CN113010038B - Ultrasonic lamb wave touch load identification method based on super-resolution reconstruction - Google Patents
Ultrasonic lamb wave touch load identification method based on super-resolution reconstruction Download PDFInfo
- Publication number
- CN113010038B CN113010038B CN202110179053.4A CN202110179053A CN113010038B CN 113010038 B CN113010038 B CN 113010038B CN 202110179053 A CN202110179053 A CN 202110179053A CN 113010038 B CN113010038 B CN 113010038B
- Authority
- CN
- China
- Prior art keywords
- touch
- resolution
- acoustic fingerprint
- lamb wave
- super
- 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
- 235000019687 Lamb Nutrition 0.000 title claims abstract description 32
- 238000000034 method Methods 0.000 title claims abstract description 29
- 238000005070 sampling Methods 0.000 claims description 9
- 238000005457 optimization Methods 0.000 claims description 5
- 230000005284 excitation Effects 0.000 claims description 3
- 238000010606 normalization Methods 0.000 claims description 3
- 238000010586 diagram Methods 0.000 description 4
- 230000007547 defect Effects 0.000 description 3
- 230000003044 adaptive effect Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 238000002360 preparation method Methods 0.000 description 2
- 229910000831 Steel Inorganic materials 0.000 description 1
- 239000000969 carrier Substances 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 239000003344 environmental pollutant Substances 0.000 description 1
- 239000011521 glass Substances 0.000 description 1
- 231100000719 pollutant Toxicity 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 239000010959 steel Substances 0.000 description 1
- 238000010897 surface acoustic wave method Methods 0.000 description 1
- 239000002023 wood Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
- G06F3/03—Arrangements for converting the position or the displacement of a member into a coded form
- G06F3/041—Digitisers, e.g. for touch screens or touch pads, characterised by the transducing means
- G06F3/0416—Control or interface arrangements specially adapted for digitisers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/01—Input arrangements or combined input and output arrangements for interaction between user and computer
- G06F3/03—Arrangements for converting the position or the displacement of a member into a coded form
- G06F3/041—Digitisers, e.g. for touch screens or touch pads, characterised by the transducing means
- G06F3/043—Digitisers, e.g. for touch screens or touch pads, characterised by the transducing means using propagating acoustic waves
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2203/00—Indexing scheme relating to G06F3/00 - G06F3/048
- G06F2203/041—Indexing scheme relating to G06F3/041 - G06F3/045
- G06F2203/04105—Pressure sensors for measuring the pressure or force exerted on the touch surface without providing the touch position
Abstract
The invention discloses an ultrasonic lamb wave touch load identification method based on super-resolution reconstruction, which utilizes a touch load lamb wave acoustic fingerprint library constructed under a low-pixel condition to obtain a high-pixel touch load lamb wave acoustic fingerprint library through expansion by the super-resolution reconstruction method. And matching the signal characteristics to be positioned with a touch load lamb wave acoustic fingerprint library by adopting a least square model of a multi-sensor pair with a weight coefficient iteration self-adaption, and finishing touch load positioning and touch force identification. The method provided by the invention can effectively improve the performance of the ultrasonic lamb wave touch screen, and provides a new thought for constructing an acoustic fingerprint library of the ultrasonic lamb wave touch screen and identifying touch load parameters.
Description
Technical Field
The invention relates to the technical field of man-machine interaction, in particular to an ultrasonic lamb wave touch load identification method based on super-resolution reconstruction in an ultrasonic lamb wave touch system.
Background
Touch screens, which are a portable human-computer interaction device, have been widely used in consumer electronics, industrial control, education, medical treatment, public services, and the like. Compared with the traditional touch screens such as resistance type and capacitance type, the acoustic wave type touch screen has two outstanding advantages, the touch force can be directly identified while positioning without an additional force sensor, and in addition, as acoustic waves can be transmitted in various solid media (glass, steel plates, wood plates and the like), desks, windows and the like in daily life can become carriers of the acoustic wave type touch screen.
