CN107292845A - Based on the pyramidal dynamic image noise-reduction method of standard deviation and device - Google Patents
Based on the pyramidal dynamic image noise-reduction method of standard deviation and device Download PDFInfo
- Publication number
- CN107292845A CN107292845A CN201710492019.6A CN201710492019A CN107292845A CN 107292845 A CN107292845 A CN 107292845A CN 201710492019 A CN201710492019 A CN 201710492019A CN 107292845 A CN107292845 A CN 107292845A
- Authority
- CN
- China
- Prior art keywords
- pyramid
- mrow
- noise reduction
- standard deviation
- layers
- 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
- 238000000034 method Methods 0.000 title claims abstract description 23
- 230000009467 reduction Effects 0.000 claims abstract description 116
- 238000011946 reduction process Methods 0.000 claims abstract description 22
- 230000002708 enhancing effect Effects 0.000 claims abstract description 19
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 6
- 230000000694 effects Effects 0.000 abstract description 3
- 238000010586 diagram Methods 0.000 description 5
- PCHJSUWPFVWCPO-UHFFFAOYSA-N gold Chemical compound [Au] PCHJSUWPFVWCPO-UHFFFAOYSA-N 0.000 description 4
- 239000010931 gold Substances 0.000 description 4
- 229910052737 gold Inorganic materials 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000001914 filtration Methods 0.000 description 3
- 238000005070 sampling Methods 0.000 description 3
- 230000003247 decreasing effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008901 benefit Effects 0.000 description 1
- 239000000686 essence Substances 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/70—Denoising; Smoothing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/73—Deblurring; Sharpening
- G06T5/75—Unsharp masking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T5/00—Image enhancement or restoration
- G06T5/90—Dynamic range modification of images or parts thereof
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10116—X-ray image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20016—Hierarchical, coarse-to-fine, multiscale or multiresolution image processing; Pyramid transform
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The present invention is applied to medical image processing technical field and is based on the pyramidal dynamic image noise-reduction method of standard deviation there is provided one kind, including：Pyramid decomposition is carried out to pending dynamic image using pyramid multiresolution overall situation enhancing algorithm, N layers of gaussian pyramid and N layers of laplacian pyramid is obtained；Neighborhood standard deviation is calculated to every layer of gaussian pyramid, N layers of standard deviation pyramid is obtained, and noise reduction mask pyramid is generated according to standard deviation pyramid；Noise reduction process is carried out to N layers of the laplacian pyramid using the noise reduction mask pyramid of generation, the laplacian pyramid after noise reduction is obtained；Pyramid Reconstruction is carried out using the laplacian pyramid after noise reduction, the dynamic image after noise reduction is obtained.A kind of noise for amplifying when can suppress to strengthen the details of dynamic image using pyramid multiresolution overall situation enhancing algorithm based on the pyramidal dynamic image noise-reduction method of standard deviation that the present invention is provided, so as to improve the display effect of dynamic image.
Description
Technical field
The pyramidal dynamic image of standard deviation is based on the invention belongs to medical image processing technical field, more particularly to one kind
Noise-reduction method and device.
Background technology
Medical x-ray Imaging enhanced, through frequently with pyramid multiresolution overall situation enhancing algorithm, this algorithm is will be original
Dynamic image is decomposed into gaussian pyramid and laplacian pyramid, then the global enhancing of laplacian pyramid progress is handled,
And be reconstructed using the laplacian pyramid after processing；Dynamic image details protrusion after this processing, good contrast, layer
Secondary sense is strong, but it does not differentiate between edge and noise during details is enhanced, all does the enhancing of intensity identical, in enhancing details
Noise is also exaggerated simultaneously, noise has a strong impact on the quality of dynamic image.
The content of the invention
The present invention provides a kind of based on the pyramidal dynamic image noise-reduction method of standard deviation and device, it is desirable to provide one kind drop
The method made an uproar is amplified when suppressing and the details of dynamic image being strengthened using pyramid multiresolution overall situation enhancing algorithm
Noise, so as to improve the display effect of dynamic image.
The pyramidal dynamic image noise-reduction method of standard deviation is based on the invention provides one kind, including：
Pyramid decomposition is carried out to pending dynamic image using pyramid multiresolution overall situation enhancing algorithm, N is obtained
The gaussian pyramid and N layers of laplacian pyramid of layer；
Neighborhood standard deviation is calculated to every layer of the gaussian pyramid, N layers of standard deviation pyramid is obtained, and according to described
Standard deviation pyramid generates noise reduction mask pyramid；
Noise reduction process is carried out to N layers of the laplacian pyramid using the noise reduction mask pyramid of generation, dropped
Laplacian pyramid after making an uproar；
Pyramid Reconstruction is carried out using the laplacian pyramid after noise reduction, the dynamic image after noise reduction is obtained.
Further, the generation pyramidal formula of standard deviation is：
Wherein, StdPyr is standard deviation pyramid, and l is pyramidal layer, and (R, C) is in the neighborhood centered on (r, c)
All coordinates, r is row coordinate, and c is row coordinate, and STD2 represents to seek the standard deviation of contiguous range pixel value, and GaussPyr is Gauss
Pyramid, StdPyr [l] (r, c) is that the coordinate of pyramidal l layers of standard deviation is the standard deviation of the pixel of (r, c), STD2
(GaussPyr [l] (R, C)) is the standard deviation of all coordinates (R, C) in l layers of the neighborhood centered on (r, c), STD2
(GaussPyr [l]) is the standard deviation of l tomographic images.
