CN105631830A - Photoacoustic image denoising method and denoising device - Google Patents
Photoacoustic image denoising method and denoising device Download PDFInfo
- Publication number
- CN105631830A CN105631830A CN201610130722.8A CN201610130722A CN105631830A CN 105631830 A CN105631830 A CN 105631830A CN 201610130722 A CN201610130722 A CN 201610130722A CN 105631830 A CN105631830 A CN 105631830A
- Authority
- CN
- China
- Prior art keywords
- matrix
- image matrix
- photoacoustic image
- denoising
- unit
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 49
- 239000011159 matrix material Substances 0.000 claims abstract description 199
- 238000013507 mapping Methods 0.000 claims description 18
- 238000010276 construction Methods 0.000 claims description 9
- 230000004069 differentiation Effects 0.000 claims description 7
- 230000002146 bilateral effect Effects 0.000 claims description 3
- 230000004304 visual acuity Effects 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 238000003384 imaging method Methods 0.000 description 3
- 230000007423 decrease Effects 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000002059 diagnostic imaging Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
Classifications
-
- G06T5/70—
Abstract
The invention discloses a photoacoustic image denoising method and a denoising device. In view of features that a singular value of a noise-containing photoacoustic image drops rapidly and low-rank performance is strong, a low-rank matrix approximation method is used, denoising processing is carried out on the noise-containing photoacoustic image, Gauss noise and impulse noise can be removed at the same time in a condition of ensuring the resolution of the photoacoustic image is not seriously damaged, and the photoacoustic image quality is improved.
Description
Technical field
The present invention relates to photoacoustic image process field, more specifically relate to a kind of photoacoustic image denoising method and device.
Background technology
Photoacoustic imaging(PAI) (PhotoacousticImaging, PAI) it is the new bio medical imaging procedure of a kind of non-invasive of getting up of development in recent years and non-ionization type, the method effectively combines pure optical image technology high-contrast and the feature of pure ultra sonic imaging technology high-penetrability, it is possible to obtain the high resolving power of deep tissues and the tissue image of high-contrast. And in actual light acoustic imaging process, inevitably containing noise in photoacoustic signal, such as process from detector Received signal strength to electronics and it is transferred to computer, whole process all likely introduces noise: in the process that photoacoustic signal is received, owing to the characteristic electron of receptor can make each signal introduce Gaussian noise; When image data or reception data, the detector of driving stepper motor or receptor need to carry out moving or rotating, and can introduce a small amount of impulse noise. The data gathering carried out in computer and image are rebuild and are had very big interference by this, it is possible to can cause that the signal to noise ratio of photoacoustic image is low, contrast gradient is low, and resolving power is low.
Summary of the invention
(1) technical problem solved
The technical problem to be solved in the present invention is the sparse impulse noise and Gaussian noise how removing in photoacoustic image.
(2) technical scheme
In order to solve the problems of the technologies described above, the present invention provides a kind of photoacoustic image denoising method, and described method comprises the following steps:
S1, input source photoacoustic image matrix, containing noise in the photoacoustic image matrix of wherein said source;
Described source photoacoustic image matrix is carried out denoising by S2, the method utilizing low-rank matrix to be similar to, and obtains target light acoustic image matrix.
Preferably, the method utilizing low-rank matrix approximate in described step S2 removes the sparse impulse noise in the photoacoustic image matrix of described source and Gaussian noise.
Preferably, described step S2 comprises the following steps:
S21, set up denoising model according to described source photoacoustic image matrix;
S22, the method utilizing described low-rank matrix to be similar to solve described denoising model, obtain the described target light acoustic image matrix after denoising.
Preferably, described denoising model is:
s.t.rank(L)��r,car(S)��k
In formula, X is described source photoacoustic image matrix, L is described target light acoustic image matrix, S is sparse impulse noise matrix, rank (L) is the order of described target light acoustic image matrix L, the number of the element that car (S) comprises for described sparse impulse noise matrix, r is predetermined rank value, and k is the preset value of the element number that described sparse impulse noise matrix comprises.
