is to detect the start point and the end point of the required speech signal and discard the irrelevant sections which are not required by us. In the voiceprint recognition technology, the end point detection plays a role in being invisible, because the good end point detection technology can improve the time robustness of the whole system and save time. The characteristic parameter extraction and identification stages can be operated more smoothly, and the influence of some irrelevant sections is less, so that the identification rate is improved.
(2) DFT transform
At first, 0 complementing processing is carried out on the signal until N is equal to 512, and then DFT is carried out, and X (k) is obtained as the frequency spectrum of the signal. The expression is as follows:
in this step, the size of the fourier series must be considered, and selecting an appropriate number of stages reduces the amount of calculation and also enables higher accuracy.
(3) Mel filter bank
Experimental observations have shown that the human ear acts like a filter bank, only paying attention to certain frequency components, while reducing or even ignoring the perception of other frequency signals. The Mel frequency analysis is based on human auditory perception experiments, linear frequency spectrums are mapped to non-linear frequency spectrums based on auditory perception by considering the auditory characteristics of human ears, and then the non-linear frequency spectrums are converted to cepstrum. Filter transfer function Hm(k):
(m) is defined as:
fland fhRespectively the lowest and highest frequencies within the Mel filter bank. N is the number of transform points of the discrete cosine transform. M is the number of the triangular filters. B-1 and B are reciprocal functions, and
(4) logarithm of energy removed
Logarithm of e (m) is taken to obtain a logarithmic spectrum s (m):
S(m)＝ln(E(m))，0≤m≤M
(5) discrete Cosine Transform (DCT)
The main purpose of the discrete cosine transform dct (discrete cosine transform) is to make the vectors independent from each other. After discrete cosine transform, the dimensionality of the characteristic parameters is reduced, training can be reduced, and the calculation of the recognition rate is simpler. Performing discrete cosine transform on S (m), wherein the formula is as follows:
the MFCC characteristic parameters are obtained, and a recognition model is needed to be used for building a database for the engineering vehicle, in a voiceprint recognition system, the biggest difference between different pattern matching methods is the method of representing the model and matching audio frequency during testing, and a probability model in a common recognition method is more flexible, and the theoretical significance of likelihood score is more convincing. Our technique uses a gaussian mixture model-a generic background model.
After the engineering vehicle has been identified, the position of the engineering vehicle needs to be determined by sound source localization, where we localize it using a time difference of arrival based localization algorithm. The algorithm is mainly based on the principle that sound signals received by a microphone sensor are utilized to calculate time delay of the sound signals reaching each receiver, the time delay corresponds to the sound path difference between each receiver and a sound source in a three-dimensional space, and then the sound source position coordinates are solved by utilizing a sound path difference and positioning equation resolving method. The main flow of the TDOA-based positioning algorithm can be divided into two steps: firstly, estimating the time difference between the sound signal and each sensor, namely estimating the time delay; in the second step, the coordinates of the sound source, i.e. the sound source localization, are calculated by means of geometrical relations. The TDOA-based two-step location technique can be illustrated by fig. 3 and 4. FIG. 3, TDOA estimation, FIG. 4, TDOA-based sound source localization process;
the generalized cross-correlation method is a conventional time delay estimation method. Since there is some correlation between signals from the same sound source, the TDOA value can be estimated by calculating the correlation function between signals received by different microphones. A flow chart of the generalized cross-correlation function algorithm is shown in fig. 5. FIG. 5 is a flow chart of a generalized cross-correlation function algorithm.
After the TDOA is obtained through the GCC, the engineering vehicle needs to be located by using a geometric location method. The method has small calculation amount and is simple and easy to realize. After the delay estimation step is completed, we know the conditions that can be used for position solution: sound velocity V, microphone array coordinates (x)i,yi,zi) I-1, 2,3,4, and the time delay τ between the microphones12，τ13，τ14. We need to derive the sound source position coordinates by position solution, i.e. (x)s,ys,zs). Taking a four-element microphone as an example, a position resolving formula of a sound source in a 3-dimensional space is as follows: wherein
Through L shaped plate (10) that sets up, when rotating threaded rod (3), threaded rod (3) drive L shaped plate (10) and rotate to fix the device, made things convenient for and installed and dismantle positioner, improved the efficiency of installation with the dismantlement, and U shaped plate (5) of setting have reduced the condition emergence of L shaped plate (10) slope when sliding, thereby have improved the stability of device.
