CN104463785A - Method and device for amplifying ultrasound image - Google Patents
Method and device for amplifying ultrasound image Download PDFInfo
- Publication number
- CN104463785A CN104463785A CN201410668306.4A CN201410668306A CN104463785A CN 104463785 A CN104463785 A CN 104463785A CN 201410668306 A CN201410668306 A CN 201410668306A CN 104463785 A CN104463785 A CN 104463785A
- Authority
- CN
- China
- Prior art keywords
- image
- interpolation
- ultrasound image
- frequency sub
- raw ultrasound
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformation in the plane of the image
- G06T3/40—Scaling the whole image or part thereof
- G06T3/4007—Interpolation-based scaling, e.g. bilinear interpolation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformation in the plane of the image
- G06T3/40—Scaling the whole image or part thereof
- G06T3/4084—Transform-based scaling, e.g. FFT domain scaling
Abstract
The invention belongs to the technical field of image processing, and provides a method and device for amplifying an ultrasound image. The method includes the steps that an original ultrasound image to be amplified is acquired; the high-frequency component of the original ultrasound image is decomposed according to a plurality of preset decomposition directions, so that high-frequency sub-band images in multiple directions are acquired; interpolation is conducted on the high-frequency sub-band image in each direction in the high-frequency sub-band images in the multiple directions, and the high-frequency sub-band amplified image in each direction is acquired; interpolation is conducted on the original ultrasound image, and an interpolation image of the original ultrasound image is acquired; frequency domain reconstruction is conducted on the acquired high-frequency sub-band amplified image in each direction and the acquired interpolation image of the original ultrasound image, and accordingly the amplified ultrasound image is acquired. The amplified ultrasound image can keep the details of the original image, ghosting-artifacts can be avoided, the medical clinic requirement can be met, and doctor diagnosis can be facilitated.
Description
Technical field
The invention belongs to technical field of image processing, particularly relate to a kind of amplification method and device of ultrasonoscopy.
Background technology
The object of Nonlinear magnify obtains high-definition picture, i.e. Super-resolution Reconstruction from low-resolution image.Super-resolution ultrasonoscopy reconstruction technique has important using value at clinical medicine.The generation method of existing super resolution image has three classes, is respectively:
1. pair single frames low-resolution image carries out interpolation: such as classical linear interpolation, cubic interpolation, this kind of amplification method is simple, is widely used, but easily causes the image blur after interpolation.
2. pair multiframe low-resolution image is rebuild: such as iterative backprojection (IBP, Iterative BackProjection), convex set projection (POCS, Projections Onto Convex Sets), these methods all have the shortcoming of edge vibration, easy generation pseudomorphism, is unfavorable for the diagnosis of doctor.Although the IBP algorithm after improving can reduce the oscillation effect of IBP with POCS algorithm, everyly can not eliminate pseudomorphism completely.
3. training study modeling: set up relational model between low-resolution image and high-definition picture by collecting a large amount of training sets (comprising the high-definition picture of low-resolution image and correspondence thereof), but the method needs complete training set and learning method, realizes difficulty large.
In sum, the multiple image in existing ultrasonic image magnifying method is rebuild and is easily produced pseudomorphism, and calculated amount is large; Simple single-frame images interpolation amplification effect is difficult to meet clinical requirement; And modeling implements difficulty greatly, be difficult to meet actual application.
Summary of the invention
Given this, the embodiment of the present invention provides a kind of amplification method and device of ultrasonoscopy, cause to solve existing ultrasonic image magnifying method the problem that the ultrasonoscopy after amplifying is fuzzy, produce pseudomorphism, make the ultrasonoscopy after amplifying can keep the details of original image.
First aspect, provides a kind of amplification method of ultrasonoscopy, and described method comprises:
Obtain raw ultrasound image to be amplified;
Decompose by the high fdrequency component of multiple decomposition directions to described raw ultrasound image preset, obtain the high-frequency sub-band images of multiple directions;
Interpolation is carried out to the high-frequency sub-band images in each direction in the high-frequency sub-band images of described multiple directions, obtains the high-frequency sub-band enlarged image in each direction;
Interpolation is carried out to raw ultrasound image, obtains the interpolation image of raw ultrasound image;
Reconstruction in frequency-domain is carried out to the high-frequency sub-band enlarged image in each obtained direction and the interpolation image of raw ultrasound image.
Further, the described high fdrequency component of multiple decomposition directions to described raw ultrasound image by presetting is decomposed, and the high-frequency sub-band images obtaining multiple directions comprises:
Fourier transform is carried out to described raw ultrasound image to be amplified, obtains the frequency spectrum of described raw ultrasound image;
Extract the radio-frequency component in described frequency spectrum, obtain the high fdrequency component of raw ultrasound image;
By multiple decomposition directions of presetting, described high fdrequency component is decomposed, obtain the high-frequency sub-band images of multiple directions.
Further, describedly carry out interpolation to the high-frequency sub-band images in each direction in the high-frequency sub-band images of described multiple directions, the high-frequency sub-band enlarged image obtaining each direction comprises:
Sharpening is carried out to the high-frequency sub-band images in each direction in the high-frequency sub-band images of described multiple directions;
Along corresponding direction, interpolation is carried out to the high-frequency sub-band images in each direction after sharpening, obtains the high-frequency sub-band enlarged image in each direction;
Describedly carry out interpolation to raw ultrasound image, the interpolation image obtaining raw ultrasound image is specially:
Newton interpolation is carried out to raw ultrasound image, obtains the interpolation image of raw ultrasound image.