Further, the generation pyramidal formula of noise reduction mask is：
And wherein,
CurveCoeff=2*log (9),
Wherein, DnsMaskPyr is noise reduction mask pyramid, and l is pyramidal layer, and curveCoeff is curve coefficients,
DnsParam [l] is l layers preset of noise reduction intensity, and StdPyr [l] is l layers of standard deviation pyramid, and dnsParam is
The 0th layer preset of noise reduction intensity, dnsLevels is the noise reduction number of plies.
Further, it is to the formula that laplacian pyramid carries out noise reduction process：
DnsLaplacePyr=DnsMaskPyr*LaplacePyr,
Wherein, DnsLaplacePyr is noise reduction laplacian pyramid, and DnsMaskPyr is noise reduction mask pyramid,
LaplacePyr is laplacian pyramid.
The pyramidal dynamic image denoising device of standard deviation is based on present invention also offers one kind, including：
Decomposing module, for carrying out golden word to pending dynamic image using pyramid multiresolution overall situation enhancing algorithm
Tower is decomposed, and obtains N layers of gaussian pyramid and N layers of laplacian pyramid；
Standard deviation pyramid generation module, for calculating neighborhood standard deviation to every layer of the gaussian pyramid, obtains N layers
Standard deviation pyramid；
Noise reduction mask pyramid generation module, for generating noise reduction mask pyramid according to the standard deviation pyramid；
Noise reduction process module, the laplacian pyramid for the noise reduction mask pyramid using generation to N layers enters
Row noise reduction process, obtains the laplacian pyramid after noise reduction；
Reconstructed module, for carrying out Pyramid Reconstruction using the laplacian pyramid after noise reduction, obtains dynamic after noise reduction
State image.
Further, the generation pyramidal formula of standard deviation is：
Wherein, StdPyr is standard deviation pyramid, and l is pyramidal layer, and (R, C) is in the neighborhood centered on (r, c)
All coordinates, r is row coordinate, and c is row coordinate, and STD2 represents to seek the standard deviation of contiguous range pixel value, and GaussPyr is Gauss
Pyramid, StdPyr [l] (r, c) is that the coordinate of pyramidal l layers of standard deviation is the standard deviation of the pixel of (r, c), STD2
(GaussPyr [l] (R, C)) is the standard deviation of all coordinates (R, C) in l layers of the neighborhood centered on (r, c), STD2
(GaussPyr [l]) is the standard deviation of l tomographic images.
Further, the generation pyramidal formula of noise reduction mask is：
And wherein,
CurveCoeff=2*log (9),
Wherein, DnsMaskPyr is noise reduction mask pyramid, and l is pyramidal layer, and curveCoeff is curve coefficients,
DnsParam [l] is l layers preset of noise reduction intensity, and StdPyr [l] is l layers of standard deviation pyramid, and dnsParam is
The 0th layer preset of noise reduction intensity, dnsLevels is the noise reduction number of plies.
Further, it is to the formula that laplacian pyramid carries out noise reduction process：
DnsLaplacePyr=DnsMaskPyr*LaplacePyr,
Wherein, DnsLaplacePyr is noise reduction laplacian pyramid, and DnsMaskPyr is noise reduction mask pyramid,
LaplacePyr is laplacian pyramid.
Compared with prior art, beneficial effect is the present invention：It is pyramidal that one kind that the present invention is provided is based on standard deviation
Dynamic image noise-reduction method and device, first, algorithm are strengthened by pending dynamic image using the pyramid multiresolution overall situation
It is decomposed into gaussian pyramid and laplacian pyramid；Then, neighborhood standard deviation is calculated to every layer of gaussian pyramid, is marked
Quasi- difference pyramid, and noise reduction mask pyramid is generated according to standard deviation pyramid；Subsequently, the golden word of noise reduction mask of generation is utilized
Tower carries out noise reduction process to laplacian pyramid, obtains the laplacian pyramid after noise reduction；Finally, the drawing after noise reduction is utilized
This pyramid of pula carries out Pyramid Reconstruction, obtains the dynamic image after noise reduction；The present invention compared with prior art, passes through introducing
Standard deviation pyramid suppresses as the reference of pyramid noise reduction intensity to the noise of different layers so that utilizing pyramid
When multiresolution overall situation enhancing algorithm strengthens details, noise can be suppressed, so as to improve dynamic image noise
Than so that the display effect of dynamic image increases.
Brief description of the drawings
Fig. 1 is a kind of realization based on the pyramidal dynamic image noise-reduction method of standard deviation provided in an embodiment of the present invention
Journey schematic diagram；
Fig. 2 is a kind of flow based on the pyramidal dynamic image noise-reduction method of standard deviation provided in an embodiment of the present invention
Figure；
Fig. 3 is the schematic diagram decomposed and reconstituted to pending dynamic image progress provided in an embodiment of the present invention；
Fig. 4 is that a kind of module based on the pyramidal dynamic image denoising device of standard deviation provided in an embodiment of the present invention is shown
It is intended to.
Embodiment
In order to make the purpose , technical scheme and advantage of the present invention be clearer, it is right below in conjunction with drawings and Examples
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.
Because presence can not suppress to enter dynamic image using pyramid multiresolution overall situation enhancing algorithm in the prior art
The technical problem of the noise amplified during row enhancing processing.