Preferably, described step S22 comprises the following steps:
S221, set described predetermined rank value r, preset value k, the maximum iteration time t of element number that described sparse impulse noise matrix comprisesmax, iteration stopping differentiation value ��; Initialize iteration number of times t is zero simultaneously, target light acoustic image matrix L described in initialize0Equal described source photoacoustic image matrix, sparse impulse noise matrix S described in initialize0It is zero;
S222, described iteration number of times t add 1;
S223, structure two random Gaussian matrix A1And A2:
Wherein, m is the line number of described source photoacoustic image matrix, and n is the row number of described source photoacoustic image matrix;
S224, ask for matrix X-St-1Bilateral random mapping value Y1And Y2:
Y1=(X-St-1)A1
A2=Y1
Y2=(X-St-1)TA2
Predetermined rank value r described in S225, comparison withSize, ifThen carry out S226; Otherwise return to S222. WhereinFor matrixOrder;
S226, to upgrade described target light acoustic image matrix be Lt:
S227, to upgrade described sparse impulse noise matrix be St: St��P��(X-Lt), wherein P��For | X-Lt| what front k was maximum is not the element of 0;
S228, calculatingValue, judge whether this value is less than �� or t >=tmaxWhether set up, if this value is less than �� or t >=tmax, then iteration is stopped, the target light acoustic image matrix L in this momenttFor final target light acoustic image matrix, otherwise return described step S222.
The invention also discloses a kind of photoacoustic image denoising device corresponding to aforesaid method, described device comprises source photoacoustic image matrix acquiring unit and denoising unit;
Described source photoacoustic image matrix acquiring unit, for inputting source photoacoustic image matrix, contains noise in the photoacoustic image matrix of wherein said source;
For utilizing, described source photoacoustic image matrix is carried out denoising to described denoising unit by the method that low-rank matrix is approximate, obtains target light acoustic image matrix.
Preferably, described denoising unit comprises denoising subelement, and the method that described denoising subelement is similar to for utilizing low-rank matrix removes the sparse impulse noise in the photoacoustic image matrix of described source and Gaussian noise.
Preferably, described denoising subelement comprises denoising model and sets up unit and processing unit;
Described denoising model sets up unit for setting up denoising model according to described source photoacoustic image matrix;
For utilizing, the method that described low-rank matrix is approximate solves described denoising model to described processing unit, obtains the described target light acoustic image matrix after denoising.
Preferably, described denoising model is:
s.t.rank(L)��r,car(S)��k
In formula, X is described source photoacoustic image matrix, L is described target light acoustic image matrix, S is sparse impulse noise matrix, rank (L) is the order of described target light acoustic image matrix L, the number of the element that car (S) comprises for described sparse impulse noise matrix, r is predetermined rank value, and k is the preset value of the element number that described sparse impulse noise matrix comprises.
Preferably, described processing unit comprises parameter setting unit, iteration number of times increases unit, random Gaussian matrix construction unit, random mapping value calculate unit, compare unit, updating block and calculate output unit;
Described parameter setting unit for setting described predetermined rank value r, preset value k, the maximum iteration time t of element number that described sparse impulse noise matrix comprisesmax, iteration stopping differentiation value ��; Initialize iteration number of times t is zero simultaneously, target light acoustic image matrix L described in initialize0Equal described source photoacoustic image matrix, sparse impulse noise matrix S described in initialize0It is zero;
Described iteration number of times increases unit and is used for arranging iteration number of times t before first time iteration and adds 1, and arranges iteration number of times t under the control continuing iterative instruction and add 1;
Described random Gaussian matrix construction unit is used for after iteration number of times t adds 1, builds two random Gaussian matrix A1And A2:
Wherein, m is the line number of described source photoacoustic image matrix, and n is the row number of described source photoacoustic image matrix;
Described random mapping value calculates unit for building two random Gaussian matrix A in described random Gaussian matrix construction unit1And A2After ask for matrix X-St-1Bilateral random mapping value Y1And Y2:
Y1=(X-St-1)A1
A2=Y1
Y2=(X-St-1)TA2
The described unit that compares calculates bilateral random mapping value Y for calculating unit in described random mapping value1And Y2After, relatively described predetermined rank value r withSize, ifThen send and upgrade instruction to described updating block; Otherwise issue described continuation iterative instruction and increase unit to described iteration number of times;
Described updating block is L for upgrading described target light acoustic image matrix under the control of described renewal instructiont:Upgrading described sparse impulse noise matrix is St: St��P��(X-Lt), wherein P��For | X-Lt| what front k was maximum is not the element of 0;
Described calculating output unit is used for calculating after updating block completes renewalValue, judge whether this value is less than �� or t >=tmaxWhether set up, if this value is less than �� or t >=tmax, then stopping iteration, the target light acoustic image matrix in this moment is LtFor final target light acoustic image matrix, otherwise issue described continuation iterative instruction and increase unit to described iteration number of times.
