CN110415179A - A kind of time complex method and device based on ultrasonic sequence similarity - Google Patents
A kind of time complex method and device based on ultrasonic sequence similarity Download PDFInfo
- Publication number
- CN110415179A CN110415179A CN201910490844.1A CN201910490844A CN110415179A CN 110415179 A CN110415179 A CN 110415179A CN 201910490844 A CN201910490844 A CN 201910490844A CN 110415179 A CN110415179 A CN 110415179A
- Authority
- CN
- China
- Prior art keywords
- signal
- class
- neighborhood
- time
- compound
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims description 33
- 150000001875 compounds Chemical class 0.000 claims abstract description 53
- 230000006798 recombination Effects 0.000 claims abstract description 37
- 238000005215 recombination Methods 0.000 claims abstract description 37
- 230000015572 biosynthetic process Effects 0.000 claims abstract description 36
- 238000004364 calculation method Methods 0.000 claims abstract description 20
- 239000002131 composite material Substances 0.000 claims abstract description 16
- 238000001914 filtration Methods 0.000 claims abstract description 11
- 230000011218 segmentation Effects 0.000 claims abstract description 10
- 230000002708 enhancing effect Effects 0.000 claims abstract description 9
- 239000000523 sample Substances 0.000 description 12
- 238000005516 engineering process Methods 0.000 description 11
- 238000003384 imaging method Methods 0.000 description 8
- 230000008901 benefit Effects 0.000 description 5
- 206010019133 Hangover Diseases 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000003759 clinical diagnosis Methods 0.000 description 3
- 238000013329 compounding Methods 0.000 description 3
- 238000003786 synthesis reaction Methods 0.000 description 3
- 238000002604 ultrasonography Methods 0.000 description 3
- 230000003044 adaptive effect Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 230000006866 deterioration Effects 0.000 description 2
- 210000004185 liver Anatomy 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 230000008520 organization Effects 0.000 description 2
- 230000004044 response Effects 0.000 description 2
- 238000012952 Resampling Methods 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 239000011248 coating agent Substances 0.000 description 1
- 238000000576 coating method Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000737 periodic effect Effects 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000002123 temporal effect Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- 230000003313 weakening effect Effects 0.000 description 1
Classifications
-
- G06T5/73—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/13—Edge detection
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H30/00—ICT specially adapted for the handling or processing of medical images
- G16H30/40—ICT specially adapted for the handling or processing of medical images for processing medical images, e.g. editing
-
- 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
- G06T2207/10132—Ultrasound 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
- G06T2207/20024—Filtering details
-
- 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
- G06T2207/20172—Image enhancement details
- G06T2207/20192—Edge enhancement; Edge preservation
Abstract
The invention discloses a kind of time set composites based on ultrasonic sequence similarity, comprising: Signal Pretreatment unit, for 2D signal to be carried out low-pass filtering and edge enhancing processing；Signal cutting unit, for 2D signal to be divided into the identical rectangular neighborhood of several sizes by treated；Taxon, for several 2D signal neighborhoods after the segmentation to be divided into edge tissues class, uniform formation's class, noise class；First computing unit, for calculate present frame 2D signal neighborhood and former frame corresponding to similarity between 2D signal neighborhood；Second computing unit, the similarity calculation time recombination coefficient for being obtained according to；Time recombiner unit, it is compound for the 2D signal neighborhood of present frame 2D signal neighborhood corresponding with previous or multiframe to be passed through the obtained time recombination coefficient progress time.
Description
Technical field
The present invention relates to ultrasonic imaging technique field more particularly to a kind of time composite squares based on ultrasonic sequence similarity
Method and device.
Background technique
Medical ultrasonic image is often since the electronic noise of its hardware system and the relevant caused speckle of tissue echo are made an uproar
Sound and make picture quality (such as resolution ratio and signal-to-noise ratio) reduce, cause clinical diagnosis to become difficult.Inhibit electronic noise and dissipates
The algorithm of spot noise can be divided into adaptive-filtering and Compound scan technology.Compound scan technology is the company that will have certain correlation
Continuous or discontinuous frame image is weighted superposition, is minimized noise, improves picture quality, including spatial compounding, frequency multiplexed
It is compound with the time.Time is compound to be realized by frame the relevant technologies, is to choose sometime putting in image temporal periodic sequence
Successive frame, then individual element application FIR (Finite Impulse Response) or IIR (Infinite Impulse
Response) digital filter, to reduce electronic noise and speckle noise.Assuming that image is absolute rest, due to not
Electronic noise between at same frame is uncorrelated, this technology can effectively reduce electronic noise and keep the acoustic picture of itself；
Speckle noise correlation can be made also weaker when tissue and ultrasonic probe light exercise, this technology can reduce speckle noise
And keep institutional framework.So signal noise ratio (snr) of image can be improved in the weighted superposition between the stronger frame of institutional framework correlation.
