CN103969651A - Self-adaptive acoustic imaging method - Google Patents

Abstract

The invention relates to a self-adaptive acoustic imaging method. The method includes the steps that acoustic wave signals are received after being transmitted at fixed focuses through an acoustic transducer array; focusing beamforming processing is carried out on the acoustic wave signals received by each array element in the acoustic transducer array to obtain acoustic radio frequency scanning line data; the focuses are used as virtual sources in synthetic aperture processing, tine delay from each virtual source to the spatial point is calculated by means of a cylindrical wave propagation model, and spatial point-by-point focusing is carried out on the scanning line data; self-adaptive weighting is conducted on the result of the spatial point-by-point focusing by the adoption of correlation coefficients or generalized correlation coefficients, the proportion of coherent components is increased, and the proportion of incoherent component is reduced; image transformation is carried out on all the scanning line data after self-adaptive weighting conducted on the result of the spatial point-by-point focusing to obtain images. According to the method, the lateral resolution and contrast ratio which are identical in depth are achieved, the number of sidelobes is effectively reduced, and therefore all-round improvement in acoustic imaging quality is achieved.

Description

Adaptive acoustic formation method

Technical field

The present invention relates to acoustic array imaging technique, relate in particular to a kind of adaptive acoustic formation method.

Background technology

Synthetic aperture acoustics imaging comes from Synthetic Aperture Radar Technique.Synthetic transmitting aperture imaging (Synthetic Transmitting Aperture Imaging, STAI) method is compared with the synthetic aperture imaging method of conventional acoustics imaging method and other modes, can improve frame per second, signal to noise ratio (S/N ratio) and contrast.But because STAI is higher to imaging device hardware requirement, particularly very high to the requirement of data transmission and processing speed, limit to a certain extent the commercialization of synthetic aperture acoustics imaging and promoted.Meanwhile, also there is transmitting array element energy shortage and the problem such as emitting times is too many in synthetic transmitting aperture imaging.Fixed-focus transmitting in conventional acoustics imaging, the strategy of reception dynamic focusing, around image resolution is higher to make to launch focus, but along with investigation depth increases, the main lobe width of wave beam increases, and lateral resolution reduces.Although synthetic aperture imaging can reach transmitting and receive dynamic focusing simultaneously in whole plane, form resolution and contrast relatively uniformly, but imaging process needs each array element to launch successively, receives successively, image taking speed is slow and signal to noise ratio (S/N ratio) is lower, electronic system complexity.

In order to overcome above shortcoming, C.Passmann and H.Ermert have provided " virtual source " (Virtual Source, VS) concept in the time within 1996, proposing synthetic aperture focusing technology.Then adopt the sub-aperture that many array element forms to focus on generation VS for improving signal to noise ratio (S/N ratio) and the resolution of imaging by C.H.Frazier.The VS reconstruct acoustic picture that M.H.Bae etc. utilize linear array to produce, improves image quality.Recently, the proposition sequence wave beams such as J.Kortbek are formed into picture, and virtual source building-up process is thought to second wave beam forms.M.Sutcliffe etc. have widened the application of virtual source compound imaging aspect industrial detection.The research of domestic virtual source imaging is the double focusing beam synthesizing method based on Virtual array in ultrasonic imaging.The research of virtual source technology shows: the synthetic aperture imaging based on virtual source can reduce the complexity of front end electronics, and can obtain between launching, receive full dynamic focusing method and fixed-focus transmitting, receiving the performance between dynamic depth focus method.

It is the coherence stack of the low resolution image signal of specific objective that acoustics imaging based on virtual source can obtain the more prerequisite of high resolving power and signal to noise ratio (S/N ratio), but this condition not can be met in actual applications.By can see the analysis of image-forming principle and process, not only comprised the signal of specific objective based on the synthetic stack of acoustic propagation time, also comprise other and had the echoed signal of the nonspecific target of same propagation time.Non-coherent addition can cause the side lobe levels in some driftlessness region to improve.Taking point target as example: when calculation level target is around when imaging results, the scattered signal stack of point target enters result, thereby produce higher pseudo-secondary lobe, the reason that these pseudo-secondary lobes produce is just that each virtual source scan-line data of stack is often not exclusively relevant, but the amplitude after stack is not low.Therefore, under relatively simple hardware condition, need to propose a kind of formation method, for realizing lateral resolution and the contrast that the degree of depth is consistent, and effectively reduce pseudo-secondary lobe.

Summary of the invention

The object of this invention is to provide a kind of adaptive acoustic formation method, after the collection and wave beam formation processing of the virtual source to spatial point, by coefficient of coherence and the stack to spatial point point-by-point focusing again of broad sense coefficient of coherence, to obtain the consistent image of the degree of depth based on virtual source algorithm, and suppress non-coherent addition based on coefficient of coherence and broad sense coefficient of coherence.

For achieving the above object, the invention provides a kind of adaptive acoustic formation method, the method comprises: receive acoustic signals by the acoustic transducer array transmitting that focuses after acoustic signals; Each array element in acoustic transducer array is received to the acoustic signals obtaining and carry out focus beam formation processing, obtain sound radio frequency reading line data; Virtual source using focus in synthetic aperture processing, adopts cylindrical wave propagation model to calculate the time delay of each virtual source to spatial point, and scan-line data is carried out to space point-by-point focusing; Adopt coefficient of coherence or broad sense coefficient of coherence to carry out the result of space point-by-point focusing adaptive weighted, improve the ratio of phase dry ingredients, reduce the ratio of incoherent composition; All scan-line datas that the result of space point-by-point focusing is carried out after adaptive weighted carry out image conversion, obtain image.