In order to solve the above-mentioned technical problem, the present invention proposes a kind of based on the pyramidal dynamic image noise-reduction method of standard deviation
And device, as shown in figure 1, carrying out gold to the pending dynamic image of input using pyramid multiresolution overall situation enhancing algorithm
Word tower is decomposed, and obtains gaussian pyramid and laplacian pyramid；Neighborhood standard deviation is calculated to every layer of the gaussian pyramid,
Standard deviation pyramid is obtained, and noise reduction mask pyramid is generated according to the standard deviation pyramid；Utilize the noise reduction mask of generation
Pyramid carries out noise reduction process to the laplacian pyramid, obtains the laplacian pyramid after noise reduction；After noise reduction
Laplacian pyramid carry out Pyramid Reconstruction, obtain the dynamic image after noise reduction, and export the Dynamic Graph after the noise reduction
Picture.
One kind that the lower mask body introduction present invention is provided is based on the pyramidal dynamic image noise-reduction method of standard deviation, such as Fig. 2
It is shown, including：
Step S1, pyramid point is carried out using pyramid multiresolution overall situation enhancing algorithm to pending dynamic image
Solve, obtain N layers of gaussian pyramid and N layers of laplacian pyramid；
Specifically, because noise is typically distributed across before laplacian pyramid two layers, pending dynamic image is decomposed
Noise reduction requirement can just be met by obtaining three layers of pyramid twice, so, the embodiment of the present invention carries out pyramid point to dynamic image
Xie Shi, it is 3 to be decomposed into three layers, i.e. N values.
As shown in figure 3, to carry out three layers of pyramid point to pending dynamic image using pyramid decomposition restructing algorithm
Solve reconstruct schematic diagram, pyramid decomposition be gaussian pyramid and laplacian pyramid, wherein, gaussian pyramid be circulation pair
What pending dynamic image filtering down-sampling was got, the G in corresponding diagram0~G2, represented with gaussian pyramid；Draw pula
L in this pyramid corresponding diagram 30~L2, wherein, first two layers is by current layer gaussian pyramid and next layer of gaussian pyramid
Up-sampling subtracts each other what is obtained after filtering again, and last layer of phase of last layer of laplacian pyramid and gaussian pyramid
Together, i.e. L2=G2.Pyramidal reconstruct, is reconstructed with laplacian pyramid, since last layer, is circulated to image
The image addition of up-sampling filtering again with last layer, until the 0th layer, because pyramid decomposition reconstruct belongs to image basis knowledge,
So will not be repeated here.
Step S2, neighborhood standard deviation is calculated to every layer of the gaussian pyramid, obtains N layers of standard deviation pyramid, and
Noise reduction mask pyramid is generated according to the standard deviation pyramid；
Specifically, the corresponding pixel value of each pixel in every layer of standard deviation pyramid is the correspondence gaussian pyramid picture
The standard deviation of all pixels in the preset neighborhood of vegetarian refreshments, wherein, the preset neighborhood in the embodiment of the present invention is 3 × 3 or 5 × 5；It is right
Every layer of the gaussian pyramid is calculated after neighborhood standard deviation, by the every laminated standard deviation as obtained calculated
Pyramid；In addition, it is to the standard deviation pyramid to generate the pyramidal process of noise reduction mask according to the standard deviation pyramid
The process being normalized.
Specifically, the generation pyramidal formula of standard deviation is：
Wherein, StdPyr is standard deviation pyramid, and l is pyramidal layer, and (R, C) is in the neighborhood centered on (r, c)
All coordinates, r is row coordinate, and c is row coordinate, and STD2 represents to seek the standard deviation of contiguous range pixel value, and GaussPyr is Gauss
Pyramid, StdPyr [l] (r, c) is that the coordinate of pyramidal l layers of standard deviation is the standard deviation of the pixel of (r, c), STD2
(GaussPyr [l] (R, C)) is the standard deviation of all coordinates (R, C) in l layers of the neighborhood centered on (r, c), STD2
(GaussPyr [l]) is the standard deviation of l tomographic images.
Specifically, in the embodiment of the present invention, l takes the 0th layer of 0,1 and 2, i.e. pyramid, layers 1 and 2, due to noise one
As be distributed in before Laplce 2 layers, that is, be distributed in the 0th layer and the 1st layer, so, noise reduction process provided in an embodiment of the present invention
Last 1 layer i.e. the 2nd layer of laplacian pyramid is not operated.
Specifically, the generation pyramidal formula of noise reduction mask is：
And wherein,
CurveCoeff=2*log (9),
Wherein, DnsMaskPyr is noise reduction mask pyramid, and l is pyramidal layer, and curveCoeff is curve coefficients,
DnsParam [l] is l layers preset of noise reduction intensity, and StdPyr [l] is l layers of standard deviation pyramid, and dnsParam is
The 0th layer preset of noise reduction intensity, dnsLevels is the noise reduction number of plies.
Specifically, noise reduction intensity is weakened with the increase of the number of plies, so, it is strong according to preset 0th layer of noise reduction
Spending the 1st layer, the 2nd layer of the noise reduction intensity calculated should be in form of successively decreasing successively.In embodiments of the present invention, dnsLevels
Value is 2.
Step S3, noise reduction process is carried out using the noise reduction mask pyramid of generation to N layers of the laplacian pyramid,
Obtain the laplacian pyramid after noise reduction；
Specifically, it is to the formula that laplacian pyramid carries out noise reduction process：
DnsLaplacePyr=DnsMaskPyr*LaplacePyr,
Wherein, DnsLaplacePyr is noise reduction laplacian pyramid, and DnsMaskPyr is noise reduction mask pyramid,
LaplacePyr is laplacian pyramid.
Specifically, every layer of progress using the formula of above-mentioned laplacian pyramid noise reduction process to laplacian pyramid
Noise reduction process, so as to obtain the laplacian pyramid after noise reduction；More specifically, using pyramidal every layer of noise reduction mask with it is right
This layer of the laplacian pyramid answered, which is multiplied, can obtain the laplacian pyramid after this layer of noise reduction, and every layer all carries out noise reduction
After processing, you can obtain the laplacian pyramid after noise reduction.