Acoustic touch screens can be further classified into surface wave type and lamb wave type. Compared with a surface acoustic wave touch screen, the ultrasonic lamb wave type touch screen is insensitive to pollutants such as scratches on the surface of the screen board, and has potential in multi-point touch recognition. The existing ultrasonic lamb wave touch screen technology still has the following defects: (1) Only the touch position is identified by calibrating an acoustic fingerprint library corresponding to the touch area, so that the load is difficult to identify; (2) The load positioning accuracy can be improved by adopting a model with multiple sensor pairs, but the weight coefficient of each sensor pair in the identification model needs to be manually set, so that automatic optimization cannot be realized; (3) The preparation of the high-resolution acoustic fingerprint library consumes long time, and high-precision load positioning and identification are difficult to realize by directly utilizing the low-resolution acoustic fingerprint library.
Aiming at the defects, the invention discloses an ultrasonic lamb wave touch load identification method based on super-resolution reconstruction, which can effectively solve the defects in the existing ultrasonic lamb wave touch technology by utilizing a multi-sensor pair identification model with super-resolution reconstruction and weight coefficient self-adaption of a low-resolution acoustic fingerprint library.
Disclosure of Invention
The invention aims to solve the problem of realizing the identification of the touch position and the touch force by utilizing low-resolution acoustic fingerprint database data in an ultrasonic lamb wave touch screen, and in order to realize the aim, the embodiment of the invention provides an ultrasonic lamb wave touch load identification method based on super-resolution reconstruction, which mainly comprises the following steps,
1) The medium of the touch sensing screen board is subjected to equal-size grid division to form N x ×N y N calibration touch areas, using a load touch (x, y) calibration area with force F and area S, the sensor pair T is noted i-j (i excitation sensor, j receiving sensor) the touch load lamb wave acoustic fingerprint signal acquired is d F,S,i,j (x, y), traversing the N calibration areas of touch control to construct an acoustic fingerprint library D with the resolution of S F,S,i,j ={d F,S,i,j (x,y)}(x=1,2,…,N x ;y=1,2,…,N y )。
2) Acoustic fingerprint signal d F,S,i,j (x, y) extracting fingerprint library D for uniform sampling waveform with M sampling points F,S,i,j The amplitude a of the M (m=1, 2, …, M) th sample point of all acoustic fingerprint signals m (x, y) to form a set of pixel values G m ={A m (x,y)}(x=1,2,…,N x ;y=1,2,…,N y )。
3) Find set G m Maximum value MAXA in (a) m And minimum MINA m Using maximum and minimum pairs G m The process is carried out in the order of [0 ],1]normalizing, converting normalized amplitude into N x ×N y And (3) an image Il of each pixel point, wherein the coordinates of each pixel point in the image are consistent with the coordinates of the corresponding calibration area.
4) Super-resolution reconstruction is carried out on the image Il by using a bicubic spline interpolation method to obtain a high-resolution image Ih, and the number of pixel points is recorded and amplified to N '' x ×N' x =n 'number, equivalent to the sense panel being divided into N' x ×N' x N' new calibration touch areas, using MAXA m And MINA m Performing inverse normalization on Ih to obtain a new pixel value set G' m ={A' m (x',y')}(x’=1,2,…,N x ’;y’=1,2,…,N y ’)。
5) A 'is marked according to the sequence of the sampling points m and the coordinates (x', y ') of the touch area to be newly marked' m (x ', y') to be integrated into a new acoustic fingerprint signal d F,S',i,j (x ', y') to obtain an acoustic fingerprint library D at high resolution F,S',i,j In which the new resolution
6) Will N F Repeating the steps 2) to 5) of the low-resolution acoustic fingerprint library prepared under the individual touch control force to obtain an acoustic fingerprint library D under the high resolution F,S',i,j (F=1,2,…,N F ) Unknown touch is controlled by using the following methodThe preliminary solution is carried out so as to obtain,
in theta F ={θ F (x',y')}(x’=1,2,…,N x ’;y’=1,2,…,N y ') characterization of unknown touchPositioning result in high-resolution acoustic fingerprint library with touch force F, ++>To satisfy the above formula theta F Is a preliminary solution to (a).
7) The preliminary solutions are substituted into a multi-sensor-pair least squares model for which the following weight coefficients are iteratively adaptive,
s.t.θ F (x',y')≥0,for all x',y' (1b)
w in F,i,j For sensor pair T i-j And (3) a weight coefficient in a high-resolution acoustic fingerprint library with the touch force of F, wherein mu is a penalty coefficient.