One side of the three pre-processes of this embodiment is connected with the voice input, the voice to be tested, and the audio signal, respectively.
The DFT of this embodiment is connected to the input of the Mel filter bank, which is connected to the input of the logarithmic energy, which is connected to the input of the DCT.
The inside of channel 2 of this embodiment is provided with baffle 7, and connecting plate 8 is installed to one side of baffle 7, and dead lever 9 is installed to one side of connecting plate 8, and dead lever 9 and L shaped plate 10 normal running fit are convenient for rotate L shaped plate 10 through the dead lever 9 that sets up.
The L-shaped plate 10 of this embodiment is provided with a spring 11 at one side thereof, and the spring 11 is connected to the baffle 7.
The rotating plate 6 is installed at the end of the threaded rod 3 of the embodiment, and the threaded rod 3 is convenient to rotate through the rotating plate 6.
The above embodiments may be combined with each other.
It should be noted that the terms "first," "second," and the like in the description and claims of this application and in the drawings described above are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the application described herein are capable of operation in sequences other than those illustrated or described herein.
In the description of the present invention, it should be understood that the orientation or positional relationship indicated by the orientation words such as "front, back, up, down, left, right", "horizontal, vertical, horizontal" and "top, bottom" etc. are usually based on the orientation or positional relationship shown in the drawings, and are only for convenience of description and simplification of description, and in the case of not making a contrary explanation, these orientation words do not indicate and imply that the device or element referred to must have a specific orientation or be constructed and operated in a specific orientation, and therefore, should not be interpreted as limiting the scope of the present invention; the terms "inner and outer" refer to the inner and outer relative to the profile of the respective component itself.
Claims (6)
1. An engineering vehicle discerns positioner based on voiceprint discernment and sound source location, its characterized in that includes: training module, test module, base plate (1), training module, test module all include: preprocessing, wherein the preprocessing is connected with feature extraction, and the two feature extractions are respectively connected with a training model and comparison judgment;
the training model is connected with a model base, the comparison judgment is connected with an output result, the model base is connected with the comparison judgment through mode matching, and the preprocessing is connected with the input end of the DFT;
channel (2) have been seted up to the inside of base plate (1), and the inside of channel (2) is provided with two fixed subassemblies, and the inside normal running fit threaded rod (3) of channel (2), the week side screw-thread fit of threaded rod (3) have two U-shaped boards (5), and fixed subassembly includes: an L-shaped plate (10) and a rotary cylinder (12);
a fixed ring (13) is installed on the periphery of the rotating cylinder (12), sliding plates (15) are installed on the outer side of the fixed ring (13), the two sliding plates (15) are respectively in sliding fit in the two U-shaped plates (5), and a limiting block (14) is installed on the periphery of the upper end of the rotating cylinder (12).
2. The engineering vehicle recognition and positioning device based on voiceprint recognition and sound source positioning as claimed in claim 1, wherein the preprocessed side is connected with a voice input, a voice to be tested and an audio signal respectively.
3. The identification and location device for engineering vehicles based on voiceprint recognition and sound source location as claimed in claim 1, wherein DFT is connected with input terminal of Mel filter bank, Mel filter bank is connected with input terminal of logarithmic energy, and logarithmic energy is connected with input terminal of DCT.
4. The identification and positioning device for engineering vehicles based on voiceprint identification and sound source positioning as claimed in claim 1, wherein a baffle (7) is arranged inside the channel (2), a connecting plate (8) is arranged on one side of the baffle (7), a fixing rod (9) is arranged on one side of the connecting plate (8), and the fixing rod (9) is rotatably matched with the L-shaped plate (10).
5. An identification and positioning device for engineering vehicles based on voiceprint identification and sound source positioning as claimed in claim 1, wherein a spring (11) is installed on one side of the L-shaped plate (10), and the spring (11) is connected with the baffle (7).