Preferably, method can be expressed as following formula:

$H(x,z)=A(x,z)\underset{k=0}{\overset{K\left(z\right)}{\mathrm{\Σ}}}W(x_k,z)s_{x_k}\left(t({\stackrel{\→}{r}}_{x_k},{\stackrel{\→}{r}}_s)\right)$

Wherein z is the degree of depth, and H (x, z) is image, and A (x, z) is the adaptive weighted coefficient of coherence with spatial variations, W (x _k, z) be x _kthe pointwise weighted value of place's virtual source, for the low-resolution scan line through pointwise time delay, represent virtual source to scattering point travel-time.

Preferably, receive acoustic signals after acoustic signals and specifically comprise by the acoustic transducer array transmitting that focuses: in the time that acoustic transducer array is linear array, sound wave is launched respectively in sub-aperture to the each array element in acoustic transducer array, and focus is positioned on sub-aperture center line.

Preferably, receive acoustic signals after acoustic signals and specifically comprise by the acoustic transducer array transmitting that focuses: in the time that acoustic transducer array is phased array, to the full aperture transmitting sound wave of the array element in acoustic transducer array, in a degree of depth of acoustic transducer array according to equal angles interval or etc. horizontal interval mode form focus.

Preferably, before the position of focus can be positioned at acoustic transducer array or after acoustic transducer array.

Preferably, each array element in acoustic transducer array is received to the acoustic signals obtaining and carry out focus beam formation processing, obtaining sound radio frequency reading line data specifically comprises: the virtual source echo that covers at first is carried out to coherence stack, obtain first scan synthesis line signal of locating and be:

$H(x,z)=\underset{k=1}{\overset{K\left(z\right)}{\mathrm{\Σ}}}W(x_k,z)I_{x_k}\left(z\right)$

Wherein z is the degree of depth, H (x, z) presentation video; represent that virtual source horizontal coordinate is x _klow resolution scan-line data, by rf data directly form: $I_{x_k}\left(z\right)=s_{x_k}\left(z^{\′}\right),$ Wherein $z^{\′}=d({\stackrel{\→}{r}}_{x_k},{\stackrel{\→}{r}}_s),$ Represent scattering point at virtual source the propagation distance of sweep trace; Variable W (x _k, z) be x _kthe pointwise weighted value of place's virtual source is the function of K (z); K (z) represents the virtual source number that z depth need superpose, and can obtain K (z) according to following formula:

$K\left(z\right)=\frac{L\left(z\right)}{d}=\frac{2|z-z_f|\tan (\mathrm{\θ}/2)}{d}=\frac{|z-z_f|F_{\#}}{d}$

Wherein L (z) be first virtual source depth z place cover lateral separation, θ is the front and back radiation angle of the virtual source of first, F _#for the F number of sub-aperture focusing.

Preferably, adopt coefficient of coherence to carry out adaptive weightedly specifically comprising to the result of space point-by-point focusing: to carry out adaptive weighted by following formula to the result of space point-by-point focusing: A (x, z)=CF, wherein, CF is coefficient of coherence,

$\mathrm{CF}=\frac{{\left(\underset{m=0}{\overset{N-1}{\mathrm{\Σ}}}\right|p\left(m\right)\left|\right)}^2}{\underset{m=0}{\overset{N-1}{\mathrm{\Σ}}}{\left|p\left(m\right)\right|}^2}$

Wherein m represents virtual source x _k, p (m) represents virtual source x _kcorresponding sweep trace ? the value in moment, N represents the number of scanning lines K (z) with the participation point-by-point focusing of change in depth.

Preferably, adopt coefficient of coherence or broad sense coefficient of coherence to carry out adaptive weightedly specifically comprising to the result of space point-by-point focusing: to carry out Fourier transform to array element numeric field data:

$p\left(m\right)\text{=}\underset{i=0}{\overset{M-1}{\mathrm{\Σ}}}{x}_i\left(k\right)e^{-j2\mathrm{\π}(k-\frac{M}{2})d\frac{m}{\mathrm{Md}}}=e^{\mathrm{j\πm}}\underset{i=0}{\overset{M-1}{\mathrm{\Σ}}}{x}_i\left(k\right)e^{-j2\mathrm{\πk}\frac{m}{M}}$

Wherein x _i(k) virtual source x _kcorresponding sweep trace ? the value in moment, p (m) is x _i(k) transform to the data after Beam Domain;

Calculating the energy of each beam direction according to following formula, obtain the relevant energy of direction and the ratio of gross energy, is broad sense coefficient of coherence GCF:

$\mathrm{GCF}\left(k\right)=\frac{\underset{m\∈(\mathrm{0,1},...,K)}{\mathrm{\Σ}}{\left|p\left(m\right)\right|}^2}{\underset{m=0}{\overset{M-1}{\mathrm{\Σ}}}{\left|p\left(m\right)\right|}^2}$

Wherein by $\underset{m=0}{\overset{M-1}{\mathrm{\Σ}}}{\left|p\left(m\right)\right|}^2=M\underset{i=0}{\overset{M-1}{\mathrm{\Σ}}}{\left|x_i\left(k\right)\right|}^2$ Known, denominator represent the gross energy of array element signals.

Preferably, the result of space point-by-point focusing being carried out to all scan-line datas after adaptive weighted carries out image conversion and specifically comprises: scan-line data is carried out to envelope processing; Scan-line data after envelope processing is carried out to interpolation; According to the data acquisition scan image after interpolation.

The present invention introduces adaptive weighted in the additive process of two-way space point-by-point focusing taking coherence as criterion, make imaging realize lateral resolution and the contrast that the degree of depth is consistent, and effectively reduce secondary lobe, thereby improve acoustics imaging quality comprehensively.