Time complex technique computation complexity is smaller, has been widely used for real-time ultrasonography system.
The basic model of current existing time compound frame the relevant technologies is formulated are as follows:
I′k=α Ik-1′+(1-α)Ik；
Wherein, α is frame correlation coefficient, 0≤α≤1, I 'kAnd IkRespectively indicate the ultrasound figure before and after kth frame frame relevant treatment
As pixel value.When kth frame and -1 frame image of kth are weighted by coefficient, need for each pixel node-by-node algorithm, calculation amount
Greatly.
Current existing frame the relevant technologies are limited in that significantly histokinesis or hand hold transducer movement are too fast
Caused by institutional framework it is smudgy, by the way that streaking phenomenon can be shown after time compound frame relevant treatment, scheme
Image distortion, and user generally selects the relevant grade of lower frame in order to avoid this distortion, cannot also make full use of frame related
For improving the advantage of signal noise ratio (snr) of image, to reduce value of the frame the relevant technologies in clinical diagnosis.
As the patent of Publication No. CN103584882A discloses a kind of spatial compound imaging method, device and ultrasonic imaging
System, wherein the method includes the steps: recombination coefficient Coefx is generated according to preset parameter；Weight is carried out to current deflection image
Sampling, the image Ax (t) current deflection image being converted under the coordinate system of 0 angle；Read the knot of last time spatial compound imaging
Fruit image C (t-1)；Using recombination coefficient Coefx respectively to the result images C (t-1) of last time spatial compound imaging and to current
Deflection image carries out the weighted calculation that the coordinate space image Ax (t) for 0 angle that resampling obtains is multiplied respectively, is added
Combination picture Coefx*C (t-1) and weighted angular image (1-Coefx) * Ax (t) is weighed, and to weighting combination picture and weighting angle
Degree image is summed to obtain the result images C (t) of this spatial compound imaging.The present invention is realized polygonal using frame correlation technique
The spatial compounding of degree combines spatial compounding and time compound phase, effectively inhibits boundary caused by different angle image, reduces
Image deterioration caused by angular deflection.Although it can solve the deterioration for different angle image, it is still deposited
Frame correlation cannot made full use of for the advantage of raising signal noise ratio (snr) of image, to reduce frame the relevant technologies in clinical diagnosis
Value.
Summary of the invention
It is multiple the purpose of the present invention is in view of the drawbacks of the prior art, providing a kind of time based on ultrasonic sequence similarity
Method and device is closed, the journey compound for the adaptive ultrasonic sequence of adjusting in the case where tissue or different probe movement speed
Degree；It is lower in the region composite degree for the strenuous exercise that organizes or pop one's head in, institutional framework obscurity boundary can be weakened and hangover is existing
As；Slow region is being moved, higher Compound Degree can be used, electronic noise and speckle noise can be effectively reduced.
In order to achieve the goal above, the invention adopts the following technical scheme:
A kind of time set composite based on ultrasonic sequence similarity, comprising:
Signal Pretreatment unit, for 2D signal to be carried out low-pass filtering and edge enhancing processing；
Signal cutting unit, for 2D signal to be divided into the identical rectangular neighborhood of several sizes by treated；
Taxon, for several 2D signal neighborhoods after the segmentation to be divided into edge tissues class, uniform formation
Class, noise class；
First computing unit, the 2D signal neighborhood for calculating present frame and 2D signal neighborhood corresponding to former frame
Between similarity；
Second computing unit, the similarity calculation time recombination coefficient for being obtained according to；
Time recombiner unit, for by the 2D signal neighborhood of present frame 2D signal neighborhood corresponding with previous or multiframe
It is compound that the time is carried out by obtained time recombination coefficient.
Further, several 2D signal neighborhoods are divided into edge tissues class in the taxon, uniform formation's class, made an uproar
Sound class is classified by preset tissue threshold value and Grads threshold；If being lower than the tissue threshold value, for noise class；If being higher than described
Threshold value is organized, then is edge tissues class and uniform formation's class；If being lower than the Grads threshold, for uniform formation's class；If being higher than
The Grads threshold is then edge tissues class.
Further, the taxon further includes that edge tissues class, uniform formation's class, noise class is respectively set is corresponding
Time recombination coefficient.
Further, further includes:
Storage unit, for 2D signal neighborhood of the time after compound to be stored in buffer zone.
Further, the 2D signal neighborhood of the present frame is two corresponding to the former frame or multiframe with buffer zone
It is compound that dimensional signal neighborhood carries out the time；The complex method be the 2D signal neighborhood of present frame and the former frame of buffer zone or
The corresponding 2D signal neighborhood of multiframe is compound by weighted superposition.