Further, describedly reconstruction in frequency-domain carried out to the high-frequency sub-band enlarged image in each obtained direction and the interpolation image of raw ultrasound image comprise:
Adopt the high-frequency sub-band enlarged image of high frequency direction wave filter to each direction of reconstruction filter banks to carry out filtering, obtain the high-frequency sub-band reconstructed image in each direction;
Adopt the low frequency anisotropic filter of reconstruction filter banks to carry out filtering to the interpolation image of raw ultrasound image, obtain the reconstructed image of raw ultrasound image;
By described each party to high-frequency sub-band reconstructed image and the reconstructed image of raw ultrasound image superpose, obtain superimposed image;
Inverse Fourier transform is carried out to described superimposed image, obtains the ultrasonoscopy after amplifying.
Further, describedly carry out Newton interpolation to raw ultrasound image, the interpolation image obtaining raw ultrasound image comprises:
Obtain the multiple interpolation points in raw ultrasound image;
Along 0,45 °, 90 ° and 135 ° of four directions, Newton interpolation is carried out to described multiple interpolation point respectively, obtain the interpolation image in each direction;
By the weights in each direction of presetting, the interpolation image of four direction is sued for peace, obtain the interpolation image of raw ultrasound image
Second aspect, provides a kind of multiplying arrangement of ultrasonoscopy, and described device comprises:
Acquisition module, for obtaining raw ultrasound image to be amplified;
Decomposing module, for decomposing by the high fdrequency component of multiple decomposition directions to described raw ultrasound image preset, obtains the high-frequency sub-band images of multiple directions;
Interpolating module, interpolation is carried out for the high-frequency sub-band images in each direction in the high-frequency sub-band images to described multiple directions, obtain the high-frequency sub-band enlarged image in each direction, and interpolation is carried out to raw ultrasound image, obtain the interpolation image of raw ultrasound image;
Reconstruction in frequency-domain module, for carrying out reconstruction in frequency-domain to the high-frequency sub-band enlarged image in each obtained direction and the interpolation image of raw ultrasound image.
Further, described decomposing module comprises:
Converter unit, for carrying out Fourier transform to described raw ultrasound image to be amplified, obtains the frequency spectrum of described raw ultrasound image;
Extraction unit, for extracting the radio-frequency component in described frequency spectrum, obtains the high fdrequency component of raw ultrasound image;
Resolving cell, for decomposing described high fdrequency component by multiple decomposition directions of presetting, obtains the high-frequency sub-band images of multiple directions.
Further, described interpolating module comprises:
Sharpening unit, carries out sharpening for the high-frequency sub-band images in each direction in the high-frequency sub-band images to described multiple directions;
First interpolating unit, for carrying out interpolation to the high-frequency sub-band images in each direction after sharpening along corresponding direction, obtains the high-frequency sub-band enlarged image in each direction;
Second interpolating unit, for carrying out Newton interpolation to raw ultrasound image, obtains the interpolation image of raw ultrasound image.
Further, described reconstruction in frequency-domain module comprises:
First reconstruction in frequency-domain unit, for adopting the high-frequency sub-band enlarged image of the high frequency direction wave filter of reconstruction filter banks to each direction to carry out filtering, obtains the high-frequency sub-band reconstructed image in each direction;
Second reconstruction in frequency-domain unit, for adopting the low frequency anisotropic filter of reconstruction filter banks to carry out filtering to the interpolation image of raw ultrasound image, obtains the reconstructed image of raw ultrasound image;
Superpositing unit, for the high-frequency sub-band reconstructed image in each direction described and the reconstructed image of raw ultrasound image being superposed, obtains superimposed image;
Inverse transformation block, for carrying out inverse Fourier transform to described superimposed image, obtains the ultrasonoscopy after amplifying.
Further, described second interpolating unit specifically for:
Obtain the multiple interpolation points in raw ultrasound image;
Along 0,45 °, 90 ° and 135 ° of four directions, Newton interpolation is carried out to described multiple interpolation point respectively, obtain the interpolation image in each direction;
By the weights in each direction of presetting, the interpolation image of four direction is sued for peace, obtain the interpolation image of raw ultrasound image.
Compared with prior art, the present invention, after getting raw ultrasound image to be amplified, decomposes by the high fdrequency component of multiple decomposition directions to described raw ultrasound image preset, and obtains the high-frequency sub-band images of multiple directions; Interpolation is carried out to the high-frequency sub-band images in each direction in the high-frequency sub-band images of described multiple directions, obtains the high-frequency sub-band enlarged image in each direction, and interpolation is carried out to raw ultrasound image, obtain the interpolation image of raw ultrasound image; Finally reconstruction in frequency-domain is carried out to the high-frequency sub-band enlarged image in each obtained direction and the interpolation image of raw ultrasound image, thus obtain the ultrasonoscopy after amplifying.Ultrasonoscopy after described amplification can keep the details of original image, and can not produce pseudomorphism, meets clinical medicine requirement, is conducive to the diagnosis of doctor.
Accompanying drawing explanation
In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, be 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, other accompanying drawing can also be obtained according to these accompanying drawings.
Fig. 1 is the realization flow figure of the ultrasonic image magnifying method that the embodiment of the present invention one provides;
Fig. 2 is the specific implementation flow process of step S102 in the ultrasonic image magnifying method that provides of the embodiment of the present invention one;
Fig. 3 is 8 decomposition direction schematic diagram that the embodiment of the present invention one provides;
Fig. 4 is the analysis filter bank in 8 decomposition directions that the embodiment of the present invention one provides;
Fig. 5 is the specific implementation flow process of step S103 in the ultrasonic image magnifying method that provides of the embodiment of the present invention one;
Fig. 6 is the specific implementation flow process of step S503 in the ultrasonic image magnifying method that provides of the embodiment of the present invention one;
Fig. 7 is the specific implementation process flow diagram carrying out Newton interpolation along a concrete direction that the embodiment of the present invention one provides;
Fig. 8 is interpolation point p (x, y) that provides of the embodiment of the present invention one and the interpolation integral point P on 45 ° of directions
1, P
2, P
3, P
4position view;
Fig. 9 is the specific implementation flow process of step S104 in the ultrasonic image magnifying method that provides of the embodiment of the present invention one;
Figure 10 is the composition structural drawing of the ultrasonoscopy multiplying arrangement that the embodiment of the present invention two provides.
Embodiment
In order to make object of the present invention, technical scheme and advantage clearly understand, below in conjunction with drawings and Examples, the present invention is further elaborated.Should be appreciated that specific embodiment described herein only in order to explain the present invention, be not intended to limit the present invention.
The embodiment of the present invention, after getting raw ultrasound image to be amplified, is decomposed by the high fdrequency component of multiple decomposition directions to described raw ultrasound image preset, and obtains the high-frequency sub-band images of multiple directions; Interpolation is carried out to the high-frequency sub-band images in each direction in the high-frequency sub-band images of described multiple directions, obtains the high-frequency sub-band enlarged image in each direction, and interpolation is carried out to raw ultrasound image, obtain the interpolation image of raw ultrasound image; Finally reconstruction in frequency-domain is carried out to the high-frequency sub-band enlarged image in each obtained direction and the interpolation image of raw ultrasound image, thus obtain the ultrasonoscopy after amplifying.Ultrasonoscopy after described amplification can keep the details of original image, and can not ultrasonic artifacts, meets clinical medicine requirement, is conducive to the diagnosis of doctor.The embodiment of the present invention additionally provides corresponding device, is described in detail respectively below.
embodiment one
Fig. 1 shows the first realization flow of the amplification method of the ultrasonoscopy that the embodiment of the present invention one provides, and for convenience of explanation, illustrate only part related to the present invention.
As shown in Figure 1, described method comprises:
In step S101, obtain raw ultrasound image to be amplified.
In embodiments of the present invention, the raw ultrasound image after log-compressed, dynamic range transform, digital scan conversion is gathered in advance.After collecting raw ultrasound image, receive the selection instruction of user, obtain the image-region (i.e. ROI image, Region of Interesting) that selection instruction is corresponding, be defined as raw ultrasound image to be amplified; Or the selection instruction according to presetting selects ROI image automatically.Show because the image after amplification can substitute original image, if the image after amplifying is excessive, its marginal portion may be hidden; And by choosing ROI image in advance, then can avoid the problem that marginal portion that is that cause is hidden because the image after amplifying is excessive.
In step s 102, decompose by the high fdrequency component of multiple decomposition directions to described raw ultrasound image preset, obtain the high-frequency sub-band images of multiple directions.
In embodiments of the present invention, preferred employing non-downsampling Contourlet conversion (nonsubsampledcontourlet transform, NSCT), by tower bank of filters (nonsubsampled pyramidfilter bank, NSPFB) multi-resolution decomposition is carried out to image, obtain high fdrequency component and low frequency component, and then adopt non-lower sampling directional filter banks (nonsubsampled directional filter bank, NSDFB) to decompose the high fdrequency component travel direction obtained.The specific implementation flow process of step S102 will be provided below.
As shown in Figure 2, the specific implementation flow process of step S102 comprises:
In step s 201, Fourier transform is carried out to described raw ultrasound image to be amplified, obtain the frequency spectrum of described raw ultrasound image to be amplified.
By carrying out Fourier transform to raw ultrasound image to be amplified, described ultrasonoscopy to be amplified being transformed to frequency domain, carrying out frequency-domain operations so that follow-up.
In step S202, extract the radio-frequency component in described frequency spectrum, obtain the high fdrequency component of raw ultrasound image.
In the present embodiment, after getting the frequency spectrum of raw ultrasound image to be amplified, transfer the Hi-pass filter group for carrying out Scale Decomposition.Extracted the radio-frequency component in described frequency spectrum by described Hi-pass filter group, expression formula is F
h=F*H
d, wherein, F is the frequency spectrum after raw ultrasound image Fourier transform to be amplified, H
dfor the Hi-pass filter group of decomposing, F
hit is filtered radio-frequency component.
Described bank of filters is preferably tower bank of filters, but is not limited to tower bank of filters, can also be other bank of filters.
In step S203, by multiple decomposition directions of presetting, described high fdrequency component is decomposed, obtain the high-frequency sub-band images of multiple directions.
In embodiments of the present invention, obtain directional filter banks, use described directional filter banks to carry out filtering to the high fdrequency component got in step S202, expression formula is: F
i=F
h* D
i.Wherein, F
hfor high fdrequency component, D
ii-th anisotropic filter, F
ifor the high-frequency sub-band images in i-th direction after Directional Decomposition.By step S203, thus the high-frequency sub-band images of multiple directions can be obtained.Direction described in the embodiment of the present invention comprises positive dirction and the negative direction of a direction.
Exemplary, the embodiment of the present invention is preferably decomposed direction by 8 and is decomposed described high fdrequency component, to obtain the high-frequency sub-band images in 8 directions.Described 8 decomposition directions can be designed to decomposing schematic representation by all directions shown in Fig. 3.Accordingly, 8 are decomposed the analysis filter bank in direction as shown in Figure 4, and wherein, ground floor is from left to right respectively direction 1, direction 2, direction 3, direction 4; The second layer is from left to right respectively to 5, direction 6, direction 7, direction 8.
In step s 103, interpolation is carried out to the high-frequency sub-band images in each direction in the high-frequency sub-band images of described multiple directions, obtains the high-frequency sub-band enlarged image in each direction.
In step S104, interpolation is carried out to raw ultrasound image, obtain the interpolation image of raw ultrasound image.
The embodiment of the present invention is after getting the high-frequency sub-band images of multiple directions, obtain the enlargement factor pre-set, and respectively interpolation is carried out, to obtain high-frequency sub-band images and the raw ultrasound image in each direction of target sizes to the high-frequency sub-band images in each direction and raw ultrasound image to be amplified according to described enlargement factor.The specific implementation flow process of step S103 will be provided below.
As shown in Figure 5, the specific implementation flow process of step S103 comprises:
In step S501, sharpening is carried out to the high-frequency sub-band images in each direction in the high-frequency sub-band images of described multiple directions.
The embodiment of the present invention, before carrying out interpolation to the high-frequency sub-band images in each direction, carries out Edge contrast to each high-frequency sub-band images described.The method of sharpening is preferably extracts corresponding radio-frequency component to each high-frequency sub-band images, then each high-frequency sub-band images is superposed with corresponding radio-frequency component.
In step S502, along corresponding direction, interpolation is carried out to the high-frequency sub-band images in each direction after sharpening, obtains the high-frequency sub-band enlarged image in each direction.
In embodiments of the present invention, interpolation is carried out in spatial domain, according to establishing enlargement factor, the high-frequency sub-band images in each direction after sharpening is carried out interpolation along respective direction.Wherein, described interpolation method includes but not limited to linear interpolation, cubic interpolation etc.By the high-frequency sub-band images in each direction is carried out interpolation along respective direction, the directional information of high-frequency sub-band images can be made more level and smooth.
Preferably, the embodiment of the present invention also directly can carry out interpolation to the high-frequency sub-band images in each direction along respective direction according to presetting enlargement factor, obtain the high-frequency sub-band enlarged image in each direction, to simplify the amplification flow process of ultrasonoscopy, improve the amplification efficiency of ultrasonoscopy further.
In embodiments of the present invention, preferably Newton interpolation is adopted to the interpolation method of raw ultrasound image to be amplified.Newton interpolation can keep the detailed information of raw ultrasound image, makes the raw ultrasound image after interpolation abundanter.The specific implementation flow process of step S104 will be provided below.
As shown in Figure 6, the specific implementation flow process of step S104 comprises:
In step s 601, the multiple interpolation points in raw ultrasound image are obtained.
Need to obtain interpolation point before carrying out Newton interpolation.The embodiment of the present invention generates the integer grid of target high-resolution image in advance, and then obtain the net point P (X, Y) needing to carry out interpolation in described target high-resolution image, described net point is integral point, and namely X, Y are integer.According to amplification factor σ (namely preset enlargement factor), obtain according to following formula and need the net point P (X, Y) carrying out interpolation to be mapped to position p (x, y) on raw ultrasound image:
Wherein, X, Y are horizontal ordinate in target high-resolution image and ordinate; X, y are horizontal ordinate in raw ultrasound image and ordinate.
Described position p (x, y) be mapped on raw ultrasound image can be integral point and floating-point.Integral point is still original position again after interpolation, therefore, the embodiment of the present invention preferably with described floating-point for interpolation point, with improve ultrasonoscopy amplify efficiency.
In step S602, along 0 °, 45 °, 90 ° and 135 ° of four directions, Newton interpolation is carried out to described multiple interpolation point respectively, obtain the interpolation image in each direction.
In step S603, by the weights in each direction of presetting, the interpolation image of four direction is sued for peace, obtain the interpolation image of raw ultrasound image.
In embodiments of the present invention, the interpolation image of 0 ° that obtains in step S402,45 °, 90 ° and 135 ° is respectively f
1, f
2, f
3, f
4, the weights that the interpolation image of 0 °, 45 °, 90 °, 135 ° four direction is corresponding are respectively ω
1, ω
2, ω
3, ω
4, then the interpolation image f of raw ultrasound image is:
f＝ω
1*f
1+ω
2*f
2+ω
3*f
3+ω
4*f
4
Wherein, ω
1+ ω
2+ ω
3+ ω
4=1.Weights ω
1, ω
2, ω
3, ω
4computing method include but not limited to direction First-order Gradient value, second order Grad, gray-scale value method.
Preferably, Fig. 7 shows the specific implementation flow process of carrying out Newton interpolation along a concrete direction that the embodiment of the present invention provides, and comprising:
In step s 701, interpolation point p (x, y) 4 interpolation integral point P are in the direction calculated
1, P
2, P
3, P
4.
Wherein, this direction is the either direction in 0 °, 45 °, 90 ° and 135 °.For 45 ° of directions, described P
1, P
2, P
3, P
4be respectively:
P
1＝[floor(y)+2,floor(x)-1]
P
2＝[floor(y)+1,floor(x)]
P
3＝[floor(y),floor(x)+1]
P
4＝[floor(y)-1,floor(x)+2]
The function of above-mentioned floor function is " rounding downwards ".Floor (x) is not more than the maximum integer of x for getting; Floor (y) is not more than the maximum integer of y for getting.For the ease of understanding, Fig. 8 gives interpolation point p (x, y) and the interpolation integral point P on 45 ° of directions
1, P
2, P
3, P
4position view.
In step S702, calculate P
1, P
2, P
3the Newton interpolation N of three interpolation integral points
1and P
2, P
3, P
4the Newton interpolation N of three interpolation integral points
2.
Exemplarily, the embodiment of the present invention adopts second order Newton interpolation, and the formula of second order Newton interpolation is:
With P
1, P
2, P
3interpolation is example, x
0for a P
2horizontal ordinate, t is interpolation point p (x, y) and P
2between distance, f
0for a P
2pixel value, Δ f for some a P
2first-order Gradient value, Δ
2f
0for a P
2second order Grad.
In step S703, by Newton interpolation N
1with Newton interpolation N
2be weighted, thus obtain the party interpolation image N upwards.
Wherein, the expression formula of weighting procedure is N=ω
1n
1+ ω
2n
2.
Preferably, weights can calculate according to the distance of floating-point to integral point, and integral point distance interpolation point is nearer, and the weight of corresponding Newton interpolation is larger.Certainly, above are only preferably weight calculation method of the present invention, can also be other weight calculation method.
In embodiments of the present invention, interpolation all supports the amplifieroperation of floating-point multiple, and enlargement factor can be the floating numbers such as 1.3,1.6.
In step S105, reconstruction in frequency-domain is carried out to the high-frequency sub-band enlarged image in each obtained direction and the interpolation image of raw ultrasound image.
Contourlet inverse transformation is carried out to the high-frequency sub-band enlarged image in each obtained direction and the interpolation image of raw ultrasound image, thus obtains the output image identical with above-mentioned target high-resolution image size.
Exemplary, will the specific implementation flow process of step S105 be provided below.
As shown in Figure 9, the specific implementation flow process of step S105 comprises:
In step S901, adopt the high-frequency sub-band enlarged image of high frequency direction wave filter to each direction of reconstruction filter banks to carry out filtering, obtain the high-frequency sub-band reconstructed image in each direction.
In step S902, adopt the low frequency anisotropic filter of reconstruction filter banks to carry out filtering to the interpolation image of raw ultrasound image, obtain the reconstructed image of raw ultrasound image.
In step S903, by described each party to high-frequency sub-band reconstructed image and the reconstructed image of raw ultrasound image superpose, obtain superimposed image.
In step S904, inverse Fourier transform is carried out to described superimposed image, obtain the ultrasonoscopy after amplifying.
The image that superimposed image exports after inverse Fourier transform is the ultrasonoscopy after amplifying, and can preserve the ultrasonoscopy after described amplification or print further.Ultrasonoscopy after the amplification obtained by the embodiment of the present invention can keep the detailed information of original image, solves existing ultrasonic image magnifying method and causes the problem that the ultrasonoscopy after amplifying is fuzzy, produce pseudomorphism.
In embodiments of the present invention, after getting raw ultrasound image to be amplified, decompose by the high fdrequency component of multiple decomposition directions to described raw ultrasound image preset, obtain the high-frequency sub-band images of multiple directions; Interpolation is carried out to the high-frequency sub-band images in each direction in the high-frequency sub-band images of described multiple directions, obtains the high-frequency sub-band enlarged image in each direction, and interpolation is carried out to raw ultrasound image, obtain the interpolation image of raw ultrasound image; Finally adopt Contourlet inverse transformation to carry out reconstruction in frequency-domain to the high-frequency sub-band enlarged image in each obtained direction and the interpolation image of raw ultrasound image, thus obtain the ultrasonoscopy after amplifying.Ultrasonoscopy after described amplification can keep the details of original image, and can not produce pseudomorphism, meets clinical medicine requirement, is conducive to the diagnosis of doctor.
embodiment two
Figure 10 shows the composition structure of the ultrasonoscopy multiplying arrangement that the embodiment of the present invention two provides, and for convenience of explanation, illustrate only part related to the present invention.
As shown in Figure 10, described device comprises:
Acquisition module 11, for obtaining raw ultrasound image to be amplified.
Decomposing module 12, for decomposing along multiple directions by the high fdrequency component of multiple exploded view pictures to described raw ultrasound image preset, obtains the high-frequency sub-band images of multiple directions.
Interpolating module 13, interpolation is carried out for the high-frequency sub-band images in each direction in the high-frequency sub-band images to described multiple directions, obtain the high-frequency sub-band enlarged image in each direction, and interpolation is carried out to raw ultrasound image, obtain the interpolation image of raw ultrasound image.
Reconstruction in frequency-domain module 14, for carrying out reconstruction in frequency-domain to the high-frequency sub-band enlarged image in each obtained direction and the interpolation image of raw ultrasound image.
The embodiment of the present invention preferably adopts Contourlet inverse transformation to carry out reconstruction in frequency-domain to the high-frequency sub-band enlarged image in each obtained direction and the interpolation image of raw ultrasound image.
Further, described decomposing module 12 comprises:
Converter unit 121, for carrying out Fourier transform to described raw ultrasound image to be amplified, obtains the frequency spectrum of described raw ultrasound image.
Extraction unit 122, for extracting the radio-frequency component in described frequency spectrum, obtains the high fdrequency component of raw ultrasound image.
Resolving cell 123, for decomposing described high fdrequency component by multiple decomposition directions of presetting, obtains the high-frequency sub-band images of multiple directions.
Further, described interpolating module 13 comprises:
Sharpening unit 131, carries out sharpening for the high-frequency sub-band images in each direction in the high-frequency sub-band images to described multiple directions.
First interpolating unit 132, for carrying out interpolation to the high-frequency sub-band images in each direction after sharpening along corresponding direction, obtains the high-frequency sub-band enlarged image in each direction.
Second interpolating unit 133, for carrying out Newton interpolation to raw ultrasound image, obtains the interpolation image of raw ultrasound image.
In embodiments of the present invention, interpolation is carried out in spatial domain, and the high-frequency sub-band images in each direction after sharpening is carried out interpolation along respective direction.Wherein, described interpolation method includes but not limited to linear interpolation, cubic interpolation etc.By the high-frequency sub-band images in each direction is carried out interpolation along respective direction, the directional information of high-frequency sub-band images can be made more level and smooth.
Preferably, the embodiment of the present invention also directly can carry out interpolation to the high-frequency sub-band images in each direction along respective direction according to default enlargement factor, obtain the high-frequency sub-band enlarged image in each direction, to simplify the amplification flow process of ultrasonoscopy, improve the amplification efficiency of ultrasonoscopy further.