Brief description of the drawings

Fig. 1 is the process flow diagram of the adaptive acoustic formation method of the embodiment of the present invention one;

Fig. 2 is that the transmitting that focuses of the transducer linear array of the embodiment of the present invention one receives acoustic signals schematic diagram after acoustic signals;

Fig. 3 is that the transmitting that focuses of the phased acoustic transducer array of the embodiment of the present invention one receives acoustic signals schematic diagram after acoustic signals;

Fig. 4 is the acoustics imaging schematic diagram of the virtual source of the embodiment of the present invention one;

Fig. 5 is the scan-line data schematic diagram after the envelope processing of the embodiment of the present invention one;

Fig. 6 be the embodiment of the present invention one envelope signal is carried out to the schematic diagram of interpolation processing process;

Fig. 7 be the embodiment of the present invention one envelope signal is carried out to the schematic diagram after interpolation processing;

Fig. 8 is the point target image of 5 kinds of algorithms of the embodiment of the present invention one;

Fig. 9 be different depth under the algorithms of different of the embodiment of the present invention one scattering point-6dB laterally and axial width;

Figure 10 is the axial resolution schematic diagram of the scattering point of different depth under the algorithms of different of the embodiment of the present invention one;

Figure 11 is the algorithms of different of the embodiment of the present invention one the first secondary lobe height on different depth;

Figure 12 is different F number-6dB transverse widths under the same virtual source location of the embodiment of the present invention one;

Figure 13 is identical F number-6dB transverse width under the different virtual source location of the embodiment of the present invention one;

Figure 14 is the imaging results schematic diagram of 5 kinds of algorithms of the embodiment of the present invention one;

Figure 15 is the relatively schematic diagram of result of 3 kinds of imaging algorithms of the embodiment of the present invention one;

Figure 16 is the result schematic diagram of array energy transducer that the radius-of-curvature of the embodiment of the present invention one the is 60mm point target imaging while focusing on 40mm.

Embodiment

Below by drawings and Examples, technical scheme of the present invention is described in further detail.

A kind of adaptive acoustic formation method provided by the invention, can make up conventional imaging intermediate-resolution along with the degree of depth increases the defect that picture quality fails, and under limited system complexity condition, obtains acoustics imaging more clearly.

Fig. 1 is the process flow diagram of the adaptive acoustic formation method of the embodiment of the present invention one, and as shown in Figure 1, adaptive acoustic formation method specifically comprises the following steps:

Step 101, receive acoustic signals by the acoustic transducer array transmitting that focuses after acoustic signals.

The acoustic signals receiving due to acoustic transducer derives from the acoustic signals of acoustic transducer transmitting, in conventional acoustics imaging often by different array element delay emissions are reached to focusing target.

Step 102, the each array element in acoustic transducer array is received to the acoustic signals obtaining carry out focus beam formation processing, obtain sound radio frequency reading line data.

Particularly, Fig. 2 is that the transducer linear array of the embodiment of the present invention one focuses after transmitting acoustic signals and receives acoustic signals schematic diagram.As shown in Figure 2, taking two sweep traces as example, two sub-aperture L in figure, have been marked _sub1and L _sub2, corresponding sweep trace is L ₀and L ₁.Sound wave is launched respectively in sub-aperture, focuses on f ₀and f ₁, be similar to and have a virtual point sound source at focus place with the two-way transmitting sound wave of certain angle of release.Suppose at L ₀and L ₁between there is a scattering point can in these two sweep traces, there is the scattered signal of this point.If can effectively excavate and utilize the scattered information between sweep trace, just likely improve imaging performance.

Fig. 3 is that the transmitting that focuses of the phased acoustic transducer array of the embodiment of the present invention one receives acoustic signals schematic diagram after acoustic signals.As shown in Figure 3, different from the work of transducer linear array, phased array produces focus point f by electronic delay control ₀and f ₁, the distribution of focus can distribute in the first-class horizontal range of the same degree of depth, also can the equal angles under same radius distribute.

Especially need explanation: Fig. 2 and Fig. 3 just illustrate the distribution situation of focus not represent that focus necessarily needs cloth to be placed on before array, the method comprises equally focus cloth is placed on to the situation that focuses on processing after array.

Step 103, virtual source using focus in synthetic aperture processing, adopt cylindrical wave propagation model to calculate the time delay of each virtual source to spatial point, and scan-line data is carried out to space point-by-point focusing.

Particularly, Fig. 4 is the acoustics imaging schematic diagram of the virtual source of the embodiment of the present invention one.As shown in Figure 4, three sub-aperture L that mark _sub1, L _sub2and L _sub3, for transmitting and receiving, focus on respectively locus A, B and C, corresponding to virtual source A, B and C.The lateral separation of adjacent virtual source is d, and focal length is z _f.Wei Zi aperture, thick dashed line overlay area L in figure _sub2overlay area under ray approximation.

Use virtual source to obtain sweep trace and further improve image quality.Suppose that having one at depth z place is scattering into picture point the scattered signal of this point is not only at sub-aperture L _sub2in the echoed signal of corresponding virtual source B, exist, and also exist in the echoed signal of virtual source A and C.Therefore, the point target imaging diffusion zone that shows as z place in imaging results is larger, and resolution and contrast are lower.Particularly, along with the increase of the degree of depth, the energy of sweep trace that covers more and each root of the z sweep trace of ordering is more weak, and image resolution ratio and signal to noise ratio (S/N ratio) are poorer.

At virtual source B sweep trace mid point travel-time be:

$t_{x_B}\left(z\right)=({2z}_f\±2|z-z_f\left|\right)/c=2z/c,---\left(1\right)$

Wherein z _ffor sub-aperture L _sub2to the axial distance of virtual source B, ± representing that scattering point is before or after virtual source, c represents the velocity of sound.

Without loss of generality, if the region of contiguous virtual source C covers scattering point with respect to the travel-time of virtual source C be:

$t({\stackrel{\→}{r}}_{x_C},{\stackrel{\→}{r}}_s)=({2z}_f\±2|{\stackrel{\→}{r}}_{x_C}-{\stackrel{\→}{r}}_s\left|\right)/c=(2z_f\±2\sqrt{{(z-z_f)}^2+{\left(\mathrm{Md}\right)}^2})/c,---\left(2\right)$

Wherein Md represents the lateral separation of virtual source C and virtual source B.

In order to improve resolution and contrast, the virtual source echo coherence stack of all covering point z, obtain the composite signal at z point place.Scanning-line signal after synthesizing is:

$H(x,z)=\underset{k=1}{\overset{K\left(z\right)}{\mathrm{\Σ}}}W(x_k,z)I_{x_k}\left(z\right),---\left(3\right)$

Wherein represent that virtual source horizontal coordinate is x _klow resolution scan-line data, H (x, z) represents high-definition picture.Variable W (x _k, z) be x _kthe pointwise weighted value of place's virtual source is the function of K (z).K (z) represents the virtual source number that z depth need superpose:

$K\left(z\right)=\frac{L\left(z\right)}{d}=\frac{2|z-z_f|\tan (\mathrm{\θ}/2)}{d}=\frac{|z-z_f|F_{\#}}{d},---\left(4\right)$

The lateral separation that wherein L (z) covers in depth z place for some virtual source, θ is the front and back radiation angle of a virtual source, F _#for the F number of sub-aperture focusing.As can be seen from the above equation, along with the increase of the degree of depth, the number of scanning lines that need to include calculating in is more and more.Low-resolution image is by rf data directly form:

$I_{x_k}\left(z\right)=s_{x_k}(z^{\′}/c),---\left(5\right)$

Wherein represent scattering point at virtual source the travel-time of sweep trace.Last high-definition picture is:

$H(x,z)=\underset{k=1}{\overset{K\left(z\right)}{\mathrm{\Σ}}}W(x_k,z)s_{x_k}\left(t({\stackrel{\→}{r}}_{x_k},{\stackrel{\→}{r}}_s)\right),---\left(6\right)$

At virtual source place, do not have adjacent sweep trace to participate in stack, so this place's picture quality is the same with classic method.But along with imaging point is away from focus, the number of scanning lines that participates in point-by-point focusing is more and more, and image quality improves.Before and after focus, the symbol of point-by-point focusing time delay is contrary, claims that this process is two-way space point-by-point focusing.

Step 104, employing coefficient of coherence or broad sense coefficient of coherence carry out adaptive weighted to the result of space point-by-point focusing, improve the ratio of phase dry ingredients, reduce the ratio of incoherent composition.

It is the coherence stack of the low resolution image signal of specific objective that above-mentioned two-way space point-by-point focusing imaging can be obtained the more prerequisite of high resolving power and signal to noise ratio (S/N ratio).But this condition not can be met in actual applications, even can say always unappeasable.By can see the analysis of image-forming principle and process, not only comprised the signal of specific objective based on the synthetic stack of acoustic propagation time, also comprise other and had the echoed signal of the nonspecific target of same propagation time.Non-coherent addition can cause the side lobe levels in some driftlessness region to improve.Taking point target as example: when calculation level target is around when imaging results, the scattered signal stack of point target enters result, thereby produces higher pseudo-secondary lobe.The reason that these pseudo-secondary lobes produce is just that each virtual source scan-line data of stack is often not exclusively relevant, but the amplitude after stack is not low.Therefore, further proposition is introduced adaptive weighted in the additive process of two-way space point-by-point focusing taking coherence as criterion, thereby reduces pseudo-secondary lobe.Adaptive weighted space point-by-point focusing method can adopt coefficient of coherence, broad sense coefficient of coherence or the two group and carry out adaptive weighted.

Traditional coefficient of coherence (CF) is defined as:

$\mathrm{CF}=\frac{{\left(\underset{m=0}{\overset{N-1}{\mathrm{\Σ}}}\right|p\left(m\right)\left|\right)}^2}{\underset{m=0}{\overset{N-1}{\mathrm{\Σ}}}{\left|p\left(m\right)\right|}^2},---\left(7\right)$

Wherein p (m) represents that array element is the data of m signal through suitable time delays.N represents to participate in the port number that wave beam forms.The physical significance of CF is composite signal coherent energy and the ratio of gross energy, and it can be used as the tolerance of composite signal degree of coherence.The weighting coefficient that adopts in the present invention CF to superpose in building-up process as image adaptive: A (x, z)=CF, p (m) wherein no longer represents that array element is the data of m signal through suitable time delays, but with virtual source x _kcorresponding sweep trace ? the value in moment is replaced, and N replaces the number of scanning lines K (z) with the participation point-by-point focusing of change in depth.