Step S4, carries out Pyramid Reconstruction using the laplacian pyramid after noise reduction, obtains the dynamic image after noise reduction.
Specifically, the laplacian pyramid after noise reduction is reconstructed using previously described Pyramid Reconstruction method,
So as to get the dynamic image after noise reduction process.
It is provided in an embodiment of the present invention to be based on the pyramidal dynamic image noise-reduction method of standard deviation, by introducing standard deviation gold
Word tower suppresses as the reference of pyramid noise reduction intensity to the noise of different layers so that utilizing pyramid multiresolution
When overall situation enhancing algorithm strengthens details, noise can be suppressed, so as to improve dynamic image signal to noise ratio.
One kind that the lower mask body introduction present invention is provided is based on the pyramidal dynamic image denoising device of standard deviation, such as Fig. 4
It is shown, including：
Decomposing module 1, for carrying out gold to pending dynamic image using pyramid multiresolution overall situation enhancing algorithm
Word tower is decomposed, and obtains N layers of gaussian pyramid and N layers of laplacian pyramid；
Specifically, because noise is typically distributed across before laplacian pyramid two layers, pending dynamic image is decomposed
Noise reduction requirement can just be met by obtaining three layers of pyramid twice, so, the embodiment of the present invention carries out pyramid point to dynamic image
Xie Shi, it is 3 to be decomposed into three layers, i.e. N values.
Standard deviation pyramid generation module 2, for calculating neighborhood standard deviation to every layer of the gaussian pyramid, obtains N
The standard deviation pyramid of layer；
Specifically, the corresponding pixel value of each pixel in every layer of standard deviation pyramid is the correspondence gaussian pyramid picture
The standard deviation of all pixels in the preset neighborhood of vegetarian refreshments, wherein, the preset neighborhood in the embodiment of the present invention is 3 × 3 or 5 × 5；It is right
Every layer of the gaussian pyramid is calculated after neighborhood standard deviation, by the every laminated standard deviation as obtained calculated
Pyramid；In addition, it is to the standard deviation pyramid to generate the pyramidal process of noise reduction mask according to the standard deviation pyramid
The process being normalized.
Specifically, the generation pyramidal formula of standard deviation is：
Wherein, StdPyr is standard deviation pyramid, and l is pyramidal layer, and (R, C) is in the neighborhood centered on (r, c)
All coordinates, r is row coordinate, and c is row coordinate, and STD2 represents to seek the standard deviation of contiguous range pixel value, and GaussPyr is Gauss
Pyramid, StdPyr [l] (r, c) is that the coordinate of pyramidal l layers of standard deviation is the standard deviation of the pixel of (r, c), STD2
(GaussPyr [l] (R, C)) is the standard deviation of all coordinates (R, C) in l layers of the neighborhood centered on (r, c), STD2
(GaussPyr [l]) is the standard deviation of l tomographic images.
Specifically, in the embodiment of the present invention, l takes the 0th layer of 0,1 and 2, i.e. pyramid, layers 1 and 2, due to noise one
As be distributed in before Laplce 2 layers, that is, be distributed in the 0th layer and the 1st layer, so, noise reduction process provided in an embodiment of the present invention
Last 1 layer i.e. the 2nd layer of laplacian pyramid is not operated.
Noise reduction mask pyramid generation module 3, for generating noise reduction mask pyramid according to the standard deviation pyramid；
Specifically, the generation pyramidal formula of noise reduction mask is：
And wherein,
CurveCoeff=2*log (9),
Wherein, DnsMaskPyr is noise reduction mask pyramid, and l is pyramidal layer, and curveCoeff is curve coefficients,
DnsParam [l] is l layers preset of noise reduction intensity, and StdPyr [l] is l layers of standard deviation pyramid, and dnsParam is
The 0th layer preset of noise reduction intensity, dnsLevels is the noise reduction number of plies.
Specifically, noise reduction intensity is weakened with the increase of the number of plies, so, it is strong according to preset 0th layer of noise reduction
Spending the 1st layer, the 2nd layer of the noise reduction intensity calculated should be in form of successively decreasing successively.In embodiments of the present invention, dnsLevels
Value is 2.
Noise reduction process module 4, the laplacian pyramid for the noise reduction mask pyramid using generation to N layers enters
Row noise reduction process, obtains the laplacian pyramid after noise reduction；
It is to the formula that laplacian pyramid carries out noise reduction process：
DnsLaplacePyr=DnsMaskPyr*LaplacePyr,
Wherein, DnsLaplacePyr is noise reduction laplacian pyramid, and DnsMaskPyr is noise reduction mask pyramid,
LaplacePyr is laplacian pyramid.
Specifically, every layer of progress using the formula of above-mentioned laplacian pyramid noise reduction process to laplacian pyramid
Noise reduction process, so as to obtain the laplacian pyramid after noise reduction；More specifically, using pyramidal every layer of noise reduction mask with it is right
This layer of the laplacian pyramid answered, which is multiplied, can obtain the laplacian pyramid after this layer of noise reduction, and every layer all carries out noise reduction
After processing, you can obtain the laplacian pyramid after noise reduction.
Reconstructed module 5, for carrying out Pyramid Reconstruction using the laplacian pyramid after noise reduction, obtains dynamic after noise reduction
State image.