8) Solving the recognition model in the step 7), and realizing automatic optimization of the weight coefficient by the following steps:
8a) Setting initial weight (w F,i,j ) 0 ;
8b) Substituting the initial weight into 7) the model to obtain an initial solution (Θ) F ) 0 ;
8c) Calculating the sum of squares of the residualsAnd update the weight coefficient
8d) Taking the new weight coefficient as the initial weight and repeating the steps 8 b) and 8 c) until the iteration is finished, so as to obtain the optimal weight and the final result theta F 。
9) Obtaining unknown touch acoustic fingerprintsPositioning result theta under each touch force F (F=1,2,…,N F ) Using all theta F Maximum and minimum value pair Θ in (x ', y') F Proceed [0,1 ]]Normalized, θ exceeding a set threshold ε (0 < ε.ltoreq.1) F The corresponding calibration touch area (x ', y') and touch force F are +.>Is provided.
Compared with the prior art, the invention has the following advantages: (1) The method is combined with an image super-resolution reconstruction method, so that the conversion of the acoustic fingerprint dictionary library from low resolution to high resolution is completed, and the preparation efficiency of the high-precision dictionary library is improved; (2) The automatic optimization of the multiple sensors on the weight coefficient in the least square positioning model is realized by utilizing the residual square sum and in an iterative updating mode; (3) By using a multi-parameter dictionary library, the touch load is positioned and the touch force is recognized.
Drawings
Fig. 1 is a schematic flow chart of an ultrasonic lamb wave touch load identification method.
Fig. 2 is a schematic diagram of demarcation of calibration areas before and after super-resolution reconstruction.
FIG. 3 is a schematic diagram of a super resolution reconstruction process of an acoustic fingerprint library.
Fig. 4 is a schematic diagram of touch to be recognized.
FIG. 5 is a schematic diagram of the touch position and touch force recognition result.
In the figure: 1-excitation sensor, 2-receiving sensor 1, 3-receiving sensor 2, 4-receiving sensor 3
Detailed Description
The invention is further described below with reference to the drawings and examples.
With reference to the flow chart shown in fig. 1, the specific implementation method of touch positioning and recognition in the embodiment of the invention is as follows:
1) The touch screen board is subjected to equal-size grid division, and the grid size is 64mm 2 As shown before super-resolution reconstruction in fig. 2,3×8=24 calibration touch areas are formed in the x and y directions, and excitation sensors 1 and receiving sensors 1,2,3 are arranged around the screen board. UsingArea s=64 mm 2 The load with the touch force of 0.4N touches each calibration area, and the sensor pair T is recorded i-j (i=1, j=1, 2, 3) acquired lamb wave acoustic fingerprint signal d F,S,i,j (x, y), traversing 24 calibration areas of touch control to construct the touch control with the resolution of 64mm 2 Acoustic fingerprint library D of (a) F,S,i,j ={d F,S,i,j (x,y)}(x=1,2,3;y=1,2,…,8)。
2) Acoustic fingerprint signal d F,S,i,j (x, y) is a uniform time domain waveform having m=5000 sampling points, as shown in fig. 3 (a). Extracting fingerprint database D F,S,i,j The amplitude a of the m (m=1, 2, …, 5000) th sample point of all acoustic fingerprint signals m (x, y) constituting a set G of pixel values as shown in FIG. 3 (b) m ={A m (x,y)}。
3) Find set G m Maximum value MAXA in (a) m And minimum MINA m Using maximum and minimum pairs G m Proceed [0,1 ]]Normalizing, namely converting the normalized amplitude into an image Il containing 3×8 pixels as shown in fig. 3 (c), wherein the coordinates of each pixel in the image are consistent with the coordinates of the corresponding calibration area.