Adaptive weighted for broad sense coefficient of coherence, first array element numeric field data is carried out to Fourier transform:

$p\left(m\right)\text{=}\underset{i=0}{\overset{M-1}{\mathrm{\Σ}}}{x}_i\left(k\right)e^{-j2\mathrm{\π}(k-\frac{M}{2})d\frac{m}{\mathrm{Md}}}=e^{\mathrm{j\πm}}\underset{i=0}{\overset{M-1}{\mathrm{\Σ}}}{x}_i\left(k\right)e^{-j2\mathrm{\πk}\frac{m}{M}}---\left(8\right)$

Wherein p (m) is for transforming to the data after Beam Domain.Then calculate the energy of each beam direction, obtain the relevant energy of direction and the ratio of gross energy, be broad sense coefficient of coherence:

$\mathrm{GCF}\left(k\right)=\frac{\underset{m\∈(\mathrm{0,1},...,K)}{\mathrm{\Σ}}{\left|p\left(m\right)\right|}^2}{\underset{m=0}{\overset{M-1}{\mathrm{\Σ}}}{\left|p\left(m\right)\right|}^2}---\left(9\right)$

Wherein by $\underset{m=0}{\overset{M-1}{\mathrm{\Σ}}}{\left|p\left(m\right)\right|}^2=M\underset{i=0}{\overset{M-1}{\mathrm{\Σ}}}{\left|x_i\left(k\right)\right|}^2$ Known, denominator represent the gross energy of array element signals.Each measures represented physical meaning in the present invention for it, consistent with the definition of above-mentioned CF.

Step 105, all scan-line datas that the result of space point-by-point focusing is carried out after adaptive weighted carry out image conversion, obtain image.

Particularly, Fig. 5-Fig. 7 is the schematic diagram that scan-line data carries out image transform processes.Fig. 5 is the scan-line data schematic diagram after the envelope processing of the embodiment of the present invention one, Fig. 6 be the embodiment of the present invention one envelope signal is carried out to the schematic diagram of interpolation processing process, Fig. 7 be the embodiment of the present invention one envelope signal is carried out to the schematic diagram after interpolation processing.Scan-line data after envelope processing as shown in Figure 5, in order to obtain figure as shown in Figure 7, need to carry out interpolation processing to envelope signal, Fig. 6 has provided Interpolation Process, round dot is given data, and trigpoint needs interpolation to obtain, and utilizes 2*2 interpolation, or 4*4 interpolation, just can obtain the scan image as shown in Figure 7 after image conversion.

Adaptive acoustic formation method based on step 101-step 105 is described in embodiment mono-can be expressed as following formula:

$H(x,z)=A(x,z)\underset{k=0}{\overset{K\left(z\right)}{\mathrm{\Σ}}}W(x_k,z)s_{x_k}\left(t({\stackrel{\→}{r}}_{x_k},{\stackrel{\→}{r}}_s)\right),---\left(10\right)$

Wherein A (x, z) is for reflection coherence is with the adaptive weighted coefficient of spatial variations, and its effect is at the outstanding phase dry ingredients in synthetic process that superposes, suppresses incoherent composition; W (x _k, z) be weighting function; for the low-resolution scan line through pointwise time delay.

In order to prove the validity of adaptive acoustic formation method provided by the invention, according to corresponding space impulse theory, point target and capsule target are carried out to emulation, the performances such as horizontal stroke, longitudinal frame and the contrast of the various algorithms of quantitative examination.In all emulation, all adopt 192 array element 3.5MHz ultrasound lines arrays; Focus imaging algorithm and dynamic focusing imaging algorithm adopts the sub-aperture of fixing 64 passages; Algorithm based on virtual source is adjusted port number according to F number.

Table 1 is simulation parameter, because the algorithm based on virtual source and virtual source parameter have very large relation, so table 1, except having provided transducer parameters, gives the parameter such as focal length and F number.Simulation process comprises focal length 5mm, 10mm and 20mm, and different F numbers 0.5,1 and 1.5.

Table 1

Parameter	Symbol	Value
			Array number	M	192
Centre frequency	f 0	3.5MHz

Sampling rate	f s	40MSPS
			Array element interval	d	0.49mm
Relative bandwidth	B	60％
			Focal length	z f	5mm，10mm，20mm
F number	F #	0.5，1，1.5
			Maximum virtual source number	N	91

First research horizontal stroke, longitudinal frame.The performance of 5 algorithms has been compared in whole emulation, respectively: algorithm 1-transmitting dynamic focusing, reception dynamic focusing (dynT-dynR); Algorithm 2-transmitting fixed-focus, receives dynamic focusing (fixT-dynR); Algorithm 3-transmitting fixed-focus, reception fixed-focus (fixT-fixR); The acoustics imaging (VS-BIPF) of algorithm 4-based on virtual source; The adaptive acoustic formation method (VS-BIAPF) (the inventive method) of algorithm 5-based on virtual source.

Fig. 8 is the point target image of 5 kinds of algorithms of the embodiment of the present invention one.As shown in Figure 8,10 point target horizontal directions are positioned at picture centre, and the vertical direction degree of depth, taking 5mm as initial, is spaced apart 10mm.A is the imaging results of 64 passage dynT-dynR algorithms.Because array aperture is fixed, higher near field resolution, far field resolution declines.And because array spacings is larger, near field place, produce graing lobe structure.B is that F number is the imaging results that 1.5 focuses are positioned at the fixT-dynR algorithm of 40mm.Near image resolution focus is the highest, leaves focus resolution decline later.C is that F number is the imaging results that 1.5 focuses are positioned at the fixT-fixR algorithm of 10mm.Obviously, away from the image resolution ratio of focus worse and worse.D is the imaging results of VS-BIPF algorithm.Can find out, the resolution of this algorithm obviously improves, but some points between 10-70mm, there is secondary lobe in left and right.E is the result of VS-BIAPF algorithm.Contrast c, d, e can find out, adopt adaptive weighted processing to weaken the side lobe effect existing in VS-BIPF.