(3) useful effect
The present invention provides a kind of photoacoustic image denoising method and device, the present invention is directed to the decline of the singular value of the photoacoustic image containing noise rapidly, the feature that low-rank is stronger utilizes low-rank matrix to be similar to method, photoacoustic image containing noise is carried out denoising, when can ensure the not serious loss of photoacoustic image resolving power, remove Gaussian noise and impulse noise, it is to increase photoacoustic image quality simultaneously.
Accompanying drawing explanation
In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, it is briefly described to the accompanying drawing used required in embodiment or description of the prior art below, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skill in the art, under the prerequisite not paying creative work, it is also possible to obtain other accompanying drawing according to these accompanying drawings.
Fig. 1 is the schema of the photoacoustic image denoising method of the present invention;
Fig. 2 is the schematic diagram of source photoacoustic image in the present invention;
Fig. 3 is the schematic diagram of target light acoustic image in the present invention.
Embodiment
Below in conjunction with drawings and Examples, the present invention is described in further detail. Following examples are for illustration of the present invention, but can not be used for limiting the scope of the invention.
A kind of photoacoustic image denoising method, as shown in Figure 1, described method comprises the following steps:
S1, input source photoacoustic image matrix, containing noise in the photoacoustic image matrix of wherein said source;
Described source photoacoustic image matrix is carried out denoising by S2, the method utilizing low-rank matrix to be similar to, and obtains target light acoustic image matrix.
Aforesaid method is rapid for the singular value decline of the photoacoustic image containing noise, the feature that low-rank is stronger utilizes low-rank matrix to be similar to method, photoacoustic image containing noise is carried out denoising, when can ensure the not serious loss of photoacoustic image resolving power, remove Gaussian noise and impulse noise, it is to increase photoacoustic image quality simultaneously. Wherein, the method utilizing low-rank matrix approximate in described step S2 can remove the sparse impulse noise in the photoacoustic image matrix of described source and Gaussian noise.
Further, described step S2 specifically comprises the following steps:
S21, set up denoising model according to described source photoacoustic image matrix; Wherein, the dimension that described source photoacoustic image matrix X is is m �� n, and the such as dimension of matrix X is 500 �� 500, and it can be broken down into low-rank part, sparse part and Gaussian noise part, i.e. X=L+S+N. Wherein, L is the low-rank part of matrix X, i.e. target light acoustic image matrix after denoising, and S is the sparse part of matrix X, namely a small amount of sparse impulse noise matrix, and N is Gaussian noise part;
S22, the method utilizing described low-rank matrix approximate, namely utilize GoDec algorithm to solve described denoising model, obtain the described target light acoustic image matrix after denoising.
Wherein, described denoising model is:
s.t.rank(L)��r,car(S)��k
In formula, X is described source photoacoustic image matrix, L is described target light acoustic image matrix, S is sparse impulse noise matrix, rank (L) is the order of described target light acoustic image matrix L, the number of the element that car (S) comprises for described sparse impulse noise matrix, r is predetermined rank value, and k is the preset value of the element number that described sparse impulse noise matrix comprises.