Correspondingly, a kind of time complex method based on ultrasonic sequence similarity is also provided, comprising steps of
S1. 2D signal is subjected to low-pass filtering and edge enhancing is handled；
S2. by treated, 2D signal is divided into the identical rectangular neighborhood of several sizes；
S3. several 2D signal neighborhoods after the segmentation are divided into edge tissues class, uniform formation's class, noise class；
S4. calculate present frame 2D signal neighborhood and former frame corresponding to similarity between 2D signal neighborhood；
S5. the similarity calculation time recombination coefficient obtained according to；
S6., the 2D signal neighborhood of present frame 2D signal neighborhood corresponding with previous or multiframe is passed through to the obtained time
It is compound that recombination coefficient carries out the time.
Further, it is described by several 2D signal neighborhoods be divided into edge tissues class, uniform formation's class, noise class pass through it is pre-
If tissue threshold value and Grads threshold classification；If being lower than the tissue threshold value, for noise class；If being higher than the tissue threshold value,
It is then edge tissues class and uniform formation's class；If being lower than the Grads threshold, for uniform formation's class；If being higher than the gradient threshold
Value, then be edge tissues class.
Further, the step S3 further include:
The corresponding time recombination coefficient of edge tissues class, uniform formation's class, noise class is respectively set.
Further, which is characterized in that further comprised the steps of: after the step S6
S7. the 2D signal neighborhood by the time after compound is stored in buffer zone.
Further, the 2D signal neighborhood of the present frame is two corresponding to the former frame or multiframe with buffer zone
It is compound that dimensional signal neighborhood carries out the time；The complex method be present frame 2D signal neighborhood and buffer zone it is preceding one or more
The corresponding 2D signal neighborhood of frame is compound by weighted superposition.
Compared with prior art, the present invention beneficial effect is:
1. the present invention does time compound processing on 2D signal, the number of scan points that this mode is handled is compared to scheming
It is small that calculation amount is processed on picture.
2. the present invention is in the recombination coefficient for setting each neighborhood based on the severe degree of tissue and probe movement, than existing
There is technology accurate.
3. the present invention can effectively inhibit the electronic noise of system, smooth speckle noise, while being able to maintain image side
Edge feature, avoid due to tissue and probe movement it is too fast caused by combination picture edge blurry even trail.
Detailed description of the invention
Fig. 1 is a kind of time set composite structure chart based on ultrasonic sequence similarity that embodiment one provides；
Fig. 2 is a kind of time complex method flow chart based on ultrasonic sequence similarity that embodiment two provides；
Fig. 3 is a kind of structure chart for ultrasonic system imaging device that embodiment one provides；
Fig. 4 is the time compound frame relative theory figure that embodiment one provides；
Fig. 5 is the neighborhood classification method figure that embodiment one provides.
Specific embodiment
Illustrate embodiments of the present invention below by way of specific specific example, those skilled in the art can be by this specification
Other advantages and efficacy of the present invention can be easily understood for disclosed content.The present invention can also pass through in addition different specific realities
The mode of applying is embodied or practiced, the various details in this specification can also based on different viewpoints and application, without departing from
Various modifications or alterations are carried out under spirit of the invention.It should be noted that in the absence of conflict, following embodiment and implementation
Feature in example can be combined with each other.
It is multiple the purpose of the present invention is in view of the drawbacks of the prior art, providing a kind of time based on ultrasonic sequence similarity
Method and device is closed, the journey compound for the adaptive ultrasonic sequence of adjusting in the case where tissue or different probe movement speed
Degree；It is lower in the region composite degree for the strenuous exercise that organizes or pop one's head in, institutional framework obscurity boundary can be weakened and hangover is existing
As；Slow region is being moved, higher Compound Degree can be used, electronic noise and speckle noise can be effectively reduced.
Embodiment one
It should be noted that a kind of executing subject of time set composite based on ultrasonic sequence similarity of the present embodiment
For ultrasonic system.
A kind of time set composite based on ultrasonic sequence similarity provided in this embodiment, as shown in Figure 1, comprising:
Signal Pretreatment unit 11, for 2D signal to be carried out low-pass filtering and edge enhancing processing；
Signal cutting unit 12, for 2D signal to be divided into the identical rectangular neighborhood of several sizes by treated；
Taxon 13, for several 2D signal neighborhoods after the segmentation to be divided into edge tissues class, uniform group
Knit class, noise class；
First computing unit 14, the 2D signal neighborhood and 2D signal corresponding to former frame for calculating present frame are adjacent
Similarity between domain；
Second computing unit 15, the similarity calculation time recombination coefficient for being obtained according to；
Time recombiner unit 16, for the 2D signal neighborhood of present frame 2D signal corresponding with previous or multiframe is adjacent
It is compound that domain passes through the obtained time recombination coefficient progress time.