4) Super-resolution reconstruction is carried out on the image Il by using a bi-cubic spline interpolation method to obtain a high-resolution image Ih shown in fig. 3 (d), the number of pixel points is amplified to 6×16=96, the method is equivalent to dividing a screen board into 96 new calibration touch areas shown in fig. 2 after super-resolution reconstruction, and MAXA is utilized m And MINA m Ih is inversely normalized to obtain a new set of pixel values G 'as shown in fig. 3 (e)' m ={A' m (x',y')}(x’=1,2,…,6;y’=1,2,…,16)。
5) A 'is marked according to the sequence of the sampling points m and the coordinates (x', y ') of the touch area to be newly marked' m (x ', y') to be integrated into a new acoustic fingerprint signal d F,S',i,j (x ', y'), and obtaining an acoustic fingerprint library D at a high resolution as shown in FIG. 3 (f) F,S',i,j In which the new resolution
6) Using s=64 mm 2 Touch load with force of 1.2N and 2N respectivelyPreparing an acoustic fingerprint library, and repeatedly executing the steps 2) to 5), so as to obtain the acoustic fingerprint library D under three touch forces and high resolution F,S',i,j The corresponding touch forces of F=1, 2 and 3 are respectively 0.4N, 1.2N and 2N. Touch control is performed at the position shown in fig. 4 by using a force of 1.2N, and the signals obtained by the sensor pairs are taken as acoustic fingerprints to be positionedAnd substituting the obtained product into the following formula to perform preliminary solution,
in theta F ={θ F (x ', y') } characterizes the unknown touchPositioning result in high-resolution acoustic fingerprint library with touch force F, ++>To satisfy the above formula theta F Is a preliminary solution to (a).
7) The preliminary solutions are substituted into a multi-sensor-pair least squares model for which the following weight coefficients are iteratively adaptive,
s.t.θ F (x',y')≥0,for all x',y' (1b)
w in F,i,j For sensor pair T i-j And (3) a weight coefficient in a high-resolution acoustic fingerprint library with the touch force of F, wherein mu is a penalty coefficient.
8) Solving the recognition model in the step 7), and realizing automatic optimization of the weight coefficient by the following steps:
8a) Setting initial weight (w F,1,2 ) 0 =0.33,(w F,1,3 ) 0 =0.33,(w F,1,4 ) 0 =0.34;
8b) Substituting the initial weight into 7) the model to obtain an initial solution (Θ) F ) 0 ;
8c) Calculating the sum of squares of the residualsAnd update the weight coefficient
8d) Taking the new weight coefficient as the initial weight and repeating the steps 8 b) and 8 c) until the iteration is finished, so as to obtain the optimal weight and the positioning and identifying result theta F 。
9) By obtaining unknown touch acoustic fingerprintsPositioning result theta under each touch force F Using all theta F Maximum and minimum value pair Θ in (x ', y') F Proceed [0,1 ]]Normalization gives the result shown in FIG. 5, wherein FIGS. 5 (a) to (c) are +.>Positioning results at 0.4N, 1.2N, 2N. As is clear from the result, θ in fig. 5 (b), the threshold value ε=0.95 is set 2 The value of (4, 8) is greater than ε, so the acoustic fingerprint to be localized +.>The corresponding touch force is 1.2N, and the touch position is the nominal touch area with coordinates (4, 8) shown in FIG. 3 (b).
Claims (5)
1. The ultrasonic lamb wave touch load identification method based on super-resolution reconstruction is characterized in that a low-pixel touch load lamb wave acoustic fingerprint library is expanded to obtain a high-pixel touch load lamb wave acoustic fingerprint library by the super-resolution reconstruction method, and a least square model with automatic iterative optimization of weight coefficients is adopted to realize identification of touch load positions and touch forces; the specific implementation steps of touch load lamb wave acoustic fingerprint library extension are as follows:
s1, constructing a low-resolution acoustic fingerprint library; the medium of the touch sensing screen board is subjected to equal-size grid division, and N is respectively formed in the x direction and the y direction x 、N y The touch control areas are marked, and the whole screen board forms N altogether x ×N y N calibration touch areas, using a load touch (x, y) calibration area with force F and area S, the sensor pair T is noted i-j The acquired touch load lamb wave acoustic fingerprint signal is d F,S,i,j (x, y), i excites the sensor, j receives the sensor, and traverses the N calibration areas of touch control to construct an acoustic fingerprint library D with the resolution ratio S F,S,i,j ={d F,S,i,j (x,y)},x=1,2,…,N x ;y=1,2,…,N y ;
S2, acoustic fingerprint signal d F,S,i,j (x, y) extracting fingerprint library D for uniform sampling waveform with M sampling points F,S,i,j Amplitude A of the m-th sampling point of all acoustic fingerprint signals m (x, y), m=1, 2, …, M, make up a set of pixel values G m ={A m (x,y)};
S3, obtaining a set G m Maximum value MAXA in (a) m And minimum MINA m Using maximum and minimum pairs G m Proceed [0,1 ]]Normalizing, converting normalized amplitude into N x ×N y An image Il of each pixel point, wherein the coordinates of each pixel point in the image are consistent with the coordinates of the corresponding calibration area;
s4, performing super-resolution reconstruction on the image Il by using a bicubic spline interpolation method to obtain a high-resolution image Ih, and recording the number of pixel points to amplify to N' x ×N' x =n 'number, equivalent to the sense panel being divided into N' x ×N' x N' new calibration touch areas, using MAXA m And MINA m Performing inverse normalization on Ih to obtain a new pixel value set G' m ={A' m (x',y')},x’=1,2,…,N’ x ;y’=1,2,…,N’ y ;
S5, according to the sequence of the sampling points m and the new calibration touch area coordinates (x ', y'), A 'is carried out' m (x ', y') to be integrated into a new acoustic fingerprint signal d F,S',i,j (x ', y') to obtain an acoustic fingerprint library D at high resolution F,S',i,j In which the new resolution
2. The ultrasonic lamb wave touch load identification method based on super-resolution reconstruction according to claim 1, wherein acoustic fingerprint signals of the touch load are signals received by a receiving sensor after lamb waves actively excited by an excitation sensor are subjected to touch load scattering action, or signals passively received by the receiving sensor after lamb waves excited by touch actions, and types of the acoustic fingerprint signals comprise time domain signals and frequency domain signals.