It is provided in an embodiment of the present invention to be based on the pyramidal dynamic image denoising device of standard deviation, by introducing standard deviation gold
Word tower suppresses as the reference of pyramid noise reduction intensity to the noise of different layers so that utilizing pyramid multiresolution
When overall situation enhancing algorithm strengthens details, noise can be suppressed, so as to improve dynamic image signal to noise ratio.
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all essences in the present invention
Any modifications, equivalent substitutions and improvements made within refreshing and principle etc., should be included in the scope of the protection.
Claims (8)
1. one kind is based on the pyramidal dynamic image noise-reduction method of standard deviation, it is characterised in that including：
Pyramid decomposition is carried out to pending dynamic image using pyramid multiresolution overall situation enhancing algorithm, N layers are obtained
Gaussian pyramid and N layers of laplacian pyramid；
Neighborhood standard deviation is calculated to every layer of the gaussian pyramid, N layers of standard deviation pyramid is obtained, and according to the standard
Poor pyramid generates noise reduction mask pyramid；
Noise reduction process is carried out to N layers of the laplacian pyramid using the noise reduction mask pyramid of generation, obtained after noise reduction
Laplacian pyramid；
Pyramid Reconstruction is carried out using the laplacian pyramid after noise reduction, the dynamic image after noise reduction is obtained.
2. dynamic image noise-reduction method as claimed in claim 1, it is characterised in that generating the pyramidal formula of standard deviation is：
<mrow>
<mi>S</mi>
<mi>t</mi>
<mi>d</mi>
<mi>P</mi>
<mi>y</mi>
<mi>r</mi>
<mo>&lsqb;</mo>
<mi>l</mi>
<mo>&rsqb;</mo>
<mrow>
<mo>(</mo>
<mi>r</mi>
<mo>,</mo>
<mi>c</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mrow>
<mi>S</mi>
<mi>T</mi>
<mi>D</mi>
<mn>2</mn>
<mrow>
<mo>(</mo>
<mi>G</mi>
<mi>a</mi>
<mi>u</mi>
<mi>s</mi>
<mi>s</mi>
<mi>P</mi>
<mi>y</mi>
<mi>r</mi>
<mo>&lsqb;</mo>
<mi>l</mi>
<mo>&rsqb;</mo>
<mo>(</mo>
<mi>R</mi>
<mo>,</mo>
<mi>C</mi>
<mo>)</mo>
</mrow>
<mo>)</mo>
</mrow>
<mrow>
<mi>S</mi>
<mi>T</mi>
<mi>D</mi>
<mn>2</mn>
<mrow>
<mo>(</mo>
<mi>G</mi>
<mi>a</mi>
<mi>u</mi>
<mi>s</mi>
<mi>s</mi>
<mi>P</mi>
<mi>y</mi>
<mi>r</mi>
<mo>&lsqb;</mo>
<mi>l</mi>
<mo>&rsqb;</mo>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
<mo>,</mo>
</mrow>
Wherein, StdPyr is standard deviation pyramid, and l is pyramidal layer, and (R, C) is all in the neighborhood centered on (r, c)
Coordinate, r is row coordinate, and c is row coordinate, and STD2 represents to seek the standard deviation of contiguous range pixel value, and GaussPyr is the golden word of Gauss
Tower, StdPyr [l] (r, c) is that the coordinate of pyramidal l layers of standard deviation is the standard deviation of the pixel of (r, c), STD2
(GaussPyr [l] (R, C)) is the standard deviation of all coordinates (R, C) in l layers of the neighborhood centered on (r, c), STD2
(GaussPyr [l]) is the standard deviation of l tomographic images.
3. dynamic image noise-reduction method as claimed in claim 2, it is characterised in that the generation pyramidal formula of noise reduction mask
For：
<mrow>
<mi>D</mi>
<mi>n</mi>
<mi>s</mi>
<mi>M</mi>
<mi>a</mi>
<mi>s</mi>
<mi>k</mi>
<mi>P</mi>
<mi>y</mi>
<mi>r</mi>
<mo>&lsqb;</mo>
<mi>l</mi>
<mo>&rsqb;</mo>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mrow>
<mn>1</mn>
<mo>+</mo>
<msup>
<mi>exp</mi>
<mrow>
<mi>c</mi>
<mi>u</mi>
<mi>r</mi>
<mi>v</mi>
<mi>e</mi>
<mi>C</mi>
<mi>o</mi>
<mi>e</mi>
<mi>f</mi>
<mi>f</mi>
<mo>*</mo>
<mrow>
<mo>(</mo>
<mi>D</mi>
<mi>n</mi>
<mi>s</mi>
<mi>P</mi>
<mi>a</mi>
<mi>r</mi>
<mi>a</mi>
<mi>m</mi>
<mo>&lsqb;</mo>
<mi>l</mi>
<mo>&rsqb;</mo>
<mo>-</mo>
<mi>S</mi>
<mi>t</mi>
<mi>d</mi>
<mi>P</mi>
<mi>y</mi>
<mi>r</mi>
<mo>&lsqb;</mo>
<mi>l</mi>
<mo>&rsqb;</mo>
<mo>)</mo>
</mrow>
</mrow>
</msup>
</mrow>
</mfrac>
<mo>,</mo>
</mrow>
And wherein,
CurveCoeff=2*log (9),
<mrow>
<mi>D</mi>
<mi>n</mi>
<mi>s</mi>
<mi>P</mi>
<mi>a</mi>
<mi>r</mi>
<mi>a</mi>
<mi>m</mi>
<mo>&lsqb;</mo>
<mi>l</mi>
<mo>&rsqb;</mo>
<mo>=</mo>
<mn>2</mn>
<mo>*</mo>
<mrow>
<mo>(</mo>
<mi>d</mi>
<mi>n</mi>
<mi>s</mi>
<mi>P</mi>
<mi>a</mi>
<mi>r</mi>
<mi>a</mi>
<mi>m</mi>
<mo>-</mo>
<mo>(</mo>
<mrow>
<mi>l</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
<mo>)</mo>
<mo>*</mo>
<mfrac>
<mrow>
<mi>d</mi>
<mi>n</mi>
<mi>s</mi>
<mi>P</mi>
<mi>a</mi>
<mi>r</mi>
<mi>a</mi>
<mi>m</mi>
</mrow>
<mrow>
<mi>d</mi>
<mi>n</mi>
<mi>s</mi>
<mi>L</mi>
<mi>e</mi>
<mi>v</mi>
<mi>e</mi>
<mi>l</mi>
<mi>s</mi>
</mrow>
</mfrac>
<mo>)</mo>
</mrow>
<mo>-</mo>
<mn>0.5</mn>
<mo>,</mo>
</mrow>
Wherein, DnsMaskPyr is noise reduction mask pyramid, and l is pyramidal layer, and curveCoeff is curve coefficients, DnsParam
[l] is l layers preset of noise reduction intensity, and StdPyr [l] is l layers of standard deviation pyramid, and dnsParam is the preset the 0th
The noise reduction intensity of layer, dnsLevels is the noise reduction number of plies.