It is a kind of ultrasonic system imaging device of the present embodiment as shown in Figure 3, including transmitting and receiving module, Beam synthesis
Module, 2D signal processing module, time set composite module, scan transformation and image processing module, display module.
Specifically, entering human body by probe transmitting ultrasonic wave, after the Tissue reflectance of human body, it is received module and connects
It receives, the echo received enters Beam synthesis module and obtains scan-line data, and scan-line data passes through 2D signal processing module
Time compound frame relevant apparatus is re-fed into be handled, by time compound scan-line data enter back into digital scan conversion,
Image procossing and image display.
Wherein, transmitting and receiving module, Beam synthesis module, 2D signal processing module, digital scan conversion, at image
Technology well known to those skilled in the art can be used in reason and image display, does not elaborate herein.
By the scan-line data that 2D signal processing module is handled, two-dimentional scan line sequence IkWith from system cache
The former frame scan line sequence I of acquirementk-1' progress is compound, generates I laterkIn ' deposit system cache.Wherein, k=0,1,
2 ... n, sequence sequential storage in the buffer, as shown in Figure 4.
Time set composite module uses processing mode provided in this embodiment, and specific processing mode such as following the description is retouched
It states.
Signal Pretreatment unit 11, for 2D signal to be carried out low-pass filtering and edge enhancing processing.
Specifically, the low-pass filtering method of 2D signal can be realized by two-dimensional Weighted Mean Filter,
Wherein, w is two dimensional filter, and f is input 2D signal, and g is output 2D signal.
Exporting 2D signal enhances marginal information by two-dimensional space sharpening filter, and filter can use second-order differential
Laplace operator or first differential gradient operator.
In the present embodiment, the noise level that signal is reduced using two-dimentional smoothing filter, using two-dimensional high-pass filter
Sharpen boundary characteristic
In the present embodiment, low-pass filtering treatment is to reduce random noise and speckle noise for neighborhood classification, meter
Calculate the influence of similarity and estimation accuracy；Enhancing processing in edge is the edge contour for highlighting tissue.Such as: liver coating
Place is since high frequency characteristics generally occurs in the violent signal of gray-value variation, and liver interior tissue is smoother, and gray value compares
Uniform, low frequency feature is then presented in signal.Therefore, the high and low frequency feature of signal, purpose can be enhanced respectively as needed
It is the accuracy for improving neighborhood classification, similarity calculation and estimation.
The present embodiment is reduced signal noise level, is enhanced image border, make neighbour by the pretreatment to 2D signal
Domain classification and similarity calculation are more acurrate.
It should be noted that 2D signal requires to carry out 2D signal before carrying out the correlation steps such as the time is compound
Pretreatment.
Signal cutting unit 12, for 2D signal to be divided into the identical rectangular neighborhood of multiple sizes by treated.
In the present embodiment, the size of neighborhood changes with scanning line density and number of scan points purpose and is changed.
Specifically, signal piecemeal, is divided into the identical rectangular neighborhood of several sizes for current frame signal.Assuming that current two dimension letter
Number size is w × h, and wherein w indicates the number of scanning lines of signal transverse direction, the number of scan points of h expression signal longitudinal direction；Signal is divided into m
Then remember that the size of rectangular neighborhood is p × q, calculation in the region of × n are as follows:
P=w/m, q=h/n；
The signal magnitude of present frame changes with the variation of scanning line density and number of scan points, so the size of neighborhood
Have corresponding variation.
The segmentation number of neighborhood can be set according to system current application and frame per second.If current organization structure is simple,
The segmentation number of neighborhood can be fewer；If structure is complicated for current organization, it is careful that neighborhood need to be divided.If current frame rate
Low, then neighborhood number can be less；If current frame rate is high, it is careful that neighborhood need to be divided.
Taxon 13, for several 2D signal neighborhoods after the segmentation to be divided into edge tissues class, uniform group
Knit class, noise class.
Wherein, edge tissues class, the classification that uniform formation's class is tissue class.
As shown in figure 5, each signal neighborhood is divided into edge tissues class, uniform formation's class, noise class, wherein will be each
It is by preset tissue threshold value and Grads threshold that a 2D signal neighborhood, which is divided into edge tissues class, uniform formation's class, noise class,
Classification.
Distinguish whether neighborhood is noise class according to signal strength average value in neighborhood, since noise signal strength is much smaller than group
The signal strength of class is knitted, so the neighborhood lower than tissue threshold value is divided into noise class, the neighborhood higher than tissue threshold value is divided into tissue class.
If tissue class, then according to the two-dimensional gradient value or standard deviation of signal distinguish neighborhood belong to uniform formation's class or
Edge tissues class.Since the gradient value or standard deviation of edge tissues are much larger than uniform formation, so being lower than the neighbour of Grads threshold
Domain is uniform formation's class, and the neighborhood higher than Grads threshold is edge tissues class.
Time recombination coefficient is set separately according to classification in the neighborhood classified, wherein edge tissues take weaker answer
Collaboration number, uniform formation take stronger recombination coefficient, and the recombination coefficient of noise is maximum.Its purpose is to maximum
Picture edge characteristic can be effectively retained while reducing electronic noise and speckle noise, weakening tissue and probe movement makes the time
Image after compound generates hangover and blooming.
In the present embodiment, time recombination coefficient is not only related with present frame, the signal similarity of former frame, also with signal
Classification it is related.For ultrasound image, edge tissues class is easier to cause streaking or fuzzy, uniform formation and makes an uproar
Sound is not easy to cause to obscure.So when calculating time correlation coefficient, consideration is associated with signal content, according to the classification of neighborhood
Set the corrected value of coefficient.
Wherein, according to the corrected value ρ of the category setting time recombination coefficient of neighborhood, then frame correlation formula are as follows:
I′k=ρ α Ik-1′+(1-ρα)Ik。
First computing unit 14, the 2D signal neighborhood and 2D signal corresponding to former frame for calculating present frame are adjacent
Similarity between domain.
In the present embodiment, it is calculated corresponding to 2D signal neighborhood and the former frame of present frame using Histogram Matching algorithm
2D signal neighborhood between similarity.
Histogram Matching is a kind of method measured based on vector similarity to the similarity of signal or image.
Its principle is the histogram for calculating separately two neighborhood signals, and histogram is normalized, then according to distance metric
The measurement of standard progress similarity.
Histogram data acquisition is carried out to neighborhood signal corresponding to the neighborhood signal of present frame and former frame first, to adopting
The respective image histogram of collection is normalized, and reuses Pasteur's Coefficient Algorithm and calculates histogram data, finally obtains
Image similarity value.The calculation formula of Pasteur's coefficient factor value λ is as follows:
Wherein p represents the histogram data of current neighborhood signal, and p ' represents the histogram of neighborhood signal corresponding to former frame
Data, being added the result obtained later to the data dot-product extraction of square root of each identical i is image similarity value (Pasteur's coefficient
Factor values), range is between 0 to 1.0 indicates extremely different；1 indicates extremely similar (identical).
The present embodiment is that histogram can be good at normalizing using the advantage of the similarity calculating method of Histogram Matching
Change, the neighborhood signal of different resolution can directly calculate similarity by its histogram, and algorithm complexity is moderate, calculate
It measures moderate.
Certainly, the present embodiment can also be used perceptual hash algorithm (PHA) calculate present frame 2D signal neighborhood with it is previous
Similarity between 2D signal neighborhood corresponding to frame.
Perceptual hash algorithm is the comparison algorithm of a kind of signal or image, and principle is that the feature that is included is raw with signal
At one group of fingerprint, the similarity of signal is calculated by comparing the difference of fingerprint.
Neighborhood signal is narrowed down to the size of 8x8 first, in total 64 scanning elements.Played the role of being the various rulers of removal
Very little and ratio difference, only retains the essential informations such as structure, light and shade.
Secondly signal is switched into 64 grades of grey scale signals.Each signal value is compared with average value, if it is greater than or wait
It is denoted as 1 in average value, is denoted as 0 less than average value.
Then cryptographic Hash is calculated, the comparison result that above-mentioned each signal value is compared with average value is combined,
The binary integer for constituting one 64, obtains the fingerprint of picture.
How many position the fingerprint of last contrast signal, calculate in 64 as.Assuming that the same digit is n, then it is similar
Angle value are as follows:
λ=n/64.
The present embodiment is no matter change the height and width either brightness of neighborhood signal using the advantage of perceptual hash algorithm,
Cryptographic Hash will not all be changed, that is to say, that will not influence the similarity value of signal.Meanwhile the complexity of perceptual hash algorithm is very
Low, arithmetic speed is fast, absolutely not influences on the real-time of ultrasonic signal and image.
It should be noted that between 2D signal neighborhood corresponding to the 2D signal neighborhood and former frame of calculating present frame
Similarity be not limited only to the present embodiment proposition Histogram Matching algorithm and perceptual hash algorithm, it is common that industry can also be used
Method, such as matrix decomposition method, the method based on characteristic point, the method for being based on Y-PSNR (PSNR) are similar with structure is based on
The method etc. of property (SSIM, structural similarity (SSIM) index measurement).
Second computing unit 15, the similarity calculation time recombination coefficient for being obtained according to.
Go out time recombination coefficient by similarity calculation, time recombination coefficient and similarity are directly proportional, can pass through nature
Exponential function is realized.
In the present embodiment, time recombination coefficient is related with neighborhood classification, and edge tissues class recombination coefficient is smaller, and uniform group
It is larger to knit class recombination coefficient, noise class recombination coefficient is maximum.
Calculate frame relevant parameter according to the similarity value of neighborhood: the frame correlation coefficient of ultrasonic system usually utilizes natural Exponents
It calculates, calculation formula are as follows:
Wherein, k is frame correlation-corrected parameter, and Fr is image display frame rate.Image display frame rate is smaller, then frame phase relation
Number is also just smaller.
The neighborhood similarity value factor is added, calculation formula becomes:
Wherein, λ is similarity value.Similarity is higher between neighborhood, then frame correlation coefficient is bigger, signal time Compound Degree
Also higher, electronic noise and speckle noise can be effectively reduced；, whereas if similarity degree is lower between neighborhood, show group
Knit or probe movement be violent, then frame correlation coefficient is smaller, and time Compound Degree is lower, can to avoid organizational boundary hangover and
The fuzzy distortion of image.
Time recombiner unit 16, for the 2D signal neighborhood of present frame 2D signal corresponding with previous or multiframe is adjacent
It is compound that domain passes through the obtained time recombination coefficient progress time.
It is compound respectively for each neighborhood progress time, time recombination coefficient is used for adjacent or non-adjacent two frame signal
The weighted superposition of middle corresponding neighborhood.
It in the present embodiment, further include storage unit, for 2D signal neighborhood of the time after compound to be stored in buffer area
Domain.
Signal after time is compound is stored into buffer zone, the signal of present frame be in buffer zone former frame or
It is compound that person is separated by several frames progress times.
Each neighborhood of present frame is weighted superposition with previous or multiframe corresponding neighborhood in caching, and weighting coefficient is frame
Relevant parameter α.
Compared with prior art, beneficial effect is the present embodiment:
1. the present embodiment is to do time compound processing on 2D signal, the number of scan points that this mode is handled compared to
It is smaller that calculation amount is processed on the image.
2. the present embodiment is to set the recombination coefficient of each neighborhood based on the severe degree of tissue and probe movement, than
The prior art is accurate.
3. the present embodiment can effectively inhibit the electronic noise of system, smooth speckle noise, while being able to maintain image
Edge feature, avoid due to tissue and probe movement it is too fast caused by combination picture edge blurry even trail.
Embodiment two
It should be noted that a kind of executing subject of time set composite based on ultrasonic sequence similarity of the present embodiment
For ultrasonic system.
A kind of time complex method based on ultrasonic sequence similarity provided in this embodiment, as shown in Figure 2, comprising:
S11. 2D signal is subjected to low-pass filtering and edge enhancing is handled；
S12. by treated, 2D signal is divided into the identical rectangular neighborhood of several sizes；
S13. several 2D signal neighborhoods after the segmentation are divided into edge tissues class, uniform formation's class, noise class；
S14. calculate present frame 2D signal neighborhood and former frame corresponding to similarity between 2D signal neighborhood；
S15. the similarity calculation time recombination coefficient obtained according to；
S16. by the 2D signal neighborhood of present frame 2D signal neighborhood corresponding with previous or multiframe pass through obtain when
Between recombination coefficient carry out the time it is compound.
It is described that several 2D signal neighborhoods are divided into edge tissues class, uniform formation's class, noise class is by preset group
Knit threshold value and Grads threshold classification；If being lower than the tissue threshold value, for noise class；If being higher than the tissue threshold value, for
Edge tissues class and uniform formation's class；If being lower than the Grads threshold, for uniform formation's class；If being higher than the Grads threshold,
It is then edge tissues class.
The step S13 further include:
Edge tissues class, uniform formation's class, time recombination coefficient corresponding to noise class is respectively set.
It is further comprised the steps of: after the step S16
S17. the 2D signal neighborhood by the time after compound is stored in buffer zone.
The 2D signal neighborhood of the present frame is that 2D signal corresponding to former frame or multiframe with buffer zone is adjacent
It is compound that domain carries out the time；The complex method is corresponding to the 2D signal neighborhood of present frame and the previous or multiframe of buffer zone
2D signal neighborhood by weighted superposition it is compound.
It should be noted that method provided in this embodiment is corresponding with embodiment one a kind of based on ultrasonic sequence phase
Like the time complex method of degree, specific embodiment is similar with embodiment one, and this will not be repeated here.
Compared with prior art, beneficial effect is the present embodiment:
1. the present embodiment is to do time compound processing on 2D signal, the number of scan points that this mode is handled compared to
It is smaller that calculation amount is processed on the image.
2. the present embodiment is to set the recombination coefficient of each neighborhood based on the severe degree of tissue and probe movement, than
The prior art is accurate.
3. the present embodiment can effectively inhibit the electronic noise of system, smooth speckle noise, while being able to maintain image
Edge feature, avoid due to tissue and probe movement it is too fast caused by combination picture edge blurry even trail.
Note that the above is only a better embodiment of the present invention and the applied technical principle.It will be appreciated by those skilled in the art that
The invention is not limited to the specific embodiments described herein, be able to carry out for a person skilled in the art it is various it is apparent variation,
It readjusts and substitutes without departing from protection scope of the present invention.Therefore, although being carried out by above embodiments to the present invention
It is described in further detail, but the present invention is not limited to the above embodiments only, without departing from the inventive concept, also
It may include more other equivalent embodiments, and the scope of the invention is determined by the scope of the appended claims.
Claims (10)
1. a kind of time set composite based on ultrasonic sequence similarity characterized by comprising
Signal Pretreatment unit, for 2D signal to be carried out low-pass filtering and edge enhancing processing；
Signal cutting unit, for 2D signal to be divided into the identical rectangular neighborhood of several sizes by treated；
Taxon, for several 2D signal neighborhoods after the segmentation to be divided into edge tissues class, uniform formation's class, are made an uproar
Sound class；
First computing unit, for calculate present frame 2D signal neighborhood and former frame corresponding between 2D signal neighborhood
Similarity；
Second computing unit, the similarity calculation time recombination coefficient for being obtained according to；
Time recombiner unit, for passing through the 2D signal neighborhood of present frame 2D signal neighborhood corresponding with previous or multiframe
It is compound that obtained time recombination coefficient carries out the time.
2. a kind of time set composite based on ultrasonic sequence similarity according to claim 1, which is characterized in that described
Several 2D signal neighborhoods are divided into edge tissues class, uniform formation's class, noise class in taxon and pass through preset tissue threshold
Value and Grads threshold classification；If being lower than the tissue threshold value, for noise class；If being higher than the tissue threshold value, for edge group
Knit class and uniform formation's class；If being lower than the Grads threshold, for uniform formation's class；If being higher than the Grads threshold, for side
Edge tissue class.
3. a kind of time set composite based on ultrasonic sequence similarity according to claim 2, which is characterized in that described
Taxon further includes that the corresponding time recombination coefficient of edge tissues class, uniform formation's class, noise class is respectively set.
4. a kind of time set composite based on ultrasonic sequence similarity according to claim 1 or 2 or 3, feature exist
In, further includes:
Storage unit, for 2D signal neighborhood of the time after compound to be stored in buffer zone.
5. a kind of time set composite based on ultrasonic sequence similarity according to claim 4, which is characterized in that described
It is multiple that the 2D signal neighborhood of present frame is that 2D signal neighborhood corresponding to former frame or multiframe with buffer zone carries out the time
It closes；The complex method is the 2D signal neighborhood 2D signal neighborhood corresponding with the previous or multiframe of buffer zone of present frame
It is compound by weighted superposition.
6. a kind of time complex method based on ultrasonic sequence similarity, which is characterized in that comprising steps of
S1. 2D signal is subjected to low-pass filtering and edge enhancing is handled；
S2. by treated, 2D signal is divided into the identical rectangular neighborhood of several sizes；
S3. several 2D signal neighborhoods after the segmentation are divided into edge tissues class, uniform formation's class, noise class；
S4. calculate present frame 2D signal neighborhood and former frame corresponding to similarity between 2D signal neighborhood；
S5. the similarity calculation time recombination coefficient obtained according to；
S6., it is compound that the 2D signal neighborhood of present frame 2D signal neighborhood corresponding with previous or multiframe is passed through to the obtained time
It is compound that coefficient carries out the time.
7. a kind of time complex method based on ultrasonic sequence similarity according to claim 6, which is characterized in that described
Several 2D signal neighborhoods are divided into edge tissues class, uniform formation's class, noise class and pass through preset tissue threshold value and gradient threshold
Value classification；If being lower than the tissue threshold value, for noise class；If being higher than the tissue threshold value, for edge tissues class and uniformly
Organize class；If being lower than the Grads threshold, for uniform formation's class；If being higher than the Grads threshold, for edge tissues class.
8. a kind of time complex method based on ultrasonic sequence similarity according to claim 7, which is characterized in that described
Step S3 further include:
The corresponding time recombination coefficient of edge tissues class, uniform formation's class, noise class is respectively set.
9. a kind of time complex method based on ultrasonic sequence similarity, feature described according to claim 6 or 7 or 8 exist
In being further comprised the steps of: after the step S6
S7. the 2D signal neighborhood by the time after compound is stored in buffer zone.