3. The ultrasonic lamb wave touch load identification method based on super-resolution reconstruction of claim 1, wherein the image super-resolution method is a cubic spline interpolation method.
4. The ultrasonic lamb wave touch load identification method based on super-resolution reconstruction of claim 1, wherein the identification of the touch load position and touch force is realized by using a least square model of a multi-sensor pair with weight coefficient iterative self-adaption, and the matching of signal characteristics to be positioned and a touch load lamb wave acoustic fingerprint library is realized by the following specific steps:
step 11. N is F The low-resolution acoustic fingerprint library prepared under the individual touch control force is reconstructed through super resolution to obtain an acoustic fingerprint library D under the high resolution F,S',i,j ,F=1,2,…,N F Unknown touch is controlled by using the following methodThe preliminary solution is carried out so as to obtain,
in theta F ={θ F (x ', y') } characterizing an unknown touchPositioning result in high-resolution acoustic fingerprint library with touch force F, x ' =1, 2, …, N ' ' x ;y’=1,2,…,N’ y ,/>To satisfy the above formula theta F Is a preliminary solution to (a);
step 12, substituting the preliminary solution into a weight coefficient iteration self-adaptive multi-sensor pair least square model,
s.t.θ F (x',y')≥0,for all x',y' (1b)
w in F,i,j For sensor pair T i-j The weight coefficient in the high-resolution acoustic fingerprint library with the touch force of F, and mu is a punishment coefficient;
step 13, obtaining unknown touch acoustic fingerprintsPositioning result theta under each touch force F ,F=1,2,…,N F Using all theta F Maximum and minimum value pair Θ in (x ', y') F Proceed [0,1 ]]Normalized, θ exceeding a set threshold ε F The corresponding calibration touch area (x ', y') and touch force F are +.>And the touch position and touch force of the touch sensor are more than 0 and less than or equal to 1.
5. The ultrasonic lamb wave touch load identification method based on super-resolution reconstruction of claim 4, wherein the least square model is characterized in that the weight coefficient automatically finds an optimal value through iteration, and the method comprises the following specific steps:
step 121. Initial values (w) F,i,j ) 0 ;
Step 122 substitutes the initial weights and the signal to be located into the least squares model to solve (Θ) F ) 0 ;
Step 123. Calculate the sum of squares of the residualsAnd update the weight coefficient
Step 124, taking the new weight coefficient in step 123 as the initial weight, repeating step 122, step 123 to the end of iteration to obtain the optimal weight and the final result Θ F 。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110179053.4A CN113010038B (en) | 2021-02-09 | 2021-02-09 | Ultrasonic lamb wave touch load identification method based on super-resolution reconstruction |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110179053.4A CN113010038B (en) | 2021-02-09 | 2021-02-09 | Ultrasonic lamb wave touch load identification method based on super-resolution reconstruction |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113010038A CN113010038A (en) | 2021-06-22 |
CN113010038B true CN113010038B (en) | 2024-02-02 |
Family
ID=76383953
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110179053.4A Active CN113010038B (en) | 2021-02-09 | 2021-02-09 | Ultrasonic lamb wave touch load identification method based on super-resolution reconstruction |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113010038B (en) |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114260927A (en) * | 2021-12-16 | 2022-04-01 | 北京工业大学 | Robot perception skin based on ultrasonic lamb waves |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103020937A (en) * | 2012-12-12 | 2013-04-03 | 湖北微驾技术有限公司 | Method for improving face image super-resolution reconfiguration |
CN106709873A (en) * | 2016-11-11 | 2017-05-24 | 浙江师范大学 | Super-resolution method based on cubic spline interpolation and iterative updating |
CN109064402A (en) * | 2018-07-27 | 2018-12-21 | 四川大学 | Based on the single image super resolution ratio reconstruction method for enhancing non local total variation model priori |
CN111459324A (en) * | 2020-03-30 | 2020-07-28 | 北京工业大学 | Ultrasonic lamb wave touch screen |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2014081721A1 (en) * | 2012-11-21 | 2014-05-30 | The Board Of Trustees Of The Leland Stanford Junior University | A multi-touch ultrasonic touch screen |
US20180101255A1 (en) * | 2016-10-12 | 2018-04-12 | The Board Of Trustees Of The Leland Stanford Junior University | Methods for multi-touch ultrasonic touchscreens |
-
2021
- 2021-02-09 CN CN202110179053.4A patent/CN113010038B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103020937A (en) * | 2012-12-12 | 2013-04-03 | 湖北微驾技术有限公司 | Method for improving face image super-resolution reconfiguration |
CN106709873A (en) * | 2016-11-11 | 2017-05-24 | 浙江师范大学 | Super-resolution method based on cubic spline interpolation and iterative updating |
CN109064402A (en) * | 2018-07-27 | 2018-12-21 | 四川大学 | Based on the single image super resolution ratio reconstruction method for enhancing non local total variation model priori |
CN111459324A (en) * | 2020-03-30 | 2020-07-28 | 北京工业大学 | Ultrasonic lamb wave touch screen |
Also Published As
Publication number | Publication date |
---|---|
CN113010038A (en) | 2021-06-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110045015B (en) | Concrete structure internal defect detection method based on deep learning | |
Loutridis et al. | A two-dimensional wavelet transform for detection of cracks in plates | |
CN110431564A (en) | Method and system for Fingerprint enhancement | |
CN106990018B (en) | A kind of prestressed concrete beam Grouted density intelligent identification Method | |
CN112364779A (en) | Underwater sound target identification method based on signal processing and deep-shallow network multi-model fusion | |
CN113010038B (en) | Ultrasonic lamb wave touch load identification method based on super-resolution reconstruction | |
CN101751555A (en) | Deformation fingerprint identification method and system | |
WO2016049984A1 (en) | Fingerprint identification device and method | |
Perelli et al. | Best basis compressive sensing of guided waves in structural health monitoring | |
CN112529842B (en) | Multi-excitation fusion plate structure damage identification method based on wavelet packet energy | |
CN112634882B (en) | End-to-end real-time voice endpoint detection neural network model and training method | |
CN111754463A (en) | Ballastless track CA mortar layer defect detection method based on convolutional neural network | |
CN112215803A (en) | Aluminum plate eddy current inspection image defect segmentation method based on improved generation countermeasure network | |
CN112137620A (en) | Ultra-wideband radar-based human body weak respiration signal detection method | |
CN115420806A (en) | Nondestructive ultrasonic detection method based on neural network and image fusion | |
CN113126099A (en) | Single-beam substrate classification method based on audio feature extraction | |
CN102521795B (en) | Cross matching fingerprint image scaling method based on global ridge distance | |
CN115561307B (en) | Grouting compactness detection method | |
Zhao et al. | Ultrasonic guided wave inversion based on deep learning restoration for fingerprint recognition | |
CN106680371B (en) | Online flaw detection device and method for electroceramic body | |
CN115656034A (en) | Structure hidden defect recognition system based on vibration signal nonlinear characteristics | |
CN116228520A (en) | Image compressed sensing reconstruction method and system based on transform generation countermeasure network | |
CN112818762B (en) | Large-size composite material and rapid nondestructive testing method for sandwich structure thereof | |
CN110210410B (en) | Handwritten number recognition method based on image characteristics | |
JP3502208B2 (en) | Process identification method and apparatus |
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 |