4. dynamic image noise-reduction method as claimed in claim 3, it is characterised in that carried out to laplacian pyramid at noise reduction
The formula of reason is：
DnsLaplacePyr=DnsMaskPyr*LaplacePyr,
Wherein, DnsLaplacePyr is noise reduction laplacian pyramid, and DnsMaskPyr is noise reduction mask pyramid,
LaplacePyr is laplacian pyramid.
5. one kind is based on the pyramidal dynamic image denoising device of standard deviation, it is characterised in that including：
Decomposing module, for carrying out pyramid point to pending dynamic image using pyramid multiresolution overall situation enhancing algorithm
Solve, obtain N layers of gaussian pyramid and N layers of laplacian pyramid；
Standard deviation pyramid generation module, for calculating neighborhood standard deviation to every layer of the gaussian pyramid, obtains N layers of mark
Quasi- difference pyramid；
Noise reduction mask pyramid generation module, for generating noise reduction mask pyramid according to the standard deviation pyramid；
Noise reduction process module, the laplacian pyramid for the noise reduction mask pyramid using generation to N layers drops
Make an uproar processing, obtain the laplacian pyramid after noise reduction；
Reconstructed module, for carrying out Pyramid Reconstruction using the laplacian pyramid after noise reduction, obtains the Dynamic Graph after noise reduction
Picture.
6. dynamic image denoising device as claimed in claim 5, it is characterised in that generating the pyramidal formula of standard deviation is：
<mrow>
<mi>S</mi>
<mi>t</mi>
<mi>d</mi>
<mi>P</mi>
<mi>y</mi>
<mi>r</mi>
<mo>&lsqb;</mo>
<mi>l</mi>
<mo>&rsqb;</mo>
<mrow>
<mo>(</mo>
<mi>r</mi>
<mo>,</mo>
<mi>c</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mrow>
<mi>S</mi>
<mi>T</mi>
<mi>D</mi>
<mn>2</mn>
<mrow>
<mo>(</mo>
<mi>G</mi>
<mi>a</mi>
<mi>u</mi>
<mi>s</mi>
<mi>s</mi>
<mi>P</mi>
<mi>y</mi>
<mi>r</mi>
<mo>&lsqb;</mo>
<mi>l</mi>
<mo>&rsqb;</mo>
<mo>(</mo>
<mi>R</mi>
<mo>,</mo>
<mi>C</mi>
<mo>)</mo>
</mrow>
<mo>)</mo>
</mrow>
<mrow>
<mi>S</mi>
<mi>T</mi>
<mi>D</mi>
<mn>2</mn>
<mrow>
<mo>(</mo>
<mi>G</mi>
<mi>a</mi>
<mi>u</mi>
<mi>s</mi>
<mi>s</mi>
<mi>P</mi>
<mi>y</mi>
<mi>r</mi>
<mo>&lsqb;</mo>
<mi>l</mi>
<mo>&rsqb;</mo>
<mo>)</mo>
</mrow>
</mrow>
</mfrac>
<mo>,</mo>
</mrow>
Wherein, StdPyr is standard deviation pyramid, and l is pyramidal layer, and (R, C) is all in the neighborhood centered on (r, c)
Coordinate, r is row coordinate, and c is row coordinate, and STD2 represents to seek the standard deviation of contiguous range pixel value, and GaussPyr is the golden word of Gauss
Tower, StdPyr [l] (r, c) is that the coordinate of pyramidal l layers of standard deviation is the standard deviation of the pixel of (r, c), STD2
(GaussPyr [l] (R, C)) is the standard deviation of all coordinates (R, C) in l layers of the neighborhood centered on (r, c), STD2
(GaussPyr [l]) is the standard deviation of l tomographic images.
7. dynamic image denoising device as claimed in claim 6, it is characterised in that the generation pyramidal formula of noise reduction mask
For：
<mrow>
<mi>D</mi>
<mi>n</mi>
<mi>s</mi>
<mi>M</mi>
<mi>a</mi>
<mi>s</mi>
<mi>k</mi>
<mi>P</mi>
<mi>y</mi>
<mi>r</mi>
<mo>&lsqb;</mo>
<mi>l</mi>
<mo>&rsqb;</mo>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mrow>
<mn>1</mn>
<mo>+</mo>
<msup>
<mi>exp</mi>
<mrow>
<mi>c</mi>
<mi>u</mi>
<mi>r</mi>
<mi>v</mi>
<mi>e</mi>
<mi>C</mi>
<mi>o</mi>
<mi>e</mi>
<mi>f</mi>
<mi>f</mi>
<mo>*</mo>
<mrow>
<mo>(</mo>
<mi>D</mi>
<mi>n</mi>
<mi>s</mi>
<mi>P</mi>
<mi>a</mi>
<mi>r</mi>
<mi>a</mi>
<mi>m</mi>
<mo>&lsqb;</mo>
<mi>l</mi>
<mo>&rsqb;</mo>
<mo>-</mo>
<mi>S</mi>
<mi>t</mi>
<mi>d</mi>
<mi>P</mi>
<mi>y</mi>
<mi>r</mi>
<mo>&lsqb;</mo>
<mi>l</mi>
<mo>&rsqb;</mo>
<mo>)</mo>
</mrow>
</mrow>
</msup>
</mrow>
</mfrac>
<mo>,</mo>
</mrow>
And wherein,
CurveCoeff=2*log (9),
<mrow>
<mi>D</mi>
<mi>n</mi>
<mi>s</mi>
<mi>P</mi>
<mi>a</mi>
<mi>r</mi>
<mi>a</mi>
<mi>m</mi>
<mo>&lsqb;</mo>
<mi>l</mi>
<mo>&rsqb;</mo>
<mo>=</mo>
<mn>2</mn>
<mo>*</mo>
<mrow>
<mo>(</mo>
<mi>d</mi>
<mi>n</mi>
<mi>s</mi>
<mi>P</mi>
<mi>a</mi>
<mi>r</mi>
<mi>a</mi>
<mi>m</mi>
<mo>-</mo>
<mo>(</mo>
<mi>l</mi>
<mo>-</mo>
<mn>1</mn>
<mo>)</mo>
</mrow>
<mo>*</mo>
<mfrac>
<mrow>
<mi>d</mi>
<mi>n</mi>
<mi>s</mi>
<mi>P</mi>
<mi>a</mi>
<mi>r</mi>
<mi>a</mi>
<mi>m</mi>
</mrow>
<mrow>
<mi>d</mi>
<mi>n</mi>
<mi>s</mi>
<mi>L</mi>
<mi>e</mi>
<mi>v</mi>
<mi>e</mi>
<mi>l</mi>
<mi>s</mi>
</mrow>
</mfrac>
<mo>)</mo>
<mo>-</mo>
<mn>0.5</mn>
<mo>,</mo>
</mrow>
Wherein, DnsMaskPyr is noise reduction mask pyramid, and l is pyramidal layer, and curveCoeff is curve coefficients, DnsParam
[l] is l layers preset of noise reduction intensity, and StdPyr [l] is l layers of standard deviation pyramid, and dnsParam is the preset the 0th
The noise reduction intensity of layer, dnsLevels is the noise reduction number of plies.
8. dynamic image denoising device as claimed in claim 7, it is characterised in that carried out to laplacian pyramid at noise reduction
The formula of reason is：
DnsLaplacePyr=DnsMaskPyr*LaplacePyr,
Wherein, DnsLaplacePyr is noise reduction laplacian pyramid, and DnsMaskPyr is noise reduction mask pyramid,
LaplacePyr is laplacian pyramid.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710492019.6A CN107292845B (en) | 2017-06-26 | 2017-06-26 | Standard deviation pyramid-based dynamic image noise reduction method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710492019.6A CN107292845B (en) | 2017-06-26 | 2017-06-26 | Standard deviation pyramid-based dynamic image noise reduction method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107292845A true CN107292845A (en) | 2017-10-24 |
| CN107292845B CN107292845B (en) | 2020-04-17 |
Family
ID=60099575
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710492019.6A Active CN107292845B (en) | 2017-06-26 | 2017-06-26 | Standard deviation pyramid-based dynamic image noise reduction method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107292845B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108230274A (en) * | 2018-01-16 | 2018-06-29 | 华南理工大学 | Multiresolution rapid denoising method and device under a kind of mixed noise model |
| CN109801279A (en) * | 2019-01-21 | 2019-05-24 | 京东方科技集团股份有限公司 | Object detection method and device, electronic equipment, storage medium in image |
| CN111260580A (en) * | 2020-01-17 | 2020-06-09 | 珠海全志科技股份有限公司 | Image denoising method based on image pyramid, computer device and computer readable storage medium |
| CN111882504A (en) * | 2020-08-05 | 2020-11-03 | 展讯通信（上海）有限公司 | Method and system for processing color noise in image, electronic device and storage medium |
| CN115205738A (en) * | 2022-07-05 | 2022-10-18 | 广州和达水务科技股份有限公司 | Emergency drainage method and system applied to urban inland inundation |
| CN115841425A (en) * | 2022-07-21 | 2023-03-24 | 爱芯元智半导体(上海)有限公司 | Video noise reduction method and device, electronic equipment and computer readable storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101425137A (en) * | 2008-11-10 | 2009-05-06 | 北方工业大学 | Face Image Fusion Method Based on Laplacian Pyramid |
| CN102646269A (en) * | 2012-02-29 | 2012-08-22 | 中山大学 | Image processing method and device based on Laplace pyramid |
| CN103778616A (en) * | 2012-10-22 | 2014-05-07 | 中国科学院研究生院 | Contrast pyramid image fusion method based on area |
| CN104616273A (en) * | 2015-01-26 | 2015-05-13 | 电子科技大学 | Multi-exposure image fusion method based on Laplacian pyramid decomposition |
| CN104637060A (en) * | 2015-02-13 | 2015-05-20 | 武汉工程大学 | Image partition method based on neighbor-hood PCA (Principal Component Analysis)-Laplace |
| CN105225213A (en) * | 2015-09-29 | 2016-01-06 | 云南大学 | A kind of Color Image Fusion method based on S-PCNN and laplacian pyramid |
-
2017
- 2017-06-26 CN CN201710492019.6A patent/CN107292845B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101425137A (en) * | 2008-11-10 | 2009-05-06 | 北方工业大学 | Face Image Fusion Method Based on Laplacian Pyramid |