10. a kind of time complex method based on ultrasonic sequence similarity according to claim 9, which is characterized in that institute
The 2D signal neighborhood for stating present frame is that 2D signal neighborhood corresponding to former frame or multiframe with buffer zone carries out the time
It is compound；The complex method is that the 2D signal neighborhood 2D signal corresponding with the previous or multiframe of buffer zone of present frame is adjacent
Domain is compound by weighted superposition.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910490844.1A CN110415179A (en) | 2019-06-06 | 2019-06-06 | A kind of time complex method and device based on ultrasonic sequence similarity |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910490844.1A CN110415179A (en) | 2019-06-06 | 2019-06-06 | A kind of time complex method and device based on ultrasonic sequence similarity |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110415179A true CN110415179A (en) | 2019-11-05 |
Family
ID=68358249
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910490844.1A Pending CN110415179A (en) | 2019-06-06 | 2019-06-06 | A kind of time complex method and device based on ultrasonic sequence similarity |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110415179A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112734665A (en) * | 2020-12-31 | 2021-04-30 | 飞依诺科技(苏州)有限公司 | Ultrasonic image time-domain filtering method, device and medium for reducing motion blur |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102915537A (en) * | 2012-09-14 | 2013-02-06 | 飞依诺科技(苏州)有限公司 | Frame-based ultrasonic image processing method and system |
| CN109767400A (en) * | 2019-01-14 | 2019-05-17 | 三峡大学 | A kind of ultrasound image speckle noise minimizing technology of three sides of guiding filtering |
-
2019
- 2019-06-06 CN CN201910490844.1A patent/CN110415179A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102915537A (en) * | 2012-09-14 | 2013-02-06 | 飞依诺科技(苏州)有限公司 | Frame-based ultrasonic image processing method and system |
| CN109767400A (en) * | 2019-01-14 | 2019-05-17 | 三峡大学 | A kind of ultrasound image speckle noise minimizing technology of three sides of guiding filtering |
Non-Patent Citations (1)
| Title |
|---|
| 郭德全等: "基于序列相关性的超声图像自适应去噪", 《计算机应用研究》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112734665A (en) * | 2020-12-31 | 2021-04-30 | 飞依诺科技(苏州)有限公司 | Ultrasonic image time-domain filtering method, device and medium for reducing motion blur |
| CN112734665B (en) * | 2020-12-31 | 2023-07-14 | 飞依诺科技股份有限公司 | Ultrasonic image time domain filtering method, equipment and medium for weakening motion blur |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Sudha et al. | Speckle noise reduction in ultrasound images by wavelet thresholding based on weighted variance | |
| CN102306377B (en) | Method and device for reducing noise in ultrasound image | |
| Michailovich et al. | A novel approach to the 2-D blind deconvolution problem in medical ultrasound | |
| JP5449852B2 (en) | Ultrasonic diagnostic equipment | |
| KR101819028B1 (en) | Method and apparatus for processing a ultrasound image | |
| CN101527034A (en) | Method and device for correlating adaptive frames | |
| CN101459766B (en) | Method for ultrasonic image reinforcement and noise suppression, and device thereof | |
| CN105205788B (en) | A kind of denoising method for high-throughput gene sequencing image | |
| US20120108973A1 (en) | Ultrasonic diagnostic apparatus and ultrasonic image processing apparatus | |
| US20100280378A1 (en) | Ultrasonographic device and method for improving ultrasonographic device image quality | |
| CN110800019B (en) | Method and system for composite ultrasound image generation | |
| CN109598680A (en) | Shearing wave conversion medicine CT image denoising method based on quick non-local mean and TV-L1 model | |
| CN108354627B (en) | Ultrasonic beam forming method for improving frame frequency | |
| Magud et al. | Medical ultrasound image speckle noise reduction by adaptive median filter | |
| EP3072104B1 (en) | Image de-noising method | |
| Bhardwaj et al. | A Novel Method for Despeckling of Ultrasound Images Using Cellular Automata-Based Despeckling Filter | |
| CN110415179A (en) | A kind of time complex method and device based on ultrasonic sequence similarity | |
| US20080097212A1 (en) | Automatic identification of orientation in medical diagnostic ultrasound | |
| Ferraioli et al. | A novel cost function for despeckling using convolutional neural networks | |
| Hiremath et al. | Speckle reducing contourlet transform for medical ultrasound images | |
| Vijay et al. | Image denoising based on adaptive spatial and Wavelet Thresholding methods | |
| CN106725612B (en) | Four-dimensional ultrasonic image optimization method and system | |
| Liu | Study on medical image enhancement based on wavelet transform fusion algorithm | |
| Resham et al. | Noise reduction, enhancement and classification for sonar images | |
| Tiruwa et al. | Comparing various filtering techniques for reducing noise in MRI |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191105 |