6. A recognition and positioning device for engineering vehicles based on voiceprint recognition and sound source positioning as claimed in claim 5, wherein the end of the threaded rod (3) is mounted with a rotating plate (6).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202022756413.6U CN214796752U (en) | 2020-11-25 | 2020-11-25 | Engineering vehicle identification and positioning device based on voiceprint identification and sound source positioning |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202022756413.6U CN214796752U (en) | 2020-11-25 | 2020-11-25 | Engineering vehicle identification and positioning device based on voiceprint identification and sound source positioning |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN214796752U true CN214796752U (en) | 2021-11-19 |
Family
ID=78709650
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202022756413.6U Active CN214796752U (en) | 2020-11-25 | 2020-11-25 | Engineering vehicle identification and positioning device based on voiceprint identification and sound source positioning |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN214796752U (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116092484A (en) * | 2023-04-07 | 2023-05-09 | 四川高速公路建设开发集团有限公司 | Signal detection method and system based on distributed optical fiber sensing in high-interference environment |
-
2020
- 2020-11-25 CN CN202022756413.6U patent/CN214796752U/en active Active
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116092484A (en) * | 2023-04-07 | 2023-05-09 | 四川高速公路建设开发集团有限公司 | Signal detection method and system based on distributed optical fiber sensing in high-interference environment |
| CN116092484B (en) * | 2023-04-07 | 2023-06-09 | 四川高速公路建设开发集团有限公司 | Signal detection method and system based on distributed optical fiber sensing in high-interference environment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109357749B (en) | DNN algorithm-based power equipment audio signal analysis method | |
| CN109036382B (en) | Audio feature extraction method based on KL divergence | |
| CN107644650B (en) | Improved sound source positioning method based on progressive serial orthogonalization blind source separation algorithm and implementation system thereof | |
| CN108169639B (en) | Method for identifying switch cabinet fault based on parallel long-time and short-time memory neural network | |
| CN107610715B (en) | Similarity calculation method based on multiple sound characteristics | |
| CN108896878B (en) | Partial discharge detection method based on ultrasonic waves | |
| US6792118B2 (en) | Computation of multi-sensor time delays | |
| CN109949823B (en) | DWPT-MFCC and GMM-based in-vehicle abnormal sound identification method | |
| CN102565759B (en) | Binaural sound source localization method based on sub-band signal to noise ratio estimation | |
| CN107170465B (en) | Audio quality detection method and audio quality detection system | |
| CN111429939B (en) | Sound signal separation method of double sound sources and pickup | |
| CN103646649A (en) | High-efficiency voice detecting method | |
| CN101494049A (en) | Method for extracting audio characteristic parameter of audio monitoring system | |
| CN102426835A (en) | Method for identifying local discharge signals of switchboard based on support vector machine model | |
| CN103278801A (en) | Noise imaging detection device and detection calculation method for transformer substation | |
| CN109256139A (en) | A kind of method for distinguishing speek person based on Triplet-Loss | |
| CN111540347A (en) | Cable tunnel monitoring method and system based on audio | |
| KR20210077389A (en) | Deep learning based abnormal symptom detection system using mechanical sound of mechanical equipment | |
| CN105118511A (en) | Thunder identification method | |
| CN112542174A (en) | VAD-based multi-dimensional characteristic parameter voiceprint identification method | |
| CN214796752U (en) | Engineering vehicle identification and positioning device based on voiceprint identification and sound source positioning | |
| CN105825857A (en) | Voiceprint-recognition-based method for assisting deaf patient in determining sound type | |
| CN116229991A (en) | Motor fault diagnosis method based on MFCC voice feature extraction and machine learning | |
| CN116935894B (en) | Micro-motor abnormal sound identification method and system based on time-frequency domain mutation characteristics | |
| CN111429916B (en) | Sound signal recording system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Assignee: HYLIGHT TECHNOLOGY CO.,LTD. Assignor: ZHONGKE WEIBO (SUZHOU) INTELLIGENT TECHNOLOGY Co.,Ltd. Contract record no.: X2023310000095 Denomination of utility model: A Construction Vehicle Identification and Location Device Based on Voiceprint Recognition and Sound Source Localization Granted publication date: 20211119 License type: Common License Record date: 20230615 |
|
| EE01 | Entry into force of recordation of patent licensing contract |