| CN102646269A (en) * | 2012-02-29 | 2012-08-22 | 中山大学 | Image processing method and device based on Laplace pyramid |
| CN103778616A (en) * | 2012-10-22 | 2014-05-07 | 中国科学院研究生院 | Contrast pyramid image fusion method based on area |
| CN104616273A (en) * | 2015-01-26 | 2015-05-13 | 电子科技大学 | Multi-exposure image fusion method based on Laplacian pyramid decomposition |
| CN104637060A (en) * | 2015-02-13 | 2015-05-20 | 武汉工程大学 | Image partition method based on neighbor-hood PCA (Principal Component Analysis)-Laplace |
| CN105225213A (en) * | 2015-09-29 | 2016-01-06 | 云南大学 | A kind of Color Image Fusion method based on S-PCNN and laplacian pyramid |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108230274A (en) * | 2018-01-16 | 2018-06-29 | 华南理工大学 | Multiresolution rapid denoising method and device under a kind of mixed noise model |
| CN109801279A (en) * | 2019-01-21 | 2019-05-24 | 京东方科技集团股份有限公司 | Object detection method and device, electronic equipment, storage medium in image |
| CN111260580A (en) * | 2020-01-17 | 2020-06-09 | 珠海全志科技股份有限公司 | Image denoising method based on image pyramid, computer device and computer readable storage medium |
| CN111882504A (en) * | 2020-08-05 | 2020-11-03 | 展讯通信（上海）有限公司 | Method and system for processing color noise in image, electronic device and storage medium |
| CN111882504B (en) * | 2020-08-05 | 2022-08-16 | 展讯通信（上海）有限公司 | Method and system for processing color noise in image, electronic device and storage medium |
| CN115205738A (en) * | 2022-07-05 | 2022-10-18 | 广州和达水务科技股份有限公司 | Emergency drainage method and system applied to urban inland inundation |
| CN115841425A (en) * | 2022-07-21 | 2023-03-24 | 爱芯元智半导体(上海)有限公司 | Video noise reduction method and device, electronic equipment and computer readable storage medium |
| CN115841425B (en) * | 2022-07-21 | 2023-11-17 | 爱芯元智半导体(宁波)有限公司 | Video noise reduction method and device, electronic equipment and computer readable storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107292845B (en) | 2020-04-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107292845A (en) | Based on the pyramidal dynamic image noise-reduction method of standard deviation and device | |
| Jiang et al. | Unsupervised decomposition and correction network for low-light image enhancement | |
| CN104715461B (en) | Image de-noising method | |
| CN103854264B (en) | A kind of Wavelet Transformation of Image Denoising method based on modified threshold function table | |
| CN104616273A (en) | Multi-exposure image fusion method based on Laplacian pyramid decomposition | |
| EP3166070A1 (en) | Method for upscaling noisy images, and apparatus for upscaling noisy images | |
| CN104182939B (en) | Medical image detail enhancement method | |
| CN103761719B (en) | A kind of adaptive wavelet threshold denoising method based on neighborhood relevance | |
| CN104166974A (en) | CT locating film image enhancing method and CT locating film image enhancing device | |
| CN101639936A (en) | X-ray image enhancing method and system thereof | |
| KR20140042809A (en) | X-ray computer tomography device and image reconstructing method | |
| CN103500442A (en) | X-ray image multi-scale detail enhancement method in integrated circuit packaging | |
| CN102999884A (en) | Image processing device and method | |
| CN112837244B (en) | Low-dose CT image denoising and artifact removing method based on progressive generation confrontation network | |
| CN101491080A (en) | Curvature-preserving filters for denoising and controlled deblurring of images | |
| US8005288B2 (en) | Layer reconstruction from dual-energy image pairs | |
| CN107274372A (en) | Dynamic image Enhancement Method and device based on pyramid local contrast | |
| CN104156931B (en) | A kind of digital subtraction angiography method | |
| CN107194903A (en) | A kind of multi-focus image fusing method based on wavelet transformation | |
| CN112085809A (en) | Neural network training method, method and device for eliminating metal artifacts | |
| CN102262775A (en) | Image denoising method and device and image processing equipment | |
| CN102682439B (en) | Medical image fusion method based on multidirectional empirical mode decomposition | |
| CN114998154A (en) | Low-dose CT image denoising method based on transformer and multi-scale features | |
| CN106233327A (en) | Recovery to low contrast structure in denoising view data | |
| CN107451986A (en) | Single infrared image enhancement method based on fusion technology |
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 | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: 404100 1-4, 1st floor, building 10, building 8, No. 26, Jiulongyuan Avenue, Jiulongpo District, Chongqing Applicant after: Ankian Technology (Chongqing) Co., Ltd Address before: 404100 1-4, 1st floor, building 10, building 8, No. 26, Jiulongyuan Avenue, Jiulongpo District, Chongqing Applicant before: Chongqing map Medical Equipment Co., Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |