Optimized delay multiplication method suitable for ultrasonic full-focusing imaging

Abstract

The invention discloses an optimized delay multiplication method suitable for ultrasonic full-focusing imaging. The method comprises the following steps: step 1, obtaining an ultrasonic array signal; step 2, calculating Hilbert transform of the ultrasonic time domain signal; step 3, dividing an ultrasonic array imaging area pixel grid, and calculating distances from all array elements to imaging pixel points; 4, calculating the arrival time of ultrasonic waves according to the obtained distances from the array elements to the pixel points, and extracting the Hilbert transform values corresponding to all the time domain signals in the step 2; 5, calculating a signed square root of the signal amplitude, and carrying out improved delay multiplication and focusing calculation; and step 6, carrying out modulus operation on the delay multiplication calculation result of the previous step to obtain a final imaging result, improving the traditional delay multiplication calculation method in the aspect of mathematical expression, greatly reducing the calculation amount of the delay multiplication, improving the operation efficiency, and helping ultrasonic nondestructive testing to obtain the imaging result with higher quality.

Description

Technical Field

[0001] The present invention is an optimized time delay multiplication method suitable for ultrasonic full-focus imaging, and in particular relates to an adaptive full-focus imaging algorithm based on spatial coherence for ultrasonic imaging. The method greatly reduces the amount of data calculation, can realize ultrasonic fast high-resolution imaging, and belongs to the field of ultrasonic non-destructive testing. Background Art

[0002] Ultrasonic phased array imaging technology controls the transmission and reception combination of multiple array elements, and uses beamforming methods to achieve ultrasonic focusing, deflection and other effects in the ultrasonic transmission stage or post-processing stage to obtain high-quality ultrasonic images. Synthetic aperture focusing and total focusing are two widely used methods for focusing imaging in post-processing.

[0003] The delay-and-addition (DAS) beamforming method is the final focusing algorithm discovered. By calculating the distance from the imaging pixel to the array element, the corresponding amplitude signal is extracted for cumulative summation calculation to obtain the imaging result. Synthetic aperture focusing and full focusing imaging based on DAS have been widely used in fields such as medicine and non-destructive testing. However, DAS is a non-adaptive method. The algorithm shows limited imaging resolution and low off-axis anti-interference ability. In recent years, adaptive weighting methods based on coherence have been proposed one after another, such as coherence factor (CF) weighting, sign coherence factor (SCF) weighting, phase coherence factor (PCF) weighting, circular coherence factor (CCF) weighting, etc. Researchers have proved that such methods can effectively improve the lateral resolution or contrast of DAS beamforming. Although coherence-based weighting is conducive to improving imaging quality, it comes at the expense of computational efficiency. Imaging using the phase information of the signal itself can ensure computational efficiency, and its imaging quality is between weighted and unweighted. Fourier domain (FD) beamforming performs block processing on the two-dimensional Fourier spectrum in the frequency domain, which can achieve higher computational efficiency, such as the phase shift algorithm and the ω-k algorithm. Some results show that the computational load of FD beamforming is an order of magnitude lower than that of classical DAS. Although the computational load is lower, the image quality of FD beamforming details, such as the original speckle pattern, may be erased or distorted by filtering. The lateral resolution of frequency domain algorithms can be improved by zero padding in the Fourier transform, which is similar to smoothing and may lose image details. Another way to improve resolution is to minimize the sampling interval, but this also increases the amount of data and sampling time.