For the further resolution of Quantitative study algorithms of different, Fig. 9 be different depth under the algorithms of different of the embodiment of the present invention one scattering point-6dB laterally and axial width.As shown in Figure 9, fixT-fixR algorithm lateral resolution is the poorest, and along with the degree of depth strengthens, resolution worse and worse; Then be fixT-dynR and dynT-dynR algorithm successively, along with the degree of depth strengthens, laterally-6dB width also increases, but recruitment is less, and before and after transmitting focus, lateral resolution is suitable; VS-BIPF algorithm can keep the resolution that the degree of depth is consistent, poorer than dynT-dynR algorithm in the time that the degree of depth is less than focal length, but is better than it while being greater than focal length; VS-BIAPF algorithm can keep constant lateral resolution equally, and resolution is better than other 4 kinds of algorithms.Figure 10 is the axial resolution schematic diagram of the scattering point of different depth under the algorithms of different of the embodiment of the present invention one, and the axial resolution difference of 5 kinds of algorithms is little as shown in figure 10, and this is consistent with the theory expectation of these algorithm principle.

In ultrasonic imaging, often characterize contrast with the first secondary lobe height.Figure 11 is the algorithms of different of the embodiment of the present invention one the first secondary lobe height on different depth.As shown in figure 11, dynT-dynR algorithm and fixT-dynR algorithm are too near owing to focusing on, higher at the secondary lobe of the point target of near field, but in darker region lower than-60dB; FixT-fixR algorithm is in the situation that keeping F number to be 1.5, and corresponding effective aperture is less, so secondary lobe height is lower; For VS-BIPF algorithm, the incoherence of additive process causes secondary lobe elevated height, exceedes-60dB in figure; And VS-BIAPF algorithm adopts adaptive coherent stack to suppress the weak composition of the non-impact point coherence of place, effectively reduce secondary lobe height.

It needs to be noted: under above simulation parameter condition, the F number of VS-BIAPF is 1.5, and focal length is 10mm, its effective aperture is only corresponding to 17 array elements, and the effective aperture of contrast conventional method is 64 array elements.Only need the port number of conventional method 1/4 can obtain suitable performance (even better), this has greatly reduced the complexity of hardware front end.

The key parameter of VS-BIPF or VS-BIAPF algorithm affects lateral resolution is focal length and F number, below study respectively under the same focal length, different F number and identical F number, different focal laterally-6dB width.Figure 12 is different F number-6dB transverse widths under the same virtual source location of the embodiment of the present invention one.As shown in figure 12, no matter for VS-BIPF or VS-BIAPF, along with F number increases, resolution all declines.But under identical F number, the lateral resolution of VS-BIAPF is higher than VS-BIPF.

Figure 13 is identical F number-6dB transverse width under the different virtual source location of the embodiment of the present invention one.As shown in figure 13, first for VS-BIPF and VS-BIAPF, under identical F number, focal length variations is less for lateral resolution impact; Secondly under identical F number, no matter focal length how far, and the lateral resolution of VS-BIAPF is all better than VS-BIPF.

For quantitative examination VS-BIAPF algorithm is for the image quality of capsule target, the strong scattering circle of three 5mm diameters of design and sound absorption circle: the intensity of strong scattering circle is background 10 times, lateral coordinates is-5mm that axial coordinate is respectively 25mm, 45mm and 65mm; In sound absorption circle scope, do not comprise scattering point, lateral coordinates is 5mm, and axial coordinate is respectively 15mm, 35mm and 55mm.

Desirable imaging results should be that strong scattering target is full of target area with black, edge clear, accurate, undistorted with white, sound absorption target.Figure 14 is the imaging results schematic diagram of 5 kinds of algorithms of the embodiment of the present invention one.In Figure 14, a is the result of dynT-dynR algorithm, although the performance of distant object declines to some extent, can both reflect clear, exactly all targets; In Figure 14, b is the result of fixT-dynR algorithm, and along with distance, to increase hydraulic performance decline more obvious, but still can all targets of clear identification; In Figure 14, c is the result of fixT-fixR algorithm, because focal position is at 10mm, only has the close-in target still can identification, and the strong scattering circle target of farther place is seriously overflowed distortion, and sound absorption circle target at a distance cannot identification; In Figure 14, d is the result of VS-BIPF algorithm, and all targets are all high-visible, and the relative first two algorithm of subjective vision effect of different target has excellently to be had badly, and main manifestations is the difference in contrast; In Figure 14, e is the result of VS-BIAPF algorithm, from subjective vision effect, and the object edge distortion minimum of this algorithm, background intensity is by reduced overall, and the contrast of strong scattering circle target generally increases, and the contrast of sound absorption circle target declines to some extent.

In order to quantize to explain the contrast of zones of different, it is the 1st, 2, No. 3 by degree of depth number consecutively respectively by two groups of targets.Then introduce contrast (Contrast Ratio, CR), be defined as the poor of the average power of center circle and the average power in external context region.Table 2 is algorithms of different contrasts to different target imaging.

Table 2

As can be seen from Table 2, for diverse location, dissimilar target, different imaging algorithm, their contrast is not identical.From quantizing contrast, for strong scattering target, VS-BIAPF is dominant generally, and particularly for distant object, not only contrast improves, and more even within the scope of entire depth; For sound absorption target, contrast overall trend is from big to small dynT-dynR, fixT-dynR, VS-BIPF, VS-BIAPF and fixT-fixR successively, now, because VS-BIAPF has significantly forced down background intensity, is unfavorable for highlighting sound absorption target.But picture shape, VS-BIAPF distortion minimum, although contrast decreases, target is full of whole red border circular areas truly, exactly, is easy to identification.