Described step S22 specifically comprises the following steps:
S221, set described predetermined rank value r, preset value k, the maximum iteration time t of element number that described sparse impulse noise matrix comprisesmax, iteration stopping differentiation value ��; Initialize iteration number of times t is zero simultaneously, target light acoustic image matrix L described in initialize0Equal described source photoacoustic image matrix, i.e. L0=X, sparse impulse noise matrix S described in initialize0It is zero; The order r=134 such as set, iteration stopping differentiation value ��=1e-4, maximum iteration time tmax=1000, parameter k gets 26��30 according to experience;
S222, described iteration number of times t add 1;
S223, structure two random Gaussian matrix A1And A2:
Wherein, m is the line number of described source photoacoustic image matrix, and n is the row number of described source photoacoustic image matrix;
S224, ask for matrix X-St-1Bilateral random mapping value Y1And Y2:
Y1=(X-St-1)A1
A2=Y1
Y2=(X-St-1)TA2
Predetermined rank value r described in S225, comparison withSize, ifThen carry out S226; Otherwise return to S222. WhereinFor matrixOrder;
S226, to upgrade described target light acoustic image matrix be Lt:
S227, to upgrade described sparse impulse noise matrix be St: St��P��(X-Lt), wherein P��For | X-Lt| what front k was maximum is not the element of 0;
S228, calculatingValue, judge whether this value is less than �� or t >=tmaxWhether set up, if this value is less than �� or t >=tmax, then iteration is stopped, the target light acoustic image matrix L in this momenttFor final target light acoustic image matrix, otherwise return described step S222.
Such as Fig. 2, shown in 3, when utilizing aforesaid method can ensure the not serious loss of photoacoustic image resolving power, remove Gaussian noise and impulse noise, it is to increase photoacoustic image quality simultaneously.
The invention also discloses a kind of photoacoustic image denoising device corresponding to aforesaid method, described device comprises source photoacoustic image matrix acquiring unit and denoising unit; Described source photoacoustic image matrix acquiring unit, for inputting source photoacoustic image matrix, contains noise in the photoacoustic image matrix of wherein said source; For utilizing, described source photoacoustic image matrix is carried out denoising to described denoising unit by the method that low-rank matrix is approximate, obtains target light acoustic image matrix.
Above-mentioned denoising unit comprises denoising subelement, and the method that described denoising subelement is similar to for utilizing low-rank matrix can remove the sparse impulse noise in the photoacoustic image matrix of described source and Gaussian noise.
Further, described denoising subelement comprises denoising model and sets up unit and processing unit; Described denoising model sets up unit for setting up denoising model according to described source photoacoustic image matrix; For utilizing, the method that described low-rank matrix is approximate solves described denoising model to described processing unit, obtains the described target light acoustic image matrix after denoising.
Wherein, described denoising model is:
s.t.rank(L)��r,car(S)��k
In formula, X is described source photoacoustic image matrix, L is described target light acoustic image matrix, S is sparse impulse noise matrix, rank (L) is the order of described target light acoustic image matrix L, the number of the element that car (S) comprises for described sparse impulse noise matrix, r is predetermined rank value, and k is the preset value of the element number that described sparse impulse noise matrix comprises.
Further, described processing unit comprises parameter setting unit, iteration number of times increases unit, random Gaussian matrix construction unit, random mapping value calculate unit, compare unit, updating block and calculate output unit.
Described parameter setting unit for setting described predetermined rank value r, preset value k, the maximum iteration time t of element number that described sparse impulse noise matrix comprisesmax, iteration stopping differentiation value ��; Initialize iteration number of times t is zero simultaneously, target light acoustic image matrix L described in initialize0Equal described source photoacoustic image matrix, sparse impulse noise matrix S described in initialize0It is zero;
Described iteration number of times increases unit and is used for arranging iteration number of times t before first time iteration and adds 1, and arranges iteration number of times t under the control continuing iterative instruction and add 1;
Described random Gaussian matrix construction unit is used for after iteration number of times t adds 1, builds two random Gaussian matrix A1And A2:
Wherein, m is the line number of described source photoacoustic image matrix, and n is the row number of described source photoacoustic image matrix;
Described random mapping value calculates unit for building two random Gaussian matrix A in described random Gaussian matrix construction unit1And A2After ask for matrix X-St-1Bilateral random mapping value Y1And Y2:
Y1=(X-St-1)A1
A2=Y1
Y2=(X-St-1)TA2
The described unit that compares calculates bilateral random mapping value Y for calculating unit in described random mapping value1And Y2After, relatively described predetermined rank value r withSize, ifThen send and upgrade instruction to described updating block; Otherwise issue described continuation iterative instruction and increase unit to described iteration number of times;
Described updating block is L for upgrading described target light acoustic image matrix under the control of described renewal instructiont:Upgrading described sparse impulse noise matrix is St: St��P��(X-Lt), wherein P��For | X-Lt| what front k was maximum is not the element of 0;
Described calculating output unit is used for calculating after updating block completes renewalValue, judge whether this value is less than �� or t >=tmaxWhether set up, if this value is less than �� or t >=tmax, then stopping iteration, the target light acoustic image matrix in this moment is LtFor final target light acoustic image matrix, otherwise issue described continuation iterative instruction and increase unit to described iteration number of times.