Preferably, in embodiments of the present invention, described second interpolating unit 133 specifically for:
Obtain the multiple interpolation floating-points in raw ultrasound image;
Along 0,45 °, 90 ° and 135 ° of four directions, Newton interpolation is carried out to described multiple interpolation floating-point respectively, obtain the interpolation image in each direction;
By the weights in each direction of presetting, the interpolation image of four direction is sued for peace, obtain the interpolation image of raw ultrasound image.
Further, described reconstruction in frequency-domain module 14 comprises:
First reconstruction in frequency-domain unit 141, for adopting the high-frequency sub-band enlarged image of the high frequency direction wave filter of reconstruction filter banks to each direction to carry out filtering, obtains the high-frequency sub-band reconstructed image in each direction.
Second reconstruction in frequency-domain unit 142, for adopting the low frequency anisotropic filter of reconstruction filter banks to carry out filtering to the interpolation image of raw ultrasound image, obtains the reconstructed image of raw ultrasound image.
Superpositing unit 143, for by described each party to high-frequency sub-band reconstructed image and the reconstructed image of raw ultrasound image superpose, obtain superimposed image.
Inverse transformation block 144, for carrying out inverse Fourier transform to described superimposed image, obtains the ultrasonoscopy after amplifying.
It should be noted that, device in the embodiment of the present invention may be used for the whole technical schemes realized in said method embodiment, the function of its each functional module can the method according to the method described above in embodiment realize, its specific implementation process can refer to the associated description in above-described embodiment, repeats no more herein.
In embodiments of the present invention, after getting raw ultrasound image to be amplified, decompose by the high fdrequency component of multiple decomposition directions to described raw ultrasound image preset, obtain the high-frequency sub-band images of multiple directions; Interpolation is carried out to the high-frequency sub-band images in each direction in the high-frequency sub-band images of described multiple directions, obtains the high-frequency sub-band enlarged image in each direction, and interpolation is carried out to raw ultrasound image, obtain the interpolation image of raw ultrasound image; Finally reconstruction in frequency-domain is carried out to the high-frequency sub-band enlarged image in each obtained direction and the interpolation image of raw ultrasound image, thus obtain the ultrasonoscopy after amplifying.Ultrasonoscopy after described amplification can keep the details of original image, and can not produce pseudomorphism, meets clinical medicine requirement, is conducive to the diagnosis of doctor.
Those of ordinary skill in the art it is also understood that, the all or part of step realized in above-described embodiment method is that the hardware that can carry out instruction relevant by program has come, described program can be stored in a computer read/write memory medium, described storage medium, comprises ROM/RAM, disk, CD etc.
The foregoing is only the preferred embodiments of the present invention, be not limited to the present invention.Such as, modules is carry out dividing according to function logic, but is not limited to above-mentioned division, as long as can realize corresponding function; In addition, the concrete title of each functional module, also just for the ease of mutual differentiation, is not limited to protection scope of the present invention.
All any amendments done within the spirit and principles in the present invention, equivalent replacement and improvement etc., all should be included within protection scope of the present invention.
Claims (10)
1. an amplification method for ultrasonoscopy, is characterized in that, described method comprises:
Obtain raw ultrasound image to be amplified;
Decompose by the high fdrequency component of multiple decomposition directions to described raw ultrasound image preset, obtain the high-frequency sub-band images of multiple directions;
Interpolation is carried out to the high-frequency sub-band images in each direction in the high-frequency sub-band images of described multiple directions, obtains the high-frequency sub-band enlarged image in each direction;
Interpolation is carried out to raw ultrasound image, obtains the interpolation image of raw ultrasound image;
Reconstruction in frequency-domain is carried out to the high-frequency sub-band enlarged image in each obtained direction and the interpolation image of raw ultrasound image.
2. the amplification method of ultrasonoscopy as claimed in claim 1, is characterized in that, the described high fdrequency component of multiple decomposition directions to described raw ultrasound image by presetting is decomposed, and the high-frequency sub-band images obtaining multiple directions comprises:
Fourier transform is carried out to described raw ultrasound image to be amplified, obtains the frequency spectrum of described raw ultrasound image;
Extract the radio-frequency component in described frequency spectrum, obtain the high fdrequency component of raw ultrasound image;
By multiple decomposition directions of presetting, described high fdrequency component is decomposed, obtain the high-frequency sub-band images of multiple directions.
3. the amplification method of ultrasonoscopy as claimed in claim 1, is characterized in that, describedly carries out interpolation to the high-frequency sub-band images in each direction in the high-frequency sub-band images of described multiple directions, and the high-frequency sub-band enlarged image obtaining each direction comprises:
Sharpening is carried out to the high-frequency sub-band images in each direction in the high-frequency sub-band images of described multiple directions;
Along corresponding direction, interpolation is carried out to the high-frequency sub-band images in each direction after sharpening, obtains the high-frequency sub-band enlarged image in each direction;
Describedly carry out interpolation to raw ultrasound image, the interpolation image obtaining raw ultrasound image is specially:
Newton interpolation is carried out to raw ultrasound image, obtains the interpolation image of raw ultrasound image.
4. the amplification method of ultrasonoscopy as claimed in claim 1, is characterized in that, describedly carries out reconstruction in frequency-domain to the high-frequency sub-band enlarged image in each obtained direction and the interpolation image of raw ultrasound image and comprises:
Adopt the high-frequency sub-band enlarged image of high frequency direction wave filter to each direction of reconstruction filter banks to carry out filtering, obtain the high-frequency sub-band reconstructed image in each direction;
Adopt the low frequency anisotropic filter of reconstruction filter banks to carry out filtering to the interpolation image of raw ultrasound image, obtain the reconstructed image of raw ultrasound image;
By described each party to high-frequency sub-band reconstructed image and the reconstructed image of raw ultrasound image superpose, obtain superimposed image;
Inverse Fourier transform is carried out to described superimposed image, obtains the ultrasonoscopy after amplifying.
5. as right will remove the amplification method of the ultrasonoscopy as described in 3, it is characterized in that, describedly carry out Newton interpolation to raw ultrasound image, the interpolation image obtaining raw ultrasound image comprises:
Obtain the multiple interpolation points in raw ultrasound image;
Along 0,45 °, 90 ° and 135 ° of four directions, Newton interpolation is carried out to described multiple interpolation point respectively, obtain the interpolation image in each direction;
By the weights in each direction of presetting, the interpolation image of four direction is sued for peace, obtain the interpolation image of raw ultrasound image.
6. a multiplying arrangement for ultrasonoscopy, is characterized in that, described device comprises:
Acquisition module, for obtaining raw ultrasound image to be amplified;
Decomposing module, for decomposing by the high fdrequency component of multiple decomposition directions to described raw ultrasound image preset, obtains the high-frequency sub-band images of multiple directions;
Interpolating module, interpolation is carried out for the high-frequency sub-band images in each direction in the high-frequency sub-band images to described multiple directions, obtain the high-frequency sub-band enlarged image in each direction, and interpolation is carried out to raw ultrasound image, obtain the interpolation image of raw ultrasound image;
Reconstruction in frequency-domain module, for carrying out reconstruction in frequency-domain to the high-frequency sub-band enlarged image in each obtained direction and the interpolation image of raw ultrasound image.
7. the multiplying arrangement of ultrasonoscopy as claimed in claim 6, it is characterized in that, described decomposing module comprises:
Converter unit, for carrying out Fourier transform to described raw ultrasound image to be amplified, obtains the frequency spectrum of described raw ultrasound image;
Extraction unit, for extracting the radio-frequency component in described frequency spectrum, obtains the high fdrequency component of raw ultrasound image;
Resolving cell, for decomposing described high fdrequency component by multiple decomposition directions of presetting, obtains the high-frequency sub-band images of multiple directions.
8. the multiplying arrangement of ultrasonoscopy as claimed in claim 6, it is characterized in that, described interpolating module comprises:
Sharpening unit, carries out sharpening for the high-frequency sub-band images in each direction in the high-frequency sub-band images to described multiple directions;
First interpolating unit, for carrying out interpolation to the high-frequency sub-band images in each direction after sharpening along corresponding direction, obtains the high-frequency sub-band enlarged image in each direction;
Second interpolating unit, for carrying out Newton interpolation to raw ultrasound image, obtains the interpolation image of raw ultrasound image.
9. the multiplying arrangement of ultrasonoscopy as claimed in claim 6, it is characterized in that, described reconstruction in frequency-domain module comprises:
First reconstruction in frequency-domain unit, for adopting the high-frequency sub-band enlarged image of the high frequency direction wave filter of reconstruction filter banks to each direction to carry out filtering, obtains the high-frequency sub-band reconstructed image in each direction;
Second reconstruction in frequency-domain unit, for adopting the low frequency anisotropic filter of reconstruction filter banks to carry out filtering to the interpolation image of raw ultrasound image, obtains the reconstructed image of raw ultrasound image;
Superpositing unit, for the high-frequency sub-band reconstructed image in each direction described and the reconstructed image of raw ultrasound image being superposed, obtains superimposed image;
Inverse transformation block, for carrying out inverse Fourier transform to described superimposed image, obtains the ultrasonoscopy after amplifying.
10., as right will remove the multiplying arrangement of the ultrasonoscopy as described in 8, it is characterized in that, described second interpolating unit specifically for:
Obtain the multiple interpolation points in raw ultrasound image;
Along 0,45 °, 90 ° and 135 ° of four directions, Newton interpolation is carried out to described multiple interpolation point respectively, obtain the interpolation image in each direction;
By the weights in each direction of presetting, the interpolation image of four direction is sued for peace, obtain the interpolation image of raw ultrasound image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410668306.4A CN104463785B (en) | 2014-11-20 | 2014-11-20 | A kind of amplification method and device of ultrasonoscopy |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410668306.4A CN104463785B (en) | 2014-11-20 | 2014-11-20 | A kind of amplification method and device of ultrasonoscopy |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104463785A true CN104463785A (en) | 2015-03-25 |
| CN104463785B CN104463785B (en) | 2018-08-21 |
Family
ID=52909773
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410668306.4A Active CN104463785B (en) | 2014-11-20 | 2014-11-20 | A kind of amplification method and device of ultrasonoscopy |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104463785B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106331429A (en) * | 2016-08-31 | 2017-01-11 | 上海交通大学 | Video detail magnifying method |
| CN106548454A (en) * | 2016-09-08 | 2017-03-29 | 清华大学 | The method and apparatus for processing medical image |
| CN106600532A (en) * | 2016-12-08 | 2017-04-26 | 广东威创视讯科技股份有限公司 | Method and apparatus for amplifying image |
| CN106780341A (en) * | 2016-12-28 | 2017-05-31 | 辽宁师范大学 | The five rank self- steering super resolution ratio reconstruction methods based on profile ripple |
| CN107845082A (en) * | 2017-12-14 | 2018-03-27 | 飞依诺科技(苏州)有限公司 | A kind of method and apparatus for being used to improve the contrast of regional area |
| CN110115598A (en) * | 2019-06-19 | 2019-08-13 | 青岛市妇女儿童医院（青岛市妇幼保健院、青岛市残疾儿童医疗康复中心、青岛市新生儿疾病筛查中心） | It is a kind of can automatic daubing coupling agent ultrasonic probe control system and method |
| CN110310227A (en) * | 2019-06-27 | 2019-10-08 | 电子科技大学 | A kind of image super-resolution rebuilding method decomposed based on high and low frequency information |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002174564A (en) * | 2000-12-06 | 2002-06-21 | Hitachi Ltd | Determination and evaluation method of quality of display device, and the display device |
| CN101303763A (en) * | 2007-12-26 | 2008-11-12 | 公安部上海消防研究所 | Method for amplifying image based on rarefaction representation |
| KR20100097858A (en) * | 2009-02-27 | 2010-09-06 | 홍익대학교 산학협력단 | Super-resolution using example-based neural networks |
| CN103500436A (en) * | 2013-09-17 | 2014-01-08 | 广东威创视讯科技股份有限公司 | Image super-resolution processing method and system |
-
2014
- 2014-11-20 CN CN201410668306.4A patent/CN104463785B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002174564A (en) * | 2000-12-06 | 2002-06-21 | Hitachi Ltd | Determination and evaluation method of quality of display device, and the display device |
| CN101303763A (en) * | 2007-12-26 | 2008-11-12 | 公安部上海消防研究所 | Method for amplifying image based on rarefaction representation |
| KR20100097858A (en) * | 2009-02-27 | 2010-09-06 | 홍익대학교 산학협력단 | Super-resolution using example-based neural networks |
| CN103500436A (en) * | 2013-09-17 | 2014-01-08 | 广东威创视讯科技股份有限公司 | Image super-resolution processing method and system |
Non-Patent Citations (2)
| Title |
|---|
| 田萌，修春波，卢少磊: "改进的单帧图像超分辨率算法研究", 《计算机工程与应用》 * |
| 薛正亮: "图像超分辨率重建技术研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106331429A (en) * | 2016-08-31 | 2017-01-11 | 上海交通大学 | Video detail magnifying method |
| CN106548454A (en) * | 2016-09-08 | 2017-03-29 | 清华大学 | The method and apparatus for processing medical image |
| CN106600532A (en) * | 2016-12-08 | 2017-04-26 | 广东威创视讯科技股份有限公司 | Method and apparatus for amplifying image |
| CN106600532B (en) * | 2016-12-08 | 2020-01-10 | 广东威创视讯科技股份有限公司 | Image amplification method and device |
| CN106780341A (en) * | 2016-12-28 | 2017-05-31 | 辽宁师范大学 | The five rank self- steering super resolution ratio reconstruction methods based on profile ripple |
| CN106780341B (en) * | 2016-12-28 | 2020-12-01 | 辽宁师范大学 | Five-order self-steering super-resolution reconstruction method based on contour wave |
| CN107845082A (en) * | 2017-12-14 | 2018-03-27 | 飞依诺科技(苏州)有限公司 | A kind of method and apparatus for being used to improve the contrast of regional area |
| CN107845082B (en) * | 2017-12-14 | 2021-10-15 | 飞依诺科技(苏州)有限公司 | Method and device for improving contrast of local area |
| CN110115598A (en) * | 2019-06-19 | 2019-08-13 | 青岛市妇女儿童医院（青岛市妇幼保健院、青岛市残疾儿童医疗康复中心、青岛市新生儿疾病筛查中心） | It is a kind of can automatic daubing coupling agent ultrasonic probe control system and method |
| CN110310227A (en) * | 2019-06-27 | 2019-10-08 | 电子科技大学 | A kind of image super-resolution rebuilding method decomposed based on high and low frequency information |
| CN110310227B (en) * | 2019-06-27 | 2020-09-08 | 电子科技大学 | Image super-resolution reconstruction method based on high-low frequency information decomposition |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104463785B (en) | 2018-08-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104463785A (en) | Method and device for amplifying ultrasound image | |
| CN102722865B (en) | Super-resolution sparse representation method | |
| CN101950365B (en) | Multi-task super-resolution image reconstruction method based on KSVD dictionary learning | |
| Hamilton et al. | Beltrami-net: domain-independent deep D-bar learning for absolute imaging with electrical impedance tomography (a-EIT) | |
| CN101504766B (en) | Image amalgamation method based on mixed multi-resolution decomposition | |
| CN106204449A (en) | A kind of single image super resolution ratio reconstruction method based on symmetrical degree of depth network | |
| CN114429422A (en) | Image super-resolution reconstruction method and system based on residual channel attention network | |
| CN106485764B (en) | The quick exact reconstruction methods of MRI image | |
| CN104166974B (en) | CT locating film image enhancing method and CT locating film image enhancing device | |
| CN106204447A (en) | The super resolution ratio reconstruction method with convolutional neural networks is divided based on total variance | |
| Zhou et al. | Adaptive tight frame based medical image reconstruction: a proof-of-concept study for computed tomography | |
| CN105046651B (en) | A kind of ultra-resolution ratio reconstructing method and device of image | |
| CN101201937A (en) | Digital image enhancement method and device based on wavelet restruction and decompose | |
| CN105046672A (en) | Method for image super-resolution reconstruction | |
| CN102881000A (en) | Super-resolution method, device and equipment for video image | |
| Claus et al. | Metal-artifact reduction using deep-learning based sinogram completion: initial results | |
| He et al. | Remote sensing image super-resolution using deep–shallow cascaded convolutional neural networks | |
| CN102184532A (en) | Single scale based medical image edge detection | |
| CN106651772A (en) | Super-resolution reconstruction method of satellite cloud picture | |
| CN104299193B (en) | Image super-resolution reconstruction method based on high-frequency information and medium-frequency information | |
| Pfister et al. | Learning sparsifying filter banks | |
| Chen et al. | A sparse representation and dictionary learning based algorithm for image restoration in the presence of Rician noise | |
| Liu et al. | An edge enhanced srgan for mri super resolution in slice-selection direction | |
| CN103390266B (en) | A kind of image super-resolution method and device | |
| CN104766290A (en) | Pixel information estimation fusion method based on quick NSCT |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |