CN111507418A - Encaustic tile quality detection method - Google Patents
Encaustic tile quality detection method Download PDFInfo
- Publication number
- CN111507418A CN111507418A CN202010316550.XA CN202010316550A CN111507418A CN 111507418 A CN111507418 A CN 111507418A CN 202010316550 A CN202010316550 A CN 202010316550A CN 111507418 A CN111507418 A CN 111507418A
- Authority
- CN
- China
- Prior art keywords
- encaustic
- sound
- encaustic tile
- tile
- real
- 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.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/28—Determining representative reference patterns, e.g. by averaging or distorting; Generating dictionaries
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/14—Fourier, Walsh or analogous domain transformations, e.g. Laplace, Hilbert, Karhunen-Loeve, transforms
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Mathematical Optimization (AREA)
- Mathematical Analysis (AREA)
- Computational Mathematics (AREA)
- Pure & Applied Mathematics (AREA)
- Algebra (AREA)
- Databases & Information Systems (AREA)
- Software Systems (AREA)
- Spectrometry And Color Measurement (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The invention discloses a method for detecting the quality of encaustic tiles, which comprises the following steps: recording the knocked sound of the complete encaustic tile, the damaged encaustic tile and the crack encaustic tile respectively, and constructing three sound signal training sets of the complete encaustic tile, the damaged encaustic tile and the crack encaustic tile; in the training stage, short-time Fourier transform is respectively carried out on three sound signal training sets, corresponding real parts and imaginary parts are extracted, dictionaries of the real parts and the imaginary parts are jointly learned based on a K-SVD algorithm, and therefore three dictionaries are obtained; and in the detection stage, recording the knocked sound of the encaustic tiles to be detected, respectively projecting the sound on the three dictionaries, reconstructing corresponding sound signals according to each sparse representation coefficient obtained by projection, and judging the category of the encaustic tiles to be detected according to the reconstruction error. The method not only utilizes the amplitude information of the sound signal frequency spectrum, but also utilizes the phase information of the signal, thereby being capable of keeping better detection capability, and the training data is easy to obtain, thereby being convenient for the industry to realize.
Description
Technical Field
The invention relates to the technical field of encaustic tile quality detection, in particular to encaustic tile quality detection technology based on dictionary learning and sparse representation.
Background
At present, the encaustic tiles become an indispensable building material in our lives, and the encaustic tiles are needed in engineering project buildings, commercial houses, rural self-building villas and self-building houses, and small pavilions and small wooden houses. The production mode of the tile is mostly industrialized flow production, and the defective tile, including the damaged encaustic tile, the cracked encaustic tile and the like, is inevitably produced in the production process. How to detect the defective tile quickly and accurately has become a problem to be solved urgently in the industry.
The existing encaustic tile quality detection technology is mainly divided into two categories, namely manual detection and automatic detection. The manual detection is mostly carried out through naked eye judgment of workers or the flawed encaustic tiles are screened out according to some physical characteristics, the method is low in efficiency, and the human resource waste is large, so that the industrial industry mostly adopts an automatic detection method nowadays. In the automatic detection method, whether the encaustic tile has flaws is mostly judged based on visual information, the basic method is to adopt a camera to collect a picture of a complete encaustic tile and a picture of a detected encaustic tile, then corresponding pixel points of the two pictures are compared, whether the difference value is larger than a preset threshold range is judged, if yes, the encaustic tile is obtained, and if not, the encaustic tile is judged to be the flaw encaustic tile, but the method has higher requirements on the quality of the collected pictures and has certain requirements on equipment such as the camera; therefore, the equipment involved is costly.
Disclosure of Invention
The invention aims to provide a method for detecting the quality of a encaustic tile, which is used for detecting the quality of the encaustic tile by mining the knocked sound information of the encaustic tile and utilizing the sparsity of sound signals, and has the advantages of higher detection precision and lower cost.
The purpose of the invention is realized by the following technical scheme:
a encaustic tile quality detection method comprises the following steps:
recording the knocked sound of the complete encaustic tile, the damaged encaustic tile and the crack encaustic tile respectively, and constructing three sound signal training sets of the complete encaustic tile, the damaged encaustic tile and the crack encaustic tile;
in the training stage, short-time Fourier transform is respectively carried out on three sound signal training sets, corresponding real parts and imaginary parts are extracted, dictionaries of the real parts and the imaginary parts are jointly learned based on a K-SVD algorithm, and therefore three dictionaries are obtained;
and in the detection stage, recording the knocked sound of the encaustic tiles to be detected, respectively projecting the sound on the three dictionaries, reconstructing corresponding sound signals according to each sparse representation coefficient obtained by projection, and judging the category of the encaustic tiles to be detected according to the reconstruction error.
According to the technical scheme provided by the invention, the invention not only utilizes the amplitude information of the sound signal frequency spectrum, but also utilizes the phase information of the signal, so that the better detection capability can be kept, and the training data is easy to obtain and is convenient to realize in the industry.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
Fig. 1 is a flowchart of a method for detecting encaustic tile quality according to an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention are clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments of the present invention without making any creative effort, shall fall within the protection scope of the present invention.
Different from the traditional detection method based on visual information, the embodiment of the invention provides a method for detecting the quality of a encaustic tile, which detects the quality of the encaustic tile by mining the sound information of the encaustic tile being knocked and utilizing the sparsity of sound signals, as shown in fig. 1, and the method mainly comprises the following steps:
Illustratively, the full, broken and cracked encaustic tiles are struck with the same tool for a period of time, e.g., about 10s to about 20s, and the sounds of the striking are recorded by a microphone, respectivelyAndand constructing three sound signal training sets of complete encaustic tiles, damaged encaustic tiles and cracked encaustic tiles.
In the embodiment of the invention, the damaged encaustic tiles refer to the existing missing parts of encaustic tiles, and the cracked encaustic tiles refer to the existing cracks of encaustic tiles but do not exist.
And 2, in a training stage, respectively carrying out short-time Fourier transform on three sound signal training sets, extracting corresponding real parts and imaginary parts, and jointly learning dictionaries of the real parts and the imaginary parts based on a K-SVD algorithm so as to obtain three dictionaries, wherein the three dictionaries are obtained, and not only amplitude spectrum information of the signals but also phase information of the signals are learned.
The preferred embodiment of the training phase is as follows:
training stage, the sound of knocking the complete encaustic tiles, the damaged encaustic tiles and the cracked encaustic tilesAndobtaining sound complex spectrum of complete encaustic tiles on time-frequency domain through short-time Fourier transformSound complex spectrum of damaged encaustic tileAnd crackle colorTile sound complex spectrumLearning the real part and the imaginary part of the three sound complex spectrums respectively based on a K-SVD algorithm to obtain corresponding dictionariesAnd
wherein the content of the first and second substances,andare respectivelyReal and imaginary parts of, Dr1And Di1Real and imaginary dictionaries, respectively;andare respectivelyReal and imaginary parts of, Dr2And Di2Real and imaginary dictionaries, respectively;andare respectivelyReal and imaginary parts of, Dr3And Di3Real and imaginary dictionaries, respectively; c1Is a sparse representation coefficient, C, of the sound signal for a complete encaustic tile2Is a sparse representation coefficient, C, of the sound signal of a broken encaustic tile3Is a sparse representation coefficient of the sound signal for the crack encaustic tile; sparse representation coefficients of the sound signal in matrix form, c1,g、c2,gAnd c3,gAre respectively C1、C2And C3Q is a sparse constraint;represented by the Frobenius norm, | · | |. luminance1Representative is the 1-norm.
And 3, in the detection stage, recording the knocked sound of the encaustic tiles to be detected, respectively projecting the sound on the three dictionaries, reconstructing corresponding sound signals according to each sparse representation coefficient obtained by projection, and judging the category of the encaustic tiles to be detected according to the reconstruction error.
The preferred embodiment of the detection stage is as follows:
in the detection stage, the sound of the knocked encaustic tile to be detected is subjected to short-time Fourier transform to calculate a corresponding complex spectrum SteGet its real partAnd imaginary partRespectively projected to three dictionariesAndand calculating sparse representation coefficients respectively:
wherein D isr1And Di1Dictionaries, D, representing the real and imaginary parts of the complex spectrum of a sound in which a complete encaustic tile has been struckr2And Di2Dictionaries, D, representing real and imaginary parts of complex spectrum of sound with broken tiles struckr3And Di3Dictionaries respectively representing the real and imaginary parts of the complex spectrum of a sound struck by a crackle colortile, E1、E2And E3Respectively candidate sparse representation coefficients projected onto three dictionaries,andrespectively selecting optimal sparse representation coefficients from the candidate sparse representation coefficients by the formula; sparse representation of coefficients in matrix form, e1,g、e2,gAnd e3,gAre respectively E1、E2And E3Q is a sparse constraint;
and calculating the real part and the imaginary part of the signal complex spectrum reconstructed by each dictionary by using the sparse representation coefficient, as shown in the following formula:
wherein the content of the first and second substances,andseparately representing using sparse representation coefficientsAnd dictionaryReal and imaginary parts of a reconstructed signal complex spectrum;andseparately representing using sparse representation coefficientsAnd dictionaryReal and imaginary parts of a reconstructed signal complex spectrum;andseparately representing using sparse representation coefficientsChinese character' HeDian (Chinese character)Real and imaginary parts of a reconstructed signal complex spectrum;
calculating the reconstruction error of the sound signal:
judging the type of the encaustic tile to be detected according to the size of the reconstruction error, if the reconstruction error is large1Minimum, sound dictionary indicating complete encaustic tileThe reconstruction error of the sound signal of the detected encaustic tile is minimum, and the sparse representation accuracy is highest, so that the encaustic tile to be detected is judged to belong to the complete encaustic tile class; if the reconstruction error is2Sound dictionary for indicating damaged encaustic tile when minimumThe reconstruction error of the sound signal of the detected encaustic tile is minimum, and the sparse representation accuracy is highest, so that the encaustic tile to be detected is judged to belong to the damaged encaustic tile; if the reconstruction error is3Sound dictionary of encaustic tile indicating crack if minimumThe reconstruction error of the sound signal of the detected encaustic tile is minimum, and the sparse representation accuracy is highest, so that the encaustic tile to be detected belongs to the category of crack encaustic tiles.
According to the scheme of the embodiment of the invention, a training method of combined dictionary learning is adopted firstly, and a real part dictionary and an imaginary part dictionary of a knocked sound signal of three types of encaustic tiles, namely complete encaustic tiles, damaged encaustic tiles and cracked encaustic tiles are learned according to the environment used by the method. In the detection stage, the sound signal of the encaustic tile to be detected is subjected to short-time Fourier transform to obtain a corresponding real part and an imaginary part, the real part and the imaginary part are respectively projected to three types of dictionaries to reconstruct the real part and the imaginary part of the signal, and the type of the encaustic tile to be detected is judged according to the reconstructed error. Therefore, the method not only utilizes the amplitude information of the sound signal, but also utilizes the phase information of the signal, thereby maintaining better detection capability, and the training data is easy to obtain and convenient to realize in the industry.
Through the above description of the embodiments, it is clear to those skilled in the art that the above embodiments can be implemented by software, and can also be implemented by software plus a necessary general hardware platform. With this understanding, the technical solutions of the embodiments can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a usb disk, a removable hard disk, etc.), and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods according to the embodiments of the present invention.
The above description is only for the preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present invention are included in the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (3)
1. A encaustic tile quality detection method is characterized by comprising the following steps:
recording the knocked sound of the complete encaustic tile, the damaged encaustic tile and the crack encaustic tile respectively, and constructing three sound signal training sets of the complete encaustic tile, the damaged encaustic tile and the crack encaustic tile;
in the training stage, short-time Fourier transform is respectively carried out on three sound signal training sets, corresponding real parts and imaginary parts are extracted, dictionaries of the real parts and the imaginary parts are jointly learned based on a K-SVD algorithm, and therefore three dictionaries are obtained;
and in the detection stage, recording the knocked sound of the encaustic tiles to be detected, respectively projecting the sound on the three dictionaries, reconstructing corresponding sound signals according to each sparse representation coefficient obtained by projection, and judging the category of the encaustic tiles to be detected according to the reconstruction error.
2. The encaustic tile quality detection method according to claim 1,
training stage, the sound of knocking the complete encaustic tiles, the damaged encaustic tiles and the cracked encaustic tilesAndobtaining sound complex spectrum of complete encaustic tiles on time-frequency domain through short-time Fourier transformSound complex spectrum of damaged encaustic tileAnd crackle colour tile sound complex spectrumLearning the real part and the imaginary part of the three sound complex spectrums respectively based on a K-SVD algorithm to obtain corresponding dictionariesAnd
wherein the content of the first and second substances,andare respectivelyReal and imaginary parts of, Dr1And Di1Real and imaginary dictionaries, respectively;andare respectivelyReal and imaginary parts of, Dr2And Di2Real and imaginary dictionaries, respectively;andare respectivelyReal and imaginary parts of, Dr3And Di3Are respectively the real part andan imaginary part dictionary; c1Is a sparse representation coefficient, C, of the sound signal for a complete encaustic tile2Is a sparse representation coefficient, C, of the sound signal of a broken encaustic tile3Is a sparse representation coefficient of the sound signal for the crack encaustic tile; sparse representation coefficients of the sound signal in matrix form, c1,g、c2,gAnd c3,gAre respectively C1、C2And C3Q is a sparse constraint;represented by the Frobenius norm, | · | |. luminance1Representative is the 1-norm.
3. The encaustic tile quality detection method according to claim 1 or 2,
in the detection stage, the sound of the knocked encaustic tile to be detected is subjected to short-time Fourier transform to calculate a corresponding complex spectrum SteGet its real partAnd imaginary partRespectively projected to three dictionariesAndand calculating sparse representation coefficients respectively:
wherein D isr1And Di1Dictionaries, D, representing the real and imaginary parts of the complex spectrum of a sound in which a complete encaustic tile has been struckr2And Di2Dictionaries, D, representing real and imaginary parts of complex spectrum of sound with broken tiles struckr3And Di3Dictionaries respectively representing the real and imaginary parts of the complex spectrum of a sound struck by a crackle colortile, E1、E2And E3Respectively candidate sparse representation coefficients projected onto three dictionaries,andrespectively selecting optimal sparse representation coefficients from the candidate sparse representation coefficients by the formula; sparse representation of coefficients in matrix form, e1,g、e2,gAnd e3,gAre respectively E1、E2And E3Q is a sparse constraint;
and calculating the real part and the imaginary part of the signal complex spectrum reconstructed by each dictionary by using the sparse representation coefficient, as shown in the following formula:
wherein the content of the first and second substances,andseparately representing using sparse representation coefficientsAnd dictionaryReal and imaginary parts of a reconstructed signal complex spectrum;andseparately representing using sparse representation coefficientsAnd dictionaryReal and imaginary parts of a reconstructed signal complex spectrum;andseparately representing using sparse representation coefficientsAnd dictionaryReal and imaginary parts of a reconstructed signal complex spectrum;
calculating the reconstruction error of the sound signal:
judging the type of the encaustic tile to be detected according to the size of the reconstruction error, if the reconstruction error is large1Minimum, sound dictionary indicating complete encaustic tileThe reconstruction error of the sound signal of the detected encaustic tile is minimum, and the sparse representation accuracy is highest, so that the encaustic tile to be detected is judged to belong to the complete encaustic tile class; if the reconstruction error is2Sound dictionary for indicating damaged encaustic tile when minimumThe reconstruction error of the sound signal of the detected encaustic tile is minimum, and the sparse representation accuracy is highest, so that the encaustic tile to be detected is judged to belong to the damaged encaustic tile; if the reconstruction error is3Sound dictionary of encaustic tile indicating crack if minimumThe reconstruction error of the sound signal of the detected encaustic tile is minimum, and the sparse representation accuracy is highest, so that the encaustic tile to be detected belongs to the category of crack encaustic tiles.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010316550.XA CN111507418B (en) | 2020-04-21 | 2020-04-21 | Encaustic tile quality detection method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010316550.XA CN111507418B (en) | 2020-04-21 | 2020-04-21 | Encaustic tile quality detection method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN111507418A true CN111507418A (en) | 2020-08-07 |
CN111507418B CN111507418B (en) | 2022-09-06 |
Family
ID=71869721