[0004] Delayed multiplication and sum (DMAS) beamforming uses the spatial coherence of defect signals to improve imaging quality. This technology was originally used in microwave imaging. In the original algorithm, the signals were combined, coupled and multiplied before summing. This operation can be interpreted as the autocorrelation function of the aperture. This means that in each calculation, spatial coherence operations are performed between all received signals collected by the array elements. Therefore, DMAS is a nonlinear beamforming algorithm. The imaging quality of DMAS is better than that of DAS, but a large number of multiplication operations are used before the addition operation, which greatly reduces the computational efficiency. Preprocessing the signal into a signed square root and then using the DMAS operation can correctly scale the amplitude and reduce a certain amount of calculation, but it is still difficult to achieve real-time imaging. When the number of array elements is N, full-focus DMAS requires N 2 (N 2 -1) / 2 multiplications and N 2 (N 2 -1) / 2 additions. Therefore, an efficient DMAS calculation method is currently needed to reduce the amount of calculation and obtain high-quality ultrasound imaging results. Summary of the invention

[0005] In order to improve the calculation efficiency of delay multiplication and sum and obtain high-quality ultrasonic imaging pictures, the present invention optimizes the mathematical expression of delay multiplication and sum, and proposes an improved delay multiplication and sum calculation method by combining Hilbert transform and signed square root operation. The feasibility of the proposed method is verified by the experimental results of ultrasonic phased array full focusing.

[0006] The specific steps are as follows:

[0007] Step 1: Obtain ultrasound array signals, especially array full matrix acquisition signals suitable for ultrasound full focus imaging. For the case where the number of array elements is N, full focus imaging requires N 2 Ultrasonic time domain signal.

[0008] Step 2: Calculate the Hilbert transform of the array signal.

[0009] Step 3: Divide the imaging area into pixel grids and calculate the distances from all array elements to the imaging pixels.

[0010] Step 4: Calculate the ultrasonic arrival time based on the distance from the array element to the pixel point, and extract the Hilbert transformed value corresponding to all array signals in step 2. For each pixel point, the number of signal amplitudes extracted by full focusing is N 2 .

[0011] Step 5: Obtain the calculation result of each pixel based on the improved delay multiplication and calculation formula.

[0012] Step 6: Modulo the delay multiplication and calculation result of the previous step to obtain the final imaging result.

[0013] In step 2, the Hilbert transform of the array signal is defined as:

[0014]

[0015] Where: x(t) is the original ultrasonic signal, is the Hilbert transform of the signal.

[0016] The improved delay multiplication and calculation formula of step 5 is characterized in that: for full focusing, the improved delay multiplication and calculation formula is:

[0017]

[0018] in:

[0019]

[0020] in: is the signal amplitude extracted after delay.

[0021] The present invention can achieve the following beneficial effects:

[0022] 1. When the number of array elements is N, the amount of calculation of the total focusing delay multiplication and sum is reduced from N 2 (N 2 -1) / 2 multiplications and N 2 (N 2 -1) / 2 additions, down to N 2 +1 multiplication and 2N 2 Addition.

[0023] 2. Introduce Hilbert transform and use complex signals for focusing operations. The signal spectrum changes from double-sided spectrum to single-sided spectrum, avoiding spectrum waste and obtaining better imaging results.

[0024] 3. The signed square root operation of the signal can correctly scale the amplitude. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 The figure is a flowchart of the optimized delay multiplication and summing process of ultrasonic full focusing imaging of the present invention.

[0026] Figure 2(a) shows the full-focus imaging result of the ultrasonic phased array time-delay superposition B-type ultrasonic test block.

[0027] Figure 2(b) shows the results of the optimized delay multiplication and full-focus imaging of the B-type ultrasonic test block using the ultrasonic phased array.

[0028] Figure 3(a) shows the result of ultrasonic phased array time-delay superposition crack full-focus imaging.

[0029] Figure 3(b) shows the results of ultrasonic phased array optimized delay multiplication and crack full-focus imaging. DETAILED DESCRIPTION

[0030] The present invention will be further described below in conjunction with the accompanying drawings and embodiments. The following embodiments are only illustrative and not restrictive, and the protection scope of the present invention cannot be limited thereto.

[0031] The present invention provides an optimized delay multiplication and sum method suitable for ultrasonic full-focus imaging, which modifies the calculation expression of the delay multiplication and sum mathematical formula. The modified calculation formula is equivalent to the original formula and reduces the amount of calculation. In addition, the signal is pre-processed with Hilbert transform and signed square root operation to help improve the imaging quality. The specific implementation steps of the method include (such as Figure 1 ):

[0032] 1. Use ultrasonic excitation detection to obtain array signals. Ultrasonic excitation detection methods include but are not limited to piezoelectric ultrasound, laser ultrasound, electromagnetic ultrasound, and ultrasonic phased array. Determine the number of array elements N, sampling frequency f, ultrasonic velocity C, aperture size, and imaging area. Select the focusing mode as full focusing. For the case where the number of array elements is N, full focusing imaging obtains N 2 Ultrasonic time domain signal x(t).

[0033] 2. Take the Hilbert transform of the time domain signal, as shown in equation (1).

[0034] 3. Divide the imaging area into pixel grids and calculate the distance from all array elements to the imaging pixel points. For the pixel coordinates r(x,z) in the imaging area, the distance from the pixel point to the excitation array element e(x0,0) is:

[0035]

[0036] The distance from the pixel to the receiving element r(x0+dx,0) is:

[0037]

[0038] 4. Extract the Hilbert transformed values ​​corresponding to all array signals, and calculate the focused time domain signal position corresponding to the pixel point using the ultrasonic modal velocity C according to the ultrasonic sampling frequency f:

[0039]

[0040] Extract the Hilbert transform amplitude at the corresponding position For each pixel, the number of signal amplitudes extracted by full focus is N 2 .

[0041] In full-focus imaging, dx is the distance from the excitation element to the receiving element.

[0042] 5. The original DMAS calculation formula for full focusing is:

[0043]

[0044] in:

[0045]

[0046] This calculation requires N 2 (N 2 -1) / 2 multiplications and N 2 (N 2 -1) / 2 additions. The above formula (7) is improved as follows:

[0047]

[0048] in:

[0049]

[0050] Therefore, the improved total focusing DMAS expression is:

[0051]

[0052] Substituting the amplitude after Hilbert transformation:

[0053]

[0054] 6. Take the modulus of the calculated result to obtain the final presentation result:

[0055] I(r)=|S DMAS (r)| (13)

[0056] In summary, in order to solve the problem of large amount of calculation in the original DMAS method, this patent modifies the mathematical expression, and the modified formula is equivalent to the original formula, and the amount of calculation is reduced. The Hilbert transform and the signed square root operation help to correctly scale the amplitude and avoid spectrum waste.

[0057] Experimental verification

[0058] The following will be combined with the implementation case 1 of the present invention to clearly and completely analyze and describe the technical solution of the present invention. The experimental implementation process includes the following steps:

[0059] Experimental Case 1

[0060] (1) Experimental process: An ultrasonic phased array device was used to obtain experimental data. The device is Phaselink, a phased array integrated advanced ultrasonic system produced by Eintik. The sampling frequency of the device is 200MHz. A 64-element linear array longitudinal wave probe is used. The probe chip spacing is 0.6mm and the center frequency is 10MHz. The experimental test objects are type B ultrasonic phased array test blocks (NB / T47013.15-2021) and test blocks with different angles of crack defects manufactured by aluminum alloy powder molten bed additive manufacturing, including cracks with a length of 3mm and angles of 30°, 50° and 70°. Full matrix acquisition is implemented during full-focus imaging to obtain 4096 ultrasonic time domain signals, and the signals are averaged once. The imaging program is written in MATLAB R2021a software, and the computer is equipped with an Intel i912900CPU processor and an NvidiaRTX 3060 graphics card.

[0061] (2) Experimental results: Figures 2 and 3 show the imaging results of the B-type ultrasonic phased array test block. Figure 2(a) shows the ultrasonic phased array time-delay superposition full-focus imaging result. Compared with the ultrasonic phased array optimized time-delay multiplication and full-focus imaging result in Figure 2(b), within the same imaging amplitude range of 0 to -40 dB, the result of the optimized time-delay multiplication and has significantly higher contrast and resolution. Better imaging results can also be obtained by increasing the number of signal averaging times. When targeting crack-type defects, as shown in Figures 3(a) and 3(b), the optimized time-delay multiplication and still shows a high resolution, proving that this method is suitable for imaging a variety of different types of defects. The pixel value of all imaging results is 350×320. The calculation time of Figure 2(a) is 2.77 seconds, the calculation time of Figure 2(b) is 3.93 seconds, the calculation time of Figure 3(a) is 2.60 seconds, and the calculation time of Figure 3(b) is 3.51 seconds. This shows that although the calculation time of the optimized delay multiplication and addition is higher than that of the delay superposition, they are at the same order of magnitude, and the difference is less than 1 / 2, which proves that the optimized delay multiplication and addition has high computational efficiency.

[0062] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. An optimized delay multiplication method suitable for ultrasonic full-focus imaging, characterized in that: The steps include: Step 1: Obtain ultrasound array signals; based on the full matrix acquisition signal of the array of ultrasound full focusing imaging, for the case where the array elements are N, full focusing imaging requires N 2 Ultrasonic time domain signal; Step 2: Calculate the Hilbert transform of the ultrasonic time domain signal; Step 3: Divide the ultrasound array imaging area into pixel grids and calculate the distances from all array elements to the imaging pixels; Step 4: Calculate the ultrasonic arrival time based on the distance from the array element to the pixel point, and extract the Hilbert transformed value corresponding to all time domain signals in step 2; for each pixel point, the number of signal amplitudes extracted by full focusing is N 2 ; Step 5: Calculate the signed square root of the signal amplitude and perform an improved delay multiplication and focusing calculation. The improved delay multiplication and calculation method for each pixel is the square of the signed square root sum minus the square sum divided by 2. Perform the above operation for all pixels. Step 6: Modulo the delay multiplication and calculation result of the previous step to obtain the final imaging result.

2. The optimized delay multiplication method for ultrasonic total focusing imaging according to claim 1, characterized in that: The specific steps of obtaining the ultrasonic array signal in step 1 are as follows: a. When full-focus imaging is required, N array elements are used. Each time an array element is excited, one array element emits an ultrasonic wave, and all N array elements receive the ultrasonic wave, obtaining N ultrasonic time domain signals at a time; N array elements are excited in sequence to obtain N 2 Ultrasonic time domain signal.

3. The optimized delay multiplication method for ultrasonic total focusing imaging according to claim 2, characterized in that: The Hilbert transform described in step 2 has the following specific steps: a. For an ultrasonic time domain signal x(t), its Hilbert transform H[x(t)] is: Where t is time and τ is the integration variable.

4. The optimized delay multiplication method for ultrasonic total focusing imaging according to claim 3, characterized in that: The specific steps of dividing the imaging area into pixel grids and calculating the distances from all array elements to the imaging pixels as described in step 3 are as follows: a. For the pixel coordinates r(x,z) in the imaging area, the distance from the imaging pixel to the excitation array element e(x0,0) is: b. The distance from the pixel to the receiving element r(x0+dx,0) is: c. In full-focus imaging, dx is the distance from the excitation element to the receiving element.

5. According to the optimized delay multiplication and sum method for ultrasonic total focusing imaging according to claim 4, step 4 extracts the Hilbert transformed values ​​corresponding to all array signals, and the specific steps are as follows: a. According to the ultrasonic sampling frequency f, use the ultrasonic modal velocity C to calculate the focused time domain signal position L corresponding to the pixel point: b. Extract the amplitude of the corresponding position For each pixel, the number of signal amplitudes extracted by full focus is N 2 6. According to the optimized delay multiplication and sum method suitable for ultrasonic total focusing imaging according to claim 5, the improved delay multiplication and sum calculation formula in step 5, wherein the specific steps are as follows: For full focusing, the improved delay multiplication and calculation formula S DMAS (r) is: For the above expression: in, sgn() represents symbolic operation.

7. The optimized delay multiplication method for ultrasonic total focusing imaging according to claim 6, characterized in that: The delayed multiplication and calculation result of step 6 is modulo, and the specific steps are as follows: The final imaging result I(r) is calculated as: I(r)=|S DMAS (r)| (7)。

Images