Owing to said apparatus is corresponding with aforesaid method to the treatment step of source photoacoustic image matrix, so the part repeated repeats no more here.
Mode of more than implementing is only for illustration of the present invention, but not limitation of the present invention. Although with reference to embodiment to invention has been detailed explanation, it will be understood by those within the art that, the technical scheme of the present invention is carried out various combination, amendment or equivalent replacement, do not depart from the spirit and scope of technical solution of the present invention, all should be encompassed in the middle of the right of the present invention.
Claims (10)
1. a photoacoustic image denoising method, it is characterised in that, described method comprises the following steps:
S1, input source photoacoustic image matrix, containing noise in the photoacoustic image matrix of wherein said source;
Described source photoacoustic image matrix is carried out denoising by S2, the method utilizing low-rank matrix to be similar to, and obtains target light acoustic image matrix.
2. method according to claim 1, it is characterised in that, the method utilizing low-rank matrix approximate in described step S2 removes the sparse impulse noise in the photoacoustic image matrix of described source and Gaussian noise.
3. method according to claim 1 and 2, it is characterised in that, described step S2 comprises the following steps:
S21, set up denoising model according to described source photoacoustic image matrix;
S22, the method utilizing described low-rank matrix to be similar to solve described denoising model, obtain the described target light acoustic image matrix after denoising.
4. method according to claim 3, it is characterised in that, described denoising model is:
s.t.rank(L)��r,car(S)��k
In formula, X is described source photoacoustic image matrix, L is described target light acoustic image matrix, S is sparse impulse noise matrix, rank (L) is the order of described target light acoustic image matrix L, the number of the element that car (S) comprises for described sparse impulse noise matrix, r is predetermined rank value, and k is the preset value of the element number that described sparse impulse noise matrix comprises.
5. method according to claim 4, it is characterised in that, described step S22 comprises the following steps:
S221, set described predetermined rank value r, preset value k, the maximum iteration time t of element number that described sparse impulse noise matrix comprisesmax, iteration stopping differentiation value ��; Initialize iteration number of times t is zero simultaneously, target light acoustic image matrix L described in initialize0Equal described source photoacoustic image matrix, sparse impulse noise matrix S described in initialize0It is zero;
S222, described iteration number of times t add 1;
S223, structure two random Gaussian matrix A1And A2:
Wherein, m is the line number of described source photoacoustic image matrix, and n is the row number of described source photoacoustic image matrix;
S224, ask for matrix X-St-1Bilateral random mapping value Y1And Y2:
Y1=(X-St-1)A1
A2=Y1
Y2=(X-St-1)TA2
Predetermined rank value r described in S225, comparison withSize, ifThen carry out S226; Otherwise return to S222; WhereinFor matrixOrder;
S226, to upgrade described target light acoustic image matrix be Lt:
S227, to upgrade described sparse impulse noise matrix be St: St��P��(X-Lt), wherein P��For | X-Lt| what front k was maximum is not the element of 0;
S228, calculatingValue, judge whether this value is less than �� or t >=tmaxWhether set up, if this value is less than �� or t >=tmax, then iteration is stopped, the target light acoustic image matrix L in this momenttFor final target light acoustic image matrix, otherwise return described step S222.
6. a photoacoustic image denoising device, it is characterised in that, described device comprises source photoacoustic image matrix acquiring unit and denoising unit;
Described source photoacoustic image matrix acquiring unit, for inputting source photoacoustic image matrix, contains noise in the photoacoustic image matrix of wherein said source;
For utilizing, described source photoacoustic image matrix is carried out denoising to described denoising unit by the method that low-rank matrix is approximate, obtains target light acoustic image matrix.