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010316550.XA Active CN111507418B (en) | 2020-04-21 | 2020-04-21 | Encaustic tile quality detection method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN111507418B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112185405A (en) * | 2020-09-10 | 2021-01-05 | 中国科学技术大学 | Bone conduction speech enhancement method based on differential operation and joint dictionary learning |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103345920A (en) * | 2013-05-29 | 2013-10-09 | 河海大学常州校区 | Self-adaptation interpolation weighted spectrum model voice conversion and reconstructing method based on Mel-KSVD sparse representation |
CN106709505A (en) * | 2016-11-23 | 2017-05-24 | 中国农业大学 | Dictionary learning-based grain type classification method and system for corn grains |
CN106841403A (en) * | 2017-01-23 | 2017-06-13 | 天津大学 | A kind of acoustics glass defect detection method based on neutral net |
US20200035256A1 (en) * | 2019-08-30 | 2020-01-30 | Lg Electronics Inc. | Artificial sound source separation method and device of thereof |
-
2020
- 2020-04-21 CN CN202010316550.XA patent/CN111507418B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103345920A (en) * | 2013-05-29 | 2013-10-09 | 河海大学常州校区 | Self-adaptation interpolation weighted spectrum model voice conversion and reconstructing method based on Mel-KSVD sparse representation |
CN106709505A (en) * | 2016-11-23 | 2017-05-24 | 中国农业大学 | Dictionary learning-based grain type classification method and system for corn grains |
CN106841403A (en) * | 2017-01-23 | 2017-06-13 | 天津大学 | A kind of acoustics glass defect detection method based on neutral net |
US20200035256A1 (en) * | 2019-08-30 | 2020-01-30 | Lg Electronics Inc. | Artificial sound source separation method and device of thereof |
Non-Patent Citations (2)
Title |
---|
YUANYUAN ZHU ET AL.: "Modified Complementary Joint Sparse Representations: A Novel Post-Filtering to MVDR Beamforming", 《2019 IEEE INTERNATIONAL WORKSHOP ON SIGNAL PROCESSING SYSTEMS (SIPS)》 * |
叶中付 等: "基于字典学习和稀疏表示的单通道语音增强算法综述", 《应用声学》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112185405A (en) * | 2020-09-10 | 2021-01-05 | 中国科学技术大学 | Bone conduction speech enhancement method based on differential operation and joint dictionary learning |
CN112185405B (en) * | 2020-09-10 | 2024-02-09 | 中国科学技术大学 | Bone conduction voice enhancement method based on differential operation and combined dictionary learning |
Also Published As
Publication number | Publication date |
---|---|
CN111507418B (en) | 2022-09-06 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Fazenda et al. | Cave acoustics in prehistory: Exploring the association of Palaeolithic visual motifs and acoustic response | |
Li et al. | Rhythmic brushstrokes distinguish van Gogh from his contemporaries: findings via automated brushstroke extraction | |
CN110880019B (en) | Method for adaptively training target domain classification model through unsupervised domain | |
US20070220265A1 (en) | Searching for a scaling factor for watermark detection | |
CN113758932B (en) | Deep learning-based visual detection method for defects of lithium battery diaphragm | |
CN108304328B (en) | Text description generation method, system and device for crowdsourcing test report | |
CN111507418B (en) | Encaustic tile quality detection method | |
CN111767874B (en) | Pavement disease detection method based on deep learning | |
WO2020015411A1 (en) | Method and device for training adaptation level evaluation model, and method and device for evaluating adaptation level | |
CN110781333A (en) | Method for processing unstructured monitoring data of cable-stayed bridge based on machine learning | |
CN107274912B (en) | Method for identifying equipment source of mobile phone recording | |
CN115471466A (en) | Steel surface defect detection method and system based on artificial intelligence | |
CN116109964A (en) | Intelligent extraction method and device for video map, storage medium and computer equipment | |
CN113283467B (en) | Weak supervision picture classification method based on average loss and category-by-category selection | |
CN117576632A (en) | Multi-mode AI large model-based power grid monitoring fire early warning system and method | |
CN109829887B (en) | Image quality evaluation method based on deep neural network | |
CN111027946A (en) | Bridge detection report generation method and system | |
CN108766465B (en) | Digital audio tampering blind detection method based on ENF general background model | |
CN110910954A (en) | Method and system for detecting low-depth whole genome gene copy number variation | |
Vijayan et al. | A survey on surface crack detection in concretes using traditional, image processing, machine learning, and deep learning techniques | |
CN111693812A (en) | Large transformer fault detection method based on sound characteristics | |
CN110751867B (en) | English teaching system | |
CN101676992A (en) | Method for analyzing echo steganography | |
Rahdari et al. | An ensemble learning model for single-ended speech quality assessment using multiple-level signal decomposition method | |
CN111353550A (en) | Method for detecting surface defects of structure based on random structure forest |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |