CN106501367B - Implementation method of phased array ultrasound imaging based on elliptical arc scan conversion - Google Patents

Abstract

Phased-array ultrasonic echo-wave imaging method based on elliptic arc scan transformation belongs to quick ultrasonic echo technical field of imaging, it is characterized in that, utilize the scan transformation technology of elliptic arc, to it is existing when domain phased array in each transmission/reception array element group obtain echo data carry out elliptic arc draw operation, target reverberation is the intersection point of multiple elliptic arcs, it is superimposed its pixel value effectively, is higher than its surrounding pixel, obtains the position of target reverberation whereby.Successively as one imaging system as made of computer, phased array probe and A/D converter concatenation of construction, the acquisition of data in each secondary detection process, it is calculated using elliptic arc scan transformation technology and is constituted using sending/receiving array element as the elliptic arc on the vertical section of focus, the final fast and accurately imaging realized to the ultrasonic reflections signal of detected object.

Description

Implementation method of phased array ultrasound imaging based on elliptical arc scan conversion

technical field

The invention relates to ultrasonic nondestructive detection technology, ultrasonic imaging technology, phased array technology and ellipse scanning conversion technology, and realizes fast and accurate imaging of the surface shape and internal structure of an object.

Background technique

Ultrasonic phased array technology (Phased Array) is a new ultrasonic imaging technology in the field of ultrasonic non-destructive detection. Compared with a single transducer, the phased array is more flexible and more sensitive to the surface shape and internal defects of the object, so it has been widely studied. It has become a research hotspot in recent years and has been widely used in heavy industry, oil and gas pipelines, energy and aviation.

The basic idea of phased array technology is to use a pulse-echo (pulse-echo) measurement mechanism to use a group of ultrasonic phased array elements (ultrasonic transducers) arrays, one phased array element at a time to send to the detection object. The pulse scan signal is used, and the obtained pulse echo signal is focused and imaged by the time delay and superposition (DAS) method (time delay or phase delay). Full Matrix Capture (Full Matrix Capture) is a new data acquisition method for phased arrays. It can acquire time-domain data of all transmit/receive array elements, obtain higher image resolution, and detect internal defects in objects. characteristics for more accurate judgment. The working model of full matrix capture is shown in Figure 1(a). Each time, a phased array element is selected to transmit ultrasonic signals to the object in the direction of θ with the X-axis to detect the interior of the object. The phased array array receives the echo signal reflected from the interface of the object or the internal reflector and saves it by sampling, repeating the process of "one transmission, all reception" until all the elements of the phased array have sent the acoustic signal, and finally Full-matrix capture The acquired full-matrix data is post-processed and the image displayed. A variety of post-processing algorithms can be divided into two categories: time domain and frequency domain according to different processing methods:

Among the time-domain phased array ultrasound imaging algorithms, the most widely used and the best imaging effect is the total focusing algorithm. The algorithm flow is shown in Figure 1(b). In order to focus on any pixel (x, z) of the target image, The time-domain phased array technology performs delay and superposition processing on the echo signals received by all (transmitting, receiving) array elements in the phased array array: set is the sampled echo signal received by the phased array element E _j in the sampling process of the i-th step (that is, the echo signal corresponding to the sending array element/receiving array element group (E _i , E _j )), is the sampling number, the sampling delay of the phased array element E _j about the pixel point (x, z) is

Among them, v is the propagation speed of ultrasonic waves in the medium, and ri _,j (x,z) is the sum of the distances between the pixel point (x,z) and the phased array elements E _i and E _j . The corresponding sample number is _fs is the sampling frequency. All the delays within the effective length L of the phased array form a delay curve. The formula for calculating L is

L=0.84λz/d (2)

λ is the wavelength of ultrasound in the medium, d is the diameter of the phased array element, then the pixel value at (x, z) is

Among them, N is the number of elements in the phased array, ω _i,j (x,z) is the apodization function, which is calculated by the following formula

ω _{i ,j} (x,z)＝ωi (x,z)· _ωj (x,z)

in,

For the array element E _i (x _i ,0), at the depth z, the range of its effective aperture is For points outside the valid range, the apodization function value is 0, and its pixel value will not be affected by E _i . As shown in Figure 1(d), the red area represents the sound field range of the array element E ₀ , the green area represents the sound field range of the array element E ₁ , and the yellow area represents the common sound field range of the two. For the sending and receiving array elements (E ₀ , The sampling signal corresponding to E ₁ ) can only affect the pixels in the yellow area.

The frequency domain phased array algorithm is based on the wave equation inversion theory. Recently, some scholars have combined the Migration Technique in Reflection Seismology with the frequency-domain phased array technique to obtain the Phase Shift Migration method. The method takes the echo signals obtained at the phased array elements as the boundary conditions of the wave equation, and then uses the phase shift symbols in the frequency domain to perform hypothesis detection on the sound fields located at different depths, thereby extrapolating the entire sound field. The algorithm mainly includes two steps: the first step is to perform a two-dimensional Fourier transform on the time-domain echo signal data obtained by the phased array element to obtain a two-dimensional spectrum; the second step is to obtain the spectrum at the image boundary, and then to The two-dimensional spectrum calculated last time is phase-shifted, and then inverse Fourier transform is performed to obtain images of different depth values in turn.

It can be seen from the principle of time-domain phased array technology that in the imaging process, it is necessary to calculate the distance between all pixel points in the image and each point pair at the scanning position of the phased array array element. The algorithm complexity is high, and it involves the mean square The root operation makes the algorithm run for a long time. In the frequency domain phase migration technology, the forward and reverse Fourier transforms need to be performed on the data during the calculation process, and the calculation efficiency is not very high.

Therefore, imaging calculations related to phased arrays are time-consuming and have large imaging accuracy errors, which cannot meet the requirements of practical engineering applications for imaging speed and accuracy. Therefore, the phased array technology needs to further improve the imaging efficiency and imaging accuracy.

Analyzing the principle of time-domain phased array technology, it can be found that the imaging method is actually a reverse calculation process: it is necessary to first determine the pixel points on the image, and then, for all transmit/receive phased array element pairs, calculate The sum of the distances ri _,j from the pixel point to the two array elements and the transmission delay t _i,j , and the pixel value is obtained by accumulating the sampled signal value S _i,j (n _t ) corresponding to the delay. During the calculation process, the position of the reflector in the measured object is unknown, and the entire image is obtained by dynamically focusing and delaying accumulation of all the pixels in the image. The time-delayed sampling signals located at the pixel points of the target reflector are uniformly superimposed to achieve signal focusing. For the surrounding pixels, the accumulation of signals is disordered, so in the image, the accumulated pixel value corresponding to the reflector is significantly larger than other pixels.

The total focusing algorithm is similar to the synthetic aperture focusing algorithm, and is mainly based on the principle of delay superposition. The difference is that in the synthetic aperture focusing method, the transmitting array element and the receiving array element are the same array element, then for a phased array containing N array elements, the synthetic aperture focusing will obtain N groups of A-scan data. In order to improve the operation efficiency of the synthetic aperture focusing algorithm, some researchers proposed a synthetic aperture focusing algorithm based on arc scan conversion (DAB-SAFT). The process is reinterpreted from the positive direction, that is, the effect of all the obtained A-scan data on the target image is calculated, and the entire imaging process is converted into multiple arc scan conversion operations, which greatly improves the performance of synthetic aperture focusing. However, the full-matrix acquisition method obtains the corresponding A-scan signals of all transmitting and receiving array elements in the phased array, that is, for a phased array containing N array elements, there are N ² groups of A-scan signals, which are called full-matrix data. . For the case where the transmitting and receiving array elements are different, the algorithm cannot handle it, so it cannot be applied to the phased array ultrasonic echo imaging.

The phased array imaging method based on elliptical arc scan conversion has the following characteristics: (1) The algorithm can be applied to the situation where the transmit-receive array elements are different, and can also process the same transmit and receive array elements. DAB-SAFT only is a special case of this algorithm. (2) This algorithm converts the imaging process into multiple elliptical arc scan conversion operations by interpreting the total focusing process in a forward direction, removing the time-consuming root mean square operation, and reducing the complexity of the algorithm by using the sparseness of the echo data. Greatly improve imaging speed.

For each data sampled by each transmit/receive array element pair, it not only acts on the pixels corresponding to the reflectors, but also acts on the pixels without the reflectors. As shown in Fig. 1(c), corresponding to the sampled data whose sending array element is E _i , receiving array element is E _j and sampling delay is t _i,j Not only participates in the imaging calculation of point p(x,z), but also participates in the calculation of other points on the curve segment arc _i,j (r _i,j ) such as p'(x',z'), ri _,j = v·t _i,j /accuracy. Among them, arc _i,j (r _i,j ) is a curve located in the sound field range of the array elements E _i and E _j , and the pixel values of all points on the curve are affected by Impact. The points on the curve arc _i,j (r _i,j ) satisfy the following conditions: the sum of the distances ri _, j between the points on the curve and the matrix elements E _i and E _j is the same, and the delays are both t _i,j . So the curve segment arc _i,j (ri _,j ) is an elliptic arc with E _i and E _j as the foci and the major axis length is ri _,j .

On the other hand, if the overall calculation process is comprehensively understood from the forward direction, the efficacy of the sampling data S _i,j (n _t ) in the imaging calculation of the entire image is equivalent to its data value ω _i,j s _i,j (n _t )/r _i,j draws an ellipse arc arc _i,j (r _i,j ) in the image, that is, the pixel value of each pixel on the ellipse arc arc _i,j (r _i,j ) is accumulated ω _i,j s _i,j (n _t )/r _i,j . Therefore, the elliptic arc (arc ₀ ) corresponding to all the data (S _0,0 (n _t ) to S _N-1,N-1 (n _t )) on the delay curve where S _i,j (n _t ) is located _,0 (r _0,0 ) to arc _N-1,N-1 (r _N-1,N-1 )) in the image intersection point is the target pixel (x,z) in the original reverse calculation process. For the target reflector, multiple elliptical arcs will pass through the reflector, making its pixel value higher than the surrounding pixels to highlight the location of the reflector. As shown in Figure 1(d), when the sending array element is E ₀ , The target reflector (x,z) has arc _0,0 (r _0,0 ), arc _0,1 (r _0,1 ), arc _0,2 (r _0,2 ) and other elliptical arcs passing through this point, the point Pixel values are effectively superimposed.

If only the efficacy of all sampled data S _i,j of the array element E _j in the detection process of the i-th step is considered in the whole image, it corresponds to a sub-image I _i,j (as shown in Figure 1(e)), The image includes multiple elliptical arcs with E _i and E _j as the foci but with different major axes. The original DAS calculation formula (3) can be re-interpreted as the superposition of the corresponding sector diagrams at each scanning position, and the pixel value in the image is:

The result is the same as formula (3), where N is the total number of elements in the phased array.

In the original algorithm process, for any pixel point p(x,z) on the image, the contribution of the sampling data S _i, j received by the array element E _j to the pixel value of the point in the i-th detection process is: In order for the new algorithm to have the same result as the original algorithm, at all pixel points p(x,z) in the sub-image I _i,j , there is one and only one elliptical arc passing through this point. As shown in Figure 2, the elliptical arcs in the sub-image _It,r are all symmetrical about the straight line AM: x=x _c =(x _i +x _j )/2, where A(x _c ,0), M(x _c , Z _depth -1), all pixel points P _k (x _c , z _k ), z _k =0,1,...,Z _depth -1 on the line segment AM, all have an elliptical arc passing through. Therefore, the sub-image I _t,r includes a total of Z _depth elliptical arcs, each elliptical arc takes (x _i ,0) and (x _j ,0) as the focus, and passes through the point P _k (x _c ,z _k ),z _k = 0, 1, ..., Z _depth -1.

Therefore, the original time-domain phased array technology can be realized by successively drawing elliptical arcs on the image according to the positions of each sending/receiving array element pair in the order of the major axes ri _,j . In computer graphics, in order to display geometric elliptical arcs on raster display devices, a variety of mature elliptical arc scan conversion techniques have been developed. Among them, the most widely used is the midpoint ellipse drawing algorithm. Using this algorithm, the position of each pixel point on the ellipse arc can be quickly obtained, without calculating the distance between the imaging point and the position point of each array element, avoiding the root mean square operation, and can Save a lot of computational overhead.

After the cyclic superposition calculation of each transmitting and receiving transducer in the algorithm is completed, the value of each pixel is displayed; the entire image will have obvious high brightness values at the pixels of the internal defects and interface, and will form a continuous White boundary line; then through the edge detection and extraction (algorithm) of the image, the shape curve of the internal defect or interface of the tested object can be obtained.

SUMMARY OF THE INVENTION

The purpose of the present invention is to propose a method for realizing the time-domain phased array ultrasonic imaging technology based on elliptical arc scan conversion, so as to improve the imaging speed and imaging accuracy.

The invention is characterized in that the ellipse scan conversion technology is extended to the ultrasonic ellipse arc scan conversion technology, and the rasterization method of the ellipse arc segment is used to realize the time-domain phased array ultrasonic fast imaging technology without calculating the imaging point and each array element. The distance between the position points avoids the root mean square operation and saves a lot of calculation operations.

The time-domain phased array ultrasonic echo imaging method based on elliptical arc scan conversion of the present invention is characterized in that the following steps are included in sequence:

Step (1): construct a time-domain phased array ultrasonic echo imaging system based on elliptical arc scan conversion that can perform damage-free detection on the longitudinal section formed in the depth direction of the measured object, hereinafter referred to as the system, which includes a A computer, a phased array probe, a phased array controller, and an analog-to-digital converter consist of:

The input end of the phased array controller is connected to the output end of the computer control signal, the input end of the phased array probe is connected to the output end of the phased array controller, and the echo signal output end of the phased array is connected to the output end of the phased array controller. The input end of the analog-to-digital converter is connected, and the output end of the analog-to-digital converter is connected with the echo sampling signal input end of the computer,

The XZ coordinate system is set on the upper surface of the measured object. The horizontal length of the measured object along the X-axis direction is M _x , which is positive to the right, and the vertical axis Z is positive downward, as the target reflector. The depth on the longitudinal section of the measured object The coordinates of the direction are (x,z)

The phased array probe has a total of N phased array elements composed of ultrasonic transducers, hereinafter referred to as array elements. E _i represents, i=0,1,2,...,N-1, i is not only the array element number, but also the sequence number of each detection step in the detection process, the distance between two adjacent pixels in the image is called image accuracy, The coordinates of the first array element in the array are (x ₀ ,0), the coordinates of the E _i -th array element are (x _i =x ₀ +i*Δx/accuracy,0), the number of pixels in the horizontal direction of the target image X _length =M _x /accuracy,

Calculate the half-power beam angle of the ultrasonic transducer β _0.5 = 0.84λ/d, λ is the wavelength of the ultrasonic wave when it propagates in the measured object, d is the diameter of the ultrasonic transducer, it is the same for all array elements, at the depth When is z, the effective aperture length of the array element is: L _z =z*β _0.5 ;

Step (2): The system implements the described phased array ultrasonic echo imaging method based on elliptical arc scan conversion according to the following steps:

Step (2.1): Step i=0, detect as follows:

Step (2.1.1): The computer controls the phased array controller to generate a transistor-transistor logic level TTL pulse, triggers the first array element E _i =E ₀ , so that it is perpendicular to the X of the measured object. An excitation pulse is emitted in the depth direction Z of the shaft;

Step (2.1.2): All N array elements including E _i are switched to the receiving mode and start timing to receive the echo signal reflected from the measured object. The echo signals received by each array element E _j are sampled for a total of N _t times, the sampling sequence number is n _t , n _t =0,1,2,...,N _t -1, and the sampling frequency is f _s , f The value of _s is preset by the analog-to-digital converter, and S _i,j (n _t ) is the sampling value obtained by the sampling of the j-th array element E _j at the n- _th time during the detection process of the i=0-th step of the phased array. ; The total sampling times N _t of each array element is:

N _t =2Z _depth ·accuracy·f _s /v,

v is the propagation speed of ultrasonic waves in the measured object,

Z _depth is the number of pixels in the vertical direction of the generated target image, which is achieved by setting Z _depth ;

Step (2.2): the computer repeats the step (2.1.1)-step (2.1.2), and sequentially reads all N inputted by the analog/digital converter in the i=1, 2, . . . , N-1 step. The echo signal sampling value of each phased array element;

Step (3): For the detection process of the i=0th step, the echo sampling signal S _i,j (n _t ) received by the j=0th array element, n _t =0,1,...,N _t -1 , solve its corresponding sub-image _It,r , that is, the contribution value of the sampling signal to all the pixels on the target image,

Step (3.1): Calculate a series of elliptical arcs on the longitudinal section with the coordinate points (x _i , 0) and (x _j , 0) as the focus in the following steps:

Step (3.1.1): take the initial coordinate value z _α =0 in the depth direction, the coordinate of the center point of the ellipse is (x _c =(x _i +x _j )/2,0), and calculate the path E _i →p(x _c ,z _α )→E _j transmission distance:

Calculate the time it takes for the ultrasound to travel in the measured object with a distance ri _,j (x _c ,z _α )

t _i,j (x _c ,z _α )=r _i,j (x _c ,z _α )·accuracy/v,

and the sampling sequence number corresponding to this time n _t =t _i,j (x _c ,z _α )·f _s ;

Step (3.1.2): determine whether the sampling signal S _i,j (n _t ) is a non-zero value, if so, execute step (3.1.3), otherwise, skip step (3.1.3) and directly execute step (3.1. 4);

Step (3.1.3): The elliptic arc corresponding to the sampled signal S _i,j (n _t ) is arc _i,j (r _i,j (x _c ,z _α )), and the elliptic arc is represented by (x _i ,0 ) and (x _j , 0) are the focal points and pass through the point (x _c , z _α ), where the major semi-axis of the ellipse is a=ri _,j (x _c ,z _α )/2, and the minor semi-axis is b=0 , the ellipse arc equation is

f(x,z)=b ² (xx _c ) ² +a ² z ² -a ² b ² =0,

Call the midpoint ellipse algorithm to get the coordinates of all pixels on the ellipse arc. For all pixels p(x _k ,z _k ) that satisfy z≥0, k=0,1,2,...,n _k -1,n _k is the number of pixels that satisfy the condition, and the values of the pixels are accumulated Among them, the apodization function ω _i,j (x,z) is:

ω _{i ,j} (x,z)＝ωi (x,z)· _ωj (x,z)

Step (3.1.4): Take the coordinate values z _α =1,2,...,Z _depth -1 in the depth direction in turn, and repeat steps (3.1.1) to (3.1.3);

Step (4): take the receiving array elements j=1,2,...,N-1 in turn, and execute step (3);

Step (5): For the detection _{steps i} =1, 2, .

Step (6): Use the edge detection algorithm to extract the points where the pixel value changes significantly in the longitudinal section image, and these pixel points constitute the shape and position of the defect (interface) in the measured object.

Computational complexity analysis of the algorithm:

Assuming that the horizontal length of the detected object includes the number of pixels X _length , the vertical direction includes the number of pixels Z _depth , and the number of array elements in the phased array probe is N. Taking the calculation of one pixel point as an example, the original time-domain phased array algorithm needs to calculate the transmission distance of "transmitting array element → pixel point → receiving array element" for all the sending/receiving array element combinations in the phased array, and then calculate Get the signal transmission time, and find the strength value of the echo signal according to the time. Root mean square operation is required for each distance calculation. Let the calculation time of each root mean square operation be T _RMS , and the time to find the echo signal value each time is T _L , then the calculation time of the original algorithm is O(X _length *Z _depth *N ² *( _TL +T _RMS )). However, using this algorithm, only the non-zero signal received by the array element is required for operation each time, and the original distance operation is replaced by the elliptic arc scan conversion operation (ie, the root mean square operation is replaced by an addition operation). Assuming that the average number of non-zero signals received by each transmit-receive array tuple is _ns , there is only one RMS operation involved in rasterizing an elliptical arc each time (calculating the position of the initial point of the elliptical arc). If the average number of pixels on the elliptical arc is _ne , the addition running time of each pixel in rasterization during the elliptical arc scan conversion is T _A , so the time of each drawing operation of the ellipse arc is T _RMS + _ne *T _A , then the running time of this algorithm is O(N ² *n _s *(T _RMS +n _e *T _A )). Assuming that X _length and Z _depth belong to the same order of magnitude (represented by O(X _length )), N is a constant (usually N<<X _length ), and the farthest transmission distance of the signal is The nearest transmission distance is 0. To simplify the calculation, it is assumed that the average transmission distance is but Therefore, n _e and X _length and Z _depth belong to the same order of magnitude (represented by O(X _length )), and n _s is a constant (represented by O(1)). The complexity of the original algorithm is O(N ² *X _length ² ), and the complexity of this algorithm is O(N ² *X _length ). Experiments show that the calculation speed of this method is improved by an order of magnitude compared with the traditional time-domain phased array imaging method.

Analysis of results:

Compared with the prior art, the present invention has the advantages of high imaging speed and more accurate imaging. For the measured object in Figure 3(a), there are five equally spaced holes with a diameter of 1 mm. The original algorithm and the new algorithm are in It runs on a machine with Core ^TM i5-4210CPU@1.7GHz and 4GB RAM. Figure 3(b) is the running result of the original algorithm, and Figure 3(c) is the running result of the new algorithm. The high pixel value area (ie, the white area) in the two images indicates that there is a defect here. Comparing the positions of the low pixel area in 3(b) and 3(c), it is found that the results of the two images are exactly the same. Figure 3(d) is a grayscale comparison of the above result images. It can be seen that the two grayscale images are almost the same, indicating that the new algorithm has the same operating results as the original algorithm. The running time of the original algorithm is 10.423s, and the running time of the phased array imaging method based on elliptical arc scan conversion is 0.224s, which is 46 times better than the original algorithm.

Description of drawings

Figure 1 is the working model and schematic diagram of the time-domain phased array ultrasound imaging technology: 1(a) is the full matrix capture process; 1(b) is the schematic diagram of the reverse calculation process of the original time-domain phased array technology; 1(c) ) is the forward explanation diagram of the time-domain phased array technology; 1(d) is the elliptic arc passing through the target reflector in the detection process of step 0: the red area represents the sound field range of the array element E ₀ , and the green area Indicates the sound field range of the array element E1, the blue area represents the sound field range of the array element E2, the yellow area represents the public sound field range of E ₀ and E ₁ , and the purple area represents the public sound field range of E ₀ and E ₂ ; 1(e) is the i-th In the process of step detection, the efficiency diagram of all the data sampled by the array element E _j in the original time-domain phased array technology.

FIG. 2 is a schematic diagram of calculating the number of elliptical arcs contained in a sub-image.

Figure 3 is a schematic diagram of the experiment: 3(a) is a schematic diagram of the structure of the tested object; 3(b) is the result image of the original algorithm; 3(c) is the result image of the imaging method based on elliptical arc scanning; 3(d) is the image (b) ) and (c) grayscale histogram comparison.

FIG. 4 is a schematic flow chart of the ultrasound imaging system.

Figure 5 is a structural diagram of the ultrasound imaging hardware system.

Figure 6 is a schematic diagram of the operation of the ultrasonic array element.

FIG. 7 is an overall flow chart of a time-domain phased array ultrasound imaging method based on elliptical arc scan conversion.

Figure 8 is a pseudo-code of a time-domain phased array ultrasound imaging method based on elliptical arc scan conversion.

Detailed ways

The specific implementation process of the present invention includes three parts (as shown in FIG. 4 ): ultrasonic data acquisition, imaging calculation and image display. The hardware platform system structure diagram is shown in Figure 5. The ultrasonic imaging system consists of a computer, a phased array probe, a phased array controller and an analog-to-digital converter. The input end of the phased array controller It is connected with the output end of the computer control signal, the input end of the phased array probe is connected with the output end of the phased array controller, the echo signal output end of the phased array is connected with the input end of the analog-to-digital converter, and the The output end of the analog-to-digital converter is connected with the echo sampling signal input end of the computer.

The horizontal length of the measured object along the X-axis is M _x , the number of one-dimensional phased array elements in the phased array probe is N, and the array elements are E _i (i=0,1,2,,…,N- 1), the distance between adjacent array elements is Δx, the coordinate of the 0th element of the phased array is E ₀ (x ₀ =0,0), then the coordinate of the i-th array element E _i is ( _xi = x ₀ +i*Δx/accuracy,0), where accuracy is the image accuracy, and the pixel value in the horizontal direction of the object is X _length =M _x /accuracy. The detection process is divided into N steps. During the i-th (i=0,1,2,...,N-1) step detection process, the phased array controller generates a TTL (transistor-transistor logic level) pulse, which triggers the array element E _i in the phased array to move toward The measured object transmits an excitation pulse in the depth direction Z perpendicular to the X axis, and then all the array elements (E ₀ , E ₁ ,..., E _N-1 ) switch to the receiving mode and start timing to receive the reflected light from the measured object. The echo signal, the analog-to-digital _converter samples the echo signal received by each array element (E ₀ , _{E 1} , . In the computer, the sampling sequence number n _t =0,1,...,N _t -1, the sampling frequency is f _s , and its value is the analog-to-digital converter preset, and _si,j (n _t ) is the i-th step detection process , the sampling value of the echo signal received by the array element E _j at time t=n _t /f _s . Wherein N _t =2Z _depth ·accuracy·f _s /v, v is the propagation speed of ultrasound in the measured object, Z _depth is the pixel value in the vertical direction of the measured object, and its value is preset by the system.

The imaging calculation is to input the sampling data obtained by the full matrix capture method as computer input, and then calculate the longitudinal section image of the measured object according to the aforementioned imaging steps. The overall flow chart is shown in Figure 7.

In the imaging step (3.1), when calculating the sampling sequence number corresponding to the transmission delay _nt = f _s ·t = f _s · 2a/v, the obtained value _nt may be a decimal, and at this time, two adjacent to it are obtained. The sampling signals corresponding to the integers are interpolated to obtain the sampling signals. For example, assuming that the obtained sampling sequence number is n _t =30.4, and the sampling signals corresponding to the adjacent integers are s _i,j (30) and s _i,j (31), then the sampling signal corresponding to n _t is s _{i ,j} (30)+(s _i,j (31)-s _i,j (30))*(30.4-30).

In the experiment described in Figure 3, the number of array elements in the phased array is 32, the coordinate of the 0th array element is (0,0), the array element interval is 1.2mm, the image accuracy is accuracy=0.05mm, the array element E _i The coordinates are (i*24,0). The diameter of the array element is d = 1mm, the frequency of the array element is f _t = 3.00MHz, the half-power beam angle is β _0.5 = 0.84λ/d = 0.431, the propagation speed of the sound wave in water is v = 1540m/s, and the sampling frequency is f _s =100.00MHz. First, use the process in Figure 1(a) to obtain echo sampling data of all (transmitting, receiving) array element combinations, and then use the imaging algorithm in Figure 8 to process the sampling data to obtain the final result image of X _length ×Z _depth .

Image display is to display the calculated two-dimensional image on the corresponding display device. The resulting image is processed with an edge detection algorithm to obtain the location of the defect.

Claims (1)

1. based on the time-domain phased array ultrasonic echo imaging method of elliptic arc scanning conversion, it is characterized in that, realize according to the following steps successively: Step (1): Construct a time-domain phased array ultrasonic echo imaging system based on elliptical arc scan conversion that can perform damage-free detection on the longitudinal section formed in the depth direction of the measured object, hereinafter referred to as the system, which consists of a A computer, a phased array probe, a phased array controller, and an analog-to-digital converter consist of: The input end of the phased array controller is connected to the output end of the computer control signal, the input end of the phased array probe is connected to the output end of the phased array controller, and the echo signal output end of the phased array is connected to the output end of the phased array controller. The input end of the analog-to-digital converter is connected, and the output end of the analog-to-digital converter is connected with the echo sampling signal input end of the computer, The XZ coordinate system is set on the upper surface of the measured object. The horizontal length of the measured object along the X-axis direction is M _x , which is positive to the right, and the vertical axis Z is positive downward, as the target reflector. The depth on the longitudinal section of the measured object The coordinates of the direction are ( x , z ), The phased array probe has a total of N phased array elements composed of ultrasonic transducers, hereinafter referred to as array elements. It is represented by E _i , i=0,1,2,...,N-1 , i is not only the number of the array element, but also the serial number of each detection step in the detection process, the distance between two adjacent pixels in the image is called the image accuracy accuracy , the coordinates of the first array element in the array are ( x ₀ , 0), the coordinates of the E _i -th array element are ( x _i = x ₀ +i* Δ x/ accuracy , 0), the number of pixels in the horizontal direction of the target image Mesh X _length = M _x / accuracy , Calculate the half-power beam angle β _0.5 =0.84 λ/d of the ultrasonic transducer, λ is the wavelength of the ultrasonic wave when it propagates in the measured object, d is the diameter of the ultrasonic transducer, the same for all array elements, at the depth When is z , the effective aperture length of the array element is: L _z = z * β _0.5 ; Step (2): The system implements the described phased array ultrasonic echo imaging method based on elliptical arc scan conversion according to the following steps: Step (2.1): Step i=0, detect as follows: Step (2.1.1): the computer controls the phased array controller to generate a transistor-transistor logic level TTL pulse, triggers the first array element E _i =E ₀ , so that it is perpendicular to the X of the measured object. An excitation pulse is emitted in the depth direction Z of the shaft; Step (2.1.2): All N array elements including E _i are switched to receive mode and start timing to receive the echo signal reflected from the measured object. The echo signals received by each array element E _j are sampled for a total of N _t times, the sampling sequence number is n _t , n _t =0,1,2,….,N _t -1 , and the sampling frequency is f _s , f _s The value is preset by the analog-to-digital converter, and S _i,j ( n _t ) is the sampling value obtained by the sampling of the j -th array element E _j at the n - _th time during the detection process of the i=0-th step of the phased array; The total sampling times N _t of each array element is: N _t =2Z _depth ·accuracy·f _s /v, v is the propagation speed of ultrasonic waves in the measured object, Z _depth is the number of pixels in the vertical direction of the generated target image, which is achieved by setting Z _depth ; Step (2.2): the computer repeats the step (2.1.1)-step (2.1.2), and sequentially reads all N inputted by the analog/digital converter described in the i= 1,2,..., N -1th step The echo signal sampling value of each phased array element; Step (3): For the detection process of the i = 0th step, the echo sampling signal S _i,j ( n _t ) received by the j =0th array element, n _t = 0,1,...,N _t -1 , solve its corresponding sub-image It _,r , that is, the contribution value of the sampling signal to all the pixels on the target image, Step (3.1): Calculate a series of elliptical arcs on the longitudinal section with the coordinate points ( x _i , 0) and ( x _j , 0) as the focus in the following steps: Step (3.1.1): take the initial coordinate value z _α = 0 in the depth direction, the coordinate of the center point of the ellipse is ( x _c =( x _i + x _j )/2,0), and calculate the path E _i →p(x _c ,z _α )→E _j transmission distance: Calculate the time it takes for the ultrasound to travel the distance ri _,j ( x _c , z _α ) in the measured object t _i,j (x _c ,z _α )=r _i,j (x _c ,z _α )·accuracy/v, and the sampling sequence number corresponding to this time n _t = t _i,j ( x _c , z _α ) f _s ; Step (3.1.2): Determine whether the sampling signal S _i,j ( n _t ) is a non-zero value, if so, execute step (3.1.3), otherwise, skip step (3.1.3) and directly execute step (3.1. 4); Step (3.1.3): The elliptic arc corresponding to the sampled signal S _i,j ( n _t ) is arc _i,j ( r _i,j ( x _c , z _α )), and the elliptic arc is represented by ( x _i , 0 ) and ( x _j , 0) are the focus and pass through the point ( x _c , z _α ), where the major semi-axis of the ellipse is a = ri _,j ( x _c , z _α )/2, and the minor semi-axis is b = 0 , the ellipse arc equation is f(x,z)=b ² (xx _c ) ² +a ² z ² -a ² b ² =0, Call the midpoint ellipse algorithm to get the coordinates of all pixels on the ellipse arc, for all pixels p ( x _k , z _k ) satisfying z ≥ 0, k =0,1,2,…, n _k -1, n _k is the number of pixels that satisfy the condition, and the values of the pixels are accumulated Among them, the apodization function ω _i,j (x,z) is: ω _{i ,j} (x,z)＝ωi (x,z)· _ωj (x,z) Step (3.1.4): Take the coordinate values z _α =1, 2, ..., Z _depth -1 in the depth direction in turn, and repeat steps (3.1.1) to (3.1.3); Step (4): take the receiving array elements j=1,2,...,N -1 in turn, and execute step (3); Step (5): For the detection _{steps i} =1, 2, . Step (6): Use the edge detection algorithm to extract points in the longitudinal section image where the pixel value changes significantly, and these pixel points constitute the shape and position of the interface in the measured object.