7. device according to claim 6, it is characterized in that, described denoising unit comprises denoising subelement, and the method that described denoising subelement is similar to for utilizing low-rank matrix removes the sparse impulse noise in the photoacoustic image matrix of described source and Gaussian noise.
8. device according to claim 7, it is characterised in that, described denoising subelement comprises denoising model and sets up unit and processing unit;
Described denoising model sets up unit for setting up denoising model according to described source photoacoustic image matrix;
For utilizing, the method that described low-rank matrix is approximate solves described denoising model to described processing unit, obtains the described target light acoustic image matrix after denoising.
9. device according to claim 8, it is characterised in that, described denoising model is:
s.t.rank(L)��r,car(S)��k
In formula, X is described source photoacoustic image matrix, L is described target light acoustic image matrix, S is sparse impulse noise matrix, rank (L) is the order of described target light acoustic image matrix L, the number of the element that car (S) comprises for described sparse impulse noise matrix, r is predetermined rank value, and k is the preset value of the element number that described sparse impulse noise matrix comprises.
10. device according to claim 9, it is characterized in that, described processing unit comprises parameter setting unit, iteration number of times increases unit, random Gaussian matrix construction unit, random mapping value calculate unit, compare unit, updating block and calculate output unit;
Described parameter setting unit for setting described predetermined rank value r, preset value k, the maximum iteration time t of element number that described sparse impulse noise matrix comprisesmax, iteration stopping differentiation value ��; Initialize iteration number of times t is zero simultaneously, target light acoustic image matrix L described in initialize0Equal described source photoacoustic image matrix, sparse impulse noise matrix S described in initialize0It is zero;
Described iteration number of times increases unit and is used for arranging iteration number of times t before first time iteration and adds 1, and arranges iteration number of times t under the control continuing iterative instruction and add 1;
Described random Gaussian matrix construction unit is used for after iteration number of times t adds 1, builds two random Gaussian matrix A1And A2:
Wherein, m is the line number of described source photoacoustic image matrix, and n is the row number of described source photoacoustic image matrix;
Described random mapping value calculates unit for building two random Gaussian matrix A in described random Gaussian matrix construction unit1And A2After ask for matrix X-St-1Bilateral random mapping value Y1And Y2:
Y1=(X-St-1)A1
A2=Y1
Y2=(X-St-1)TA2;
The described unit that compares calculates bilateral random mapping value Y for calculating unit in described random mapping value1And Y2After, relatively described predetermined rank value r withSize, ifThen send and upgrade instruction to described updating block; Otherwise issue described continuation iterative instruction and increase unit to described iteration number of times;
Described updating block is L for upgrading described target light acoustic image matrix under the control of described renewal instructiont:Upgrading described sparse impulse noise matrix is St: St��P��(X-Lt), wherein P��For | X-Lt| what front k was maximum is not the element of 0;
Described calculating output unit is used for calculating after updating block completes renewalValue, judge whether this value is less than �� or t >=tmaxWhether set up, if this value is less than �� or t >=tmax, then stopping iteration, the target light acoustic image matrix in this moment is LtFor final target light acoustic image matrix, otherwise issue described continuation iterative instruction and increase unit to described iteration number of times.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610130722.8A CN105631830A (en) | 2016-03-08 | 2016-03-08 | Photoacoustic image denoising method and denoising device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610130722.8A CN105631830A (en) | 2016-03-08 | 2016-03-08 | Photoacoustic image denoising method and denoising device |
Publications (1)
Publication Number | Publication Date |
---|---|
CN105631830A true CN105631830A (en) | 2016-06-01 |
Family
ID=56046718
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610130722.8A Pending CN105631830A (en) | 2016-03-08 | 2016-03-08 | Photoacoustic image denoising method and denoising device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105631830A (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108848352A (en) * | 2018-07-21 | 2018-11-20 | 杨建怀 | A kind of cloud service video monitoring system |