For performance and the effect of further verification algorithm, adopt open ultrasonic image-forming system platform USCAS-32 and the KS107BD body mould of Acoustical Inst., Chinese Academy of Sciences's development to carry out actual imaging experiment.USCAS-32 system supports 32 passages to work alone simultaneously, and maximum probe array number is 128.Radiating portion: high emission pulse can reach 200Vpp, the minimum adjustable delay of interchannel is 5ns; Receiving unit: bandwidth 15MHz is adjustable, high sampling rate is 50MSPS, Beam-former is supported single-point and point-by-point focusing; This system also supports static pre-wave beam to form data acquisition, can be used for various high-resolution ultrasonic imagings and the research of high frame per second imaging scheduling algorithm.In experimentation, adopting centre frequency is the 128 array element R60 abdominal ultrasonic probes of 3.5MHz.

Figure 15 is the relatively schematic diagram of result of 3 kinds of imaging algorithms of the embodiment of the present invention one.In Figure 15, a is the imaging results of fixT-fixR focus at 60mm.As can be seen from the figure, although all targets all can see, extending transversely very large away from the point target of focus, resolution obviously reduces, particularly at the point at 10mm-40mm place; The result that in Figure 15, b is VS-BIPF, can see that the resolution of near field significantly improves; The result that in Figure 15, c is VS-BIAPF, can obviously find out that resolution and contrast that the degree of depth is less than the point target of 60mm all improve a lot.But from this result, find, be greater than 60mm region for the degree of depth, resolution improves and is little, and contrast improves a little, and energy seems also to weaken to some extent.In order to explain this phenomenon, scrutinize the R60 probe adopting in experiment, Figure 16 is the result schematic diagram of array energy transducer that the radius-of-curvature of the embodiment of the present invention one the is 60mm point target imaging while focusing on 40mm.Point target imaging when the array energy transducer that radius-of-curvature of emulation is 60mm focuses on 40mm, result as shown in figure 16.Can find out, sound field is dispersed suddenly after 40mm, no longer meets virtual source hypothesis after focus is sentenced.Therefore, the experimental result of VS-BIPF and VS-BIAPF is consistent with the expection of theoretical and emulation before focus, and space point-by-point focusing is respond well.

To strengthen the problem of resolution decline along with the degree of depth in practical medical ultrasonic image-forming system, the present invention proposes a kind of coefficient of coherence or broad sense coefficient of coherence adaptive acoustic formation method based on virtual source.The virtual source that this algorithm makes full use of sub-array of apertures generation on the one hand carries out space point-by-point focusing, improves resolution; On the other hand in the point-by-point focusing process of space, the secondary lobe problem of utilizing the adaptive weighted inhibition non-coherent addition of coefficient of coherence or broad sense coefficient of coherence to bring.This method can also be generalized to other applications, and such as acoustic array is imaged on application, acoustic imaging in the water and the acoustic microscope etc. of Non-Destructive Testing.

Adaptive acoustic formation method provided by the invention, after the collection and wave beam formation processing of the virtual source to spatial point, by coefficient of coherence and the stack to spatial point point-by-point focusing again of broad sense coefficient of coherence, to obtain the consistent image of the degree of depth based on virtual source algorithm, and suppress non-coherent addition based on coefficient of coherence and broad sense coefficient of coherence.The present invention introduces adaptive weighted in the additive process of two-way space point-by-point focusing taking coherence as criterion, make imaging realize lateral resolution and the contrast that the degree of depth is consistent, and effectively reduce secondary lobe, thereby improve acoustics imaging quality comprehensively.

Professional should further recognize, unit and the algorithm steps of each example of describing in conjunction with embodiment disclosed herein, can realize with electronic hardware, computer software or the combination of the two, for the interchangeability of hardware and software is clearly described, composition and the step of each example described according to function in the above description in general manner.These functions are carried out with hardware or software mode actually, depend on application-specific and the design constraint of technical scheme.Professional and technical personnel can realize described function with distinct methods to each specifically should being used for, but this realization should not thought and exceeds scope of the present invention.

The software module that the method for describing in conjunction with embodiment disclosed herein or the step of algorithm can use hardware, processor to carry out, or the combination of the two is implemented.Software module can be placed in the storage medium of any other form known in random access memory (RAM), internal memory, ROM (read-only memory) (ROM), electrically programmable ROM, electrically erasable ROM, register, hard disk, moveable magnetic disc, CD-ROM or technical field.

Above-described embodiment; object of the present invention, technical scheme and beneficial effect are further described; institute is understood that; the foregoing is only the specific embodiment of the present invention; the protection domain being not intended to limit the present invention; within the spirit and principles in the present invention all, any amendment of making, be equal to replacement, improvement etc., within all should being included in protection scope of the present invention.

Claims (9)

1. an adaptive acoustic formation method, is characterized in that, described method comprises:

Receive described acoustic signals by the acoustic transducer array transmitting that focuses after acoustic signals;

Each array element in described acoustic transducer array is received to the acoustic signals obtaining and carry out focus beam formation processing, obtain sound radio frequency reading line data;

Virtual source using described focus in synthetic aperture processing, adopts cylindrical wave propagation model to calculate the time delay of each virtual source to spatial point, and described scan-line data is carried out to space point-by-point focusing;

Adopt coefficient of coherence or broad sense coefficient of coherence to carry out the result of described space point-by-point focusing adaptive weighted, improve the ratio of phase dry ingredients, reduce the ratio of incoherent composition;

All scan-line datas that the result of described space point-by-point focusing is carried out after adaptive weighted carry out image conversion, obtain image.

2. adaptive acoustic formation method according to claim 1, is characterized in that, described method can be expressed as following formula:

$H(x,z)=A(x,z)\underset{k=0}{\overset{K\left(z\right)}{\mathrm{\Σ}}}W(x_k,z)s_{x_k}\left(t({\stackrel{\→}{r}}_{x_k},{\stackrel{\→}{r}}_s)\right)$

Wherein z is the degree of depth, and H (x, z) is image, and A (x, z) is the adaptive weighted coefficient of coherence with spatial variations, W (x _k, z) be x _kthe pointwise weighted value of place's virtual source, for the low-resolution scan line through pointwise time delay, represent virtual source to scattering point travel-time.

3. adaptive acoustic formation method according to claim 1, it is characterized in that, describedly receive described acoustic signals after acoustic signals and specifically comprise by the acoustic transducer array transmitting that focuses: in the time that described acoustic transducer array is linear array, sound wave is launched respectively in sub-aperture to the each array element in described acoustic transducer array, and focus is positioned on sub-aperture center line.

4. adaptive acoustic formation method according to claim 1, it is characterized in that, describedly receive described acoustic signals after acoustic signals and specifically comprise by the acoustic transducer array transmitting that focuses: in the time that described acoustic transducer array is phased array, to the full aperture transmitting sound wave of the array element in described acoustic transducer array, in a degree of depth of described acoustic transducer array according to equal angles interval or etc. horizontal interval mode form focus.

5. according to the adaptive acoustic formation method described in claim 3 or 4, it is characterized in that, before the position of described focus can be positioned at described acoustic transducer array or after described acoustic transducer array.

6. adaptive acoustic formation method according to claim 1, is characterized in that, describedly each array element in described acoustic transducer array is received to the acoustic signals obtaining carries out focus beam formation processing, obtains sound radio frequency reading line data and specifically comprises:

The virtual source echo that covers at first is carried out to coherence stack, obtains described first scan synthesis line signal of locating and be:

$H(x,z)=\underset{k=1}{\overset{K\left(z\right)}{\mathrm{\Σ}}}W(x_k,z)I_{x_k}\left(z\right)$

Wherein z is the degree of depth, H (x, z) presentation video; represent that virtual source horizontal coordinate is x _klow resolution scan-line data, by rf data directly form: $I_{x_k}\left(z\right)=s_{x_k}\left(z^{\′}\right),$ Wherein $z^{\′}=d({\stackrel{\→}{r}}_{x_k},{\stackrel{\→}{r}}_s),$ Represent scattering point at virtual source the propagation distance of sweep trace; Variable W (x _k, z) be x _kthe pointwise weighted value of place's virtual source is the function of K (z); K (z) represents the virtual source number that z depth need superpose, and can obtain K (z) according to following formula:

$K\left(z\right)=\frac{L\left(z\right)}{d}=\frac{2|z-z_f|\tan (\mathrm{\θ}/2)}{d}=\frac{|z-z_f|F_{\#}}{d}$

Wherein L (z) be described first virtual source depth z place cover lateral separation, θ is the front and back radiation angle of the virtual source of described first, F _#for the F number of sub-aperture focusing.

7. adaptive acoustic formation method according to claim 1, is characterized in that, described employing coefficient of coherence carries out adaptive weightedly specifically comprising to the result of described space point-by-point focusing:

Carry out adaptive weighted by following formula to the result of described space point-by-point focusing: A (x, z)=CF, wherein, CF is coefficient of coherence,

$\mathrm{CF}=\frac{{\left(\underset{m=0}{\overset{N-1}{\mathrm{\Σ}}}\right|p\left(m\right)\left|\right)}^2}{\underset{m=0}{\overset{N-1}{\mathrm{\Σ}}}{\left|p\left(m\right)\right|}^2}$

Wherein m represents virtual source x _k, p (m) represents virtual source x _kcorresponding sweep trace ? the value in moment, N represents the number of scanning lines K (z) with the participation point-by-point focusing of change in depth.

8. adaptive acoustic formation method according to claim 1, is characterized in that, described employing coefficient of coherence or broad sense coefficient of coherence carry out adaptive weightedly specifically comprising to the result of described space point-by-point focusing:

Array element numeric field data is carried out to Fourier transform:

$p\left(m\right)\text{=}\underset{i=0}{\overset{M-1}{\mathrm{\Σ}}}{x}_i\left(k\right)e^{-j2\mathrm{\π}(k-\frac{M}{2})d\frac{m}{\mathrm{Md}}}=e^{\mathrm{j\πm}}\underset{i=0}{\overset{M-1}{\mathrm{\Σ}}}{x}_i\left(k\right)e^{-j2\mathrm{\πk}\frac{m}{M}}$

Wherein x _i(k) virtual source x _kcorresponding sweep trace ? the value in moment, p (m) is x _i(k) transform to the data after Beam Domain;

Calculating the energy of each beam direction according to following formula, obtain the relevant energy of direction and the ratio of gross energy, is broad sense coefficient of coherence GCF:

$\mathrm{GCF}\left(k\right)=\frac{\underset{m\∈(\mathrm{0,1},...,K)}{\mathrm{\Σ}}{\left|p\left(m\right)\right|}^2}{\underset{m=0}{\overset{M-1}{\mathrm{\Σ}}}{\left|p\left(m\right)\right|}^2}$

Wherein by $\underset{m=0}{\overset{M-1}{\mathrm{\Σ}}}{\left|p\left(m\right)\right|}^2=M\underset{i=0}{\overset{M-1}{\mathrm{\Σ}}}{\left|x_i\left(k\right)\right|}^2$ Known, denominator represent the gross energy of array element signals.

9. adaptive acoustic formation method according to claim 1, is characterized in that, the described result to described space point-by-point focusing is carried out all scan-line datas after adaptive weighted and carried out image conversion and specifically comprise:

Described scan-line data is carried out to envelope processing;

Scan-line data after envelope processing is carried out to interpolation;

According to the data acquisition scan image after interpolation.