CN110675331A (en) * | 2019-08-13 | 2020-01-10 | 南京人工智能高等研究院有限公司 | Image denoising method and device, computer readable storage medium and electronic device |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130204137A1 (en) * | 2012-02-03 | 2013-08-08 | Delphinus Medical Technologies, Inc. | Method and System for Denoising Acoustic Travel Times and Imaging a Volume of Tissue |
CN103745445A (en) * | 2014-01-21 | 2014-04-23 | 中国科学院地理科学与资源研究所 | Gaussian and pulse mixed noise removing method and device |
CN104159003A (en) * | 2014-08-21 | 2014-11-19 | 武汉大学 | Method and system of video denoising based on 3D cooperative filtering and low-rank matrix reconstruction |
CN104299216A (en) * | 2014-10-22 | 2015-01-21 | 北京航空航天大学 | Multimodality medical image fusion method based on multiscale anisotropic decomposition and low rank analysis |
-
2016
- 2016-03-08 CN CN201610130722.8A patent/CN105631830A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20130204137A1 (en) * | 2012-02-03 | 2013-08-08 | Delphinus Medical Technologies, Inc. | Method and System for Denoising Acoustic Travel Times and Imaging a Volume of Tissue |
CN103745445A (en) * | 2014-01-21 | 2014-04-23 | 中国科学院地理科学与资源研究所 | Gaussian and pulse mixed noise removing method and device |
CN104159003A (en) * | 2014-08-21 | 2014-11-19 | 武汉大学 | Method and system of video denoising based on 3D cooperative filtering and low-rank matrix reconstruction |
CN104299216A (en) * | 2014-10-22 | 2015-01-21 | 北京航空航天大学 | Multimodality medical image fusion method based on multiscale anisotropic decomposition and low rank analysis |
Non-Patent Citations (1)
Title |
---|
袁珍 等: "滤除图像中混合噪声的LSE模型", 《信号处理》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108848352A (en) * | 2018-07-21 | 2018-11-20 | 杨建怀 | A kind of cloud service video monitoring system |
CN110675331A (en) * | 2019-08-13 | 2020-01-10 | 南京人工智能高等研究院有限公司 | Image denoising method and device, computer readable storage medium and electronic device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2023092813A1 (en) | Swin-transformer image denoising method and system based on channel attention | |
WO2021232653A1 (en) | Pet image reconstruction algorithm combining filtered back-projection algorithm and neural network | |
Vedula et al. | Towards CT-quality ultrasound imaging using deep learning | |
Ibrahim et al. | A medical image enhancement based on generalized class of fractional partial differential equations | |
CN104156919B (en) | A kind of based on wavelet transformation with the method for restoring motion blurred image of Hopfield neutral net | |
Shtok et al. | Sparsity-based sinogram denoising for low-dose computed tomography | |
CN105631830A (en) | Photoacoustic image denoising method and denoising device | |
CN110189260B (en) | Image noise reduction method based on multi-scale parallel gated neural network | |
CN108095756B (en) | Super-high resolution plane wave ultrasonic imaging method based on SOFI | |
CN105528766A (en) | CT metal artifact processing method and device | |
Wang et al. | Can a single image denoising neural network handle all levels of gaussian noise? | |
CN110880196A (en) | Tumor photoacoustic image rapid reconstruction method and device based on deep learning | |
Brickson et al. | Reverberation noise suppression in the aperture domain using 3D fully convolutional neural networks | |
CN114611387A (en) | Method and device for improving measurement precision of electromagnetic characteristics of equipment | |
Maiseli | Nonlinear anisotropic diffusion methods for image denoising problems: Challenges and future research opportunities | |
CN113870178A (en) | Plaque artifact correction and component analysis method and device based on artificial intelligence | |
CN116728291A (en) | Robot polishing system state monitoring method and device based on edge calculation | |
CN104754183A (en) | Real-time monitoring video adaptive filtering method and real-time monitoring video adaptive filtering system | |
CN111028182B (en) | Image sharpening method, device, electronic equipment and computer readable storage medium | |
CN108742627B (en) | Detection apparatus based on brain medical image fusion classification | |
CN111539885A (en) | Image enhancement defogging method based on multi-scale network | |
CN112435177B (en) | Recursive infrared image non-uniform correction method based on SRU and residual error network | |
Chen et al. | A novel edge-preserving filter for medical image enhancement | |
CN113469904A (en) | General image quality enhancement method and device based on cycle consistency loss | |
CN114529463A (en) | Image denoising method and system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160601 |