A method for terahertz continuous wave extended phase tomography in electronics

Abstract

The present invention relates to an electronic terahertz continuous-wave extended phase tomography method. The method includes terahertz extended phase tomography data acquisition 1 and terahertz extended phase tomography image reconstruction 2. A terahertz extended phase data acquisition device 3 acquires the unextended phase, and a terahertz extended phase extraction method 4 extracts the extended phase. Extended phase data preprocessing 5 then implements projection offset correction and Gaussian smoothing noise reduction to generate Gaussian model-corrected extended phase data. Extended phase image reconstruction 6 is then performed based on this data. This method is highly sensitive to phase changes caused by the interaction between the terahertz beam and the sample, and denoises and corrects the projection data, enabling high-definition imaging with superior imaging performance compared to amplitude imaging.

Description

Technical Field

[0001] The present invention belongs to the technical field of terahertz imaging, and in particular relates to an electronic terahertz continuous wave extended phase tomography method. Background Art

[0002] The principle of continuous terahertz wave tomography technology is as follows: a continuous wave source generates terahertz radiation with an intensity higher than the pulse radiation energy, and allows it to penetrate the sample to be tested, detecting the intensity information of the transmitted terahertz wave. Since the losses, irregular defects and edge effects inside the object have a scattering effect on the terahertz wave, thereby affecting the intensity distribution of the terahertz wave, after obtaining the intensity distribution of the transmitted terahertz wave, it is linearly converted into a grayscale image. The light and dark distribution of the grayscale image can be used to judge and identify the losses, irregular defects and edges inside the object.

[0003] Unlike X-rays, the interaction between a terahertz beam and a sample is characterized by its limited width, resulting in both reflection and refraction at the sample's edge. Furthermore, the beam will refract within the sample, and multiple reflections may occur if defects exist within the sample. For transmission tomography systems, if the amplitude of the detected terahertz beam is used for imaging, the reflected signal from the beam cannot be detected, and this energy is mistakenly interpreted as loss from the sample, introducing significant reflection errors. However, if the phase of the detected terahertz beam is used for imaging, reflection errors can be avoided because the sample's reflection of the beam does not affect the phase change after the beam passes through the sample.

[0004] Continuous terahertz wave phase tomography technology uses the phase information of the terahertz beam passing through the sample to achieve image reconstruction. Since low-refractive-index samples refract the terahertz beam less, the introduced phase error is smaller and the influence of reflection error can be avoided.

[0005] Currently, reported terahertz tomography data acquisition methods have the following problems: 1) The detected terahertz wave signal is an amplitude signal, which is easily affected by reflection and diffraction, such as the "Three-dimensional terahertz tomography system and scanning and image reconstruction method" with patent number CN 107631995A; 2) The influence of refraction on the imaging effect is reduced by using a built-in refractive index matching liquid, but different matching liquids need to be replaced for samples with different refractive indices, and different separation membranes need to be attached to samples with different shapes, such as the "Terahertz wave computer-assisted tomography device and method" with patent number CN108458987A.

[0006] Currently, reported terahertz tomography image reconstruction methods have the following problems: 1) Image reconstruction is achieved using the classic filtered back-projection algorithm, but simple filtering processing cannot effectively remove artifacts in the sinusoidal pattern and cannot enhance image clarity. For example, Patent No. CN112666814A, "A method for off-axis digital holographic diffraction tomography based on continuous terahertz waves"; 2) Coherent tomography technology is used to extract interference intensity information to obtain the three-dimensional structure of the object being measured, but the designed system components are numerous, requiring multiple sets of reflective mirrors, lenses, and beam splitters, making the test system more complicated. For example, Patent No. CN 111157486A, "A three-dimensional terahertz imaging method and imaging system thereof" Summary of the Invention

[0007] In response to the problems existing in existing terahertz tomography data acquisition methods and image reconstruction methods, the present invention proposes an electronic terahertz continuous wave extended phase tomography method, which includes terahertz extended phase tomography data acquisition 1 and terahertz extended phase tomography image reconstruction 2. The unexpanded phase is acquired by a terahertz extended phase data acquisition device 3, and the extended phase is extracted by a terahertz extended phase extraction method 4. The extended phase data is then preprocessed by 5 to achieve projection offset correction and Gaussian smoothing noise reduction, thereby forming extended phase data corrected based on a Gaussian model, and extended phase image reconstruction 6 is achieved based on the data.

[0008] The present invention addresses the problems existing in existing terahertz tomography data acquisition methods and image reconstruction methods, and adopts 3 and 4 to acquire extended phase data.

[0009] Among them, 3 includes an electronic terahertz continuous wave radiation source, a terahertz radiation antenna, a first terahertz beam reflection mirror, a terahertz load three-dimensional rotation displacement console, a second terahertz beam reflection mirror, a terahertz receiving antenna, a phase shift extraction circuit, and a computer. The electronic terahertz continuous wave radiation source is used to generate a continuous terahertz wave signal, which is radiated by the terahertz radiation antenna and then focused by the first terahertz beam reflection mirror to form a Gaussian beam. The Gaussian beam passes through a sample placed on the terahertz load three-dimensional rotation displacement console, is reflected by the second terahertz beam reflection mirror, and is focused on the terahertz receiving antenna, realizing the interaction process between the terahertz beam and the sample. The phase shift extraction circuit then extracts the extended phase value of the interaction between the terahertz beam and the sample and stores it in the computer memory.

[0010] Preferably, the first terahertz beam reflecting mirror and the second terahertz beam reflecting mirror are ellipsoidal reflecting mirrors, and the terahertz object unit rotation displacement control console includes two linear stepping motors and one rotary stepping motor.

[0011] Among them, 4 is the extended phase extraction technology, which uses a mixing circuit to mix the terahertz wave signal after interacting with the sample and the reference signal, and performs analog-to-digital conversion. The real and imaginary parts of the above mixing signal are extracted through digital mixing technology, and the unexpanded phase is obtained by numerical calculation, and the extended phase is obtained through phase expansion technology.

[0012] The present invention addresses the problems existing in existing terahertz tomography data acquisition methods and image reconstruction methods, and adopts 5 and 6 to process extended phase data.

[0013] Step 5 includes projection offset correction and Gaussian smoothing noise reduction. Projection offset correction is used to correct the projection misalignment of the terahertz extended phase data collected in steps 3 and 4, thereby obtaining high-precision projection data. Gaussian smoothing noise reduction is used to perform Gaussian beam model correction on the terahertz extended phase data collected in steps 3 and 4, so that it conforms to the projection data collection method under the ray model, thereby forming terahertz extended phase data based on the Gaussian model.

[0014] 6 includes filtered back-projection image reconstruction and iterative image reconstruction. The filtered back-projection image reconstruction performs nearest neighbor interpolation back-projection on the terahertz extended phase data pre-processed by 5. The iterative image reconstruction performs a finite number of iterations on the terahertz extended phase data pre-processed by 5 to achieve image reconstruction. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 This is a block diagram of the electronic terahertz continuous wave extended phase tomography method proposed by the present invention.

[0016] Figure 2 This is a schematic diagram of a terahertz extended phase data acquisition device proposed in the present invention.

[0017] Figure 3 This is a flow chart of the terahertz extended phase extraction method proposed in the present invention.

[0018] Figure 4 This is a flowchart of the terahertz extended phase data preprocessing proposed by the present invention.

[0019] Figure 5 This is a flow chart of the Gaussian model image reconstruction algorithm proposed by the present invention.

[0020] Figure 6 This is an expanded phase imaging result diagram of an embodiment proposed in the present invention. DETAILED DESCRIPTION

[0021] In order to more specifically and clearly express the technical solutions, advantages and creativity of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0022] The present invention relates to an electronic terahertz continuous wave extended phase tomography method, and the method provided in the embodiment is as follows Figure 1 As shown, the method comprises terahertz extended phase tomography data acquisition 1 and terahertz extended phase tomography image reconstruction 2. The unextended phase is acquired by a terahertz extended phase data acquisition device 3, and the extended phase is extracted by a terahertz extended phase extraction method 4. Projection offset correction and Gaussian smoothing noise reduction are then performed by extended phase data preprocessing 5 to generate extended phase data corrected by a Gaussian model. Extended phase image reconstruction 6 is then performed based on this data.

[0023] In the embodiment, Figure 2 The basic composition of 3 is shown as follows. A 110GHz vector network analyzer 7 is used to generate a continuous wave terahertz signal. The terahertz beam is radiated by the terahertz transmitting antenna 8. The terahertz wave is reflected and focused by the first ellipsoidal reflector 9 to form a Gaussian beam. The terahertz beam passes through the sample 11 placed on the terahertz carrier three-dimensional rotation displacement console 10. The second ellipsoidal reflector 12 reflects and focuses the transmitted terahertz beam. The terahertz transmitted wave signal is received by the terahertz receiving antenna 13 and sent to 7 for data calculation and display. 360° scanning is achieved under the control of the computer 14 and the terahertz carrier three-dimensional rotation displacement console controller 15, and the data is finally stored in the memory 14.

[0024] In the embodiment, the method described in 4 is implemented using phase expansion technology, such as Figure 3 As shown, it is divided into two processes:

[0025] S1: The real and imaginary parts of the terahertz beam after the interaction between the terahertz beam and the sample are recorded by the terahertz tomography extended phase data acquisition device, which are denoted as real and imaginary. The unexpanded phase is

[0026]

[0027] Since both real and imaginary are normalized, the phase wrapped ∈(-π,π), so the phase value exceeding (-π,π) will subtract the period and fall within this interval, forming an unextended phase.

[0028] S2: The unexpanded phase obtained in S1 is unwrapped by the phase expansion algorithm to recover the total change in the terahertz beam phase after the terahertz beam interacts with the sample. The specific implementation process is as follows:

[0029] Step 1: Traverse the unextended phase array. For a two-dimensional array, this is divided into row loop and column loop.

[0030] Step 2: Based on the row and column characteristics of the sinusoidal graph, take the difference between two adjacent values ​​in each column and determine whether the difference is greater than an integer multiple of π. If so, record the row number at that time, such as pos(i). If not, skip the next data point.

[0031] Step 3: Repeat step 2 until the entire unextended phase array is traversed.

[0032] Step 4: According to the row number obtained in step 2, such as pos(i), calculate the phase difference required for expansion, such as

[0033] Δphase=phase(pos(i)+1)-phase(pos(i))

[0034] Step 4: According to the row number obtained in step 2, such as pos(i), expand the phase of j∈(pos(i),pos(i+1)), such as

[0035] phase unwrapped =phase(j)-Δphase

[0036] Finally, the extended phase is obtained.

[0037] In the embodiment, the processing of step 5 is for the extended phase data obtained in step 4, such as Figure 4 As shown, it is divided into 2 processes:

[0038] S1: Projection offset correction, divided into the following steps:

[0039] Step 1: Determine whether the extended phase data has projection offset. If yes, perform offset correction; if no, continue with subsequent processing.

[0040] Step 2: If there is a projection offset, first locate the projection target eigenvalues ​​of the three adjacent columns in the unwrapped phase matrix, and record them as peak(i-1), peak(i) and peak(i+1).

[0041] Step 3: Calculate the projection offset value

[0042]

[0043] Step 4: Correcting the Projection Offset

[0044] projection(i)=projection(i)-offset(i)

[0045] S2: Gaussian smoothing filter, divided into the following steps:

[0046] Step 1: Determine whether there are stripe artifacts in the sinogram. If so, perform Gaussian smoothing filtering; if not, continue with subsequent processing.

[0047] Step 2: If stripe artifacts exist, design a Gaussian filter function, such as

[0048]

[0049] Step 3: Perform Gaussian filtering on the unwrapped phase matrix.

[0050] In the embodiment, step 6 includes performing a filtered back projection operation on the extended phase projection value processed by step 5 based on a filtered back projection algorithm to achieve image reconstruction.

[0051] In the embodiment, step 6 includes performing an iterative operation on the extended phase value processed in step 5 based on an iterative algorithm to achieve image reconstruction.

[0052] Existing imaging methods consider the terahertz beam as a ray model, without considering the reflection and diffraction effects of the terahertz beam, which will cause the reconstructed image to be distorted due to reflection and diffraction; at the same time, since the Gaussian beam has a certain width, this width will cause the interval of the projection data characteristic value to become wider, thereby widening the sinusoidal characteristic curve, and ultimately causing the reconstructed image edge to be blurred and distorted. In response to the above problems, the present invention proposes an image reconstruction method based on the Gaussian beam model, such as Figure 5 As shown, the reconstruction method of the present invention is divided into the following two processes:

[0053] S1: Creates a Gaussian convolution kernel with adaptive width based on the Gaussian beam model. The width is determined by the waist size of the Gaussian beam. The basic parameters of the Gaussian beam are as follows:

[0054]

[0055]

[0056] Where E is the Gaussian beam electric field, w(z) is the beam radius, and the minimum beam radius is the beam waist, denoted as w0. The processed terahertz extended phase projection matrix is ​​Fourier transformed to obtain the Fourier extended phase projection matrix. This matrix is ​​then deconvolved with an adaptive width Gaussian convolution kernel to form an adaptive de-Gaussian extended phase projection matrix.

[0057] S2: Image reconstruction of the adaptive de-Gaussian extended phase projection matrix is ​​achieved by filtered back projection algorithm and iterative algorithm.

[0058] In the embodiment, the present invention performs extended phase imaging on a polystyrene foam cylinder with a diameter of 70 mm and an alumina cylinder with a diameter of 10 mm. Figure 6Figure 6 (A) shows the sample structure. Figure 6 (B) shows the amplitude imaging result. In the reconstructed image, the edges of the polystyrene foam cylinder are difficult to distinguish, and the edges of the aluminum oxide cylinder are blurred. Figure 6 (C) shows the extended phase imaging result. In the reconstructed image, the edges of the polystyrene foam cylinder are clearly distinguishable, and the edges of the aluminum oxide cylinder are clear and consistent with the sample. The reconstruction method of the present invention not only effectively eliminates distortion in the reconstructed image, but also effectively eliminates the blurring of the image edges, resulting in a clearer image outline, demonstrating superior performance compared to amplitude imaging.

[0059] Those skilled in the art will readily understand that the foregoing is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, replacements, and improvements to the concepts and principles of the present invention shall fall within the scope of protection of the present invention.

Claims

1. An electronic terahertz continuous wave extended phase tomography image reconstruction method, characterized by: The invention comprises terahertz extended phase tomography data acquisition and terahertz extended phase tomography image reconstruction; the terahertz extended phase tomography data acquisition collects the unexpanded phase data of the measured terahertz wave after the interaction between the terahertz Gaussian beam and the sample by a terahertz extended phase data acquisition device, and further extracts the unexpanded phase data of the measured terahertz wave by a terahertz extended phase extraction method; the terahertz extended phase tomography image reconstruction performs extended phase data preprocessing on the obtained terahertz extended phase data of the measured terahertz wave, and further reconstructs the image using the preprocessed terahertz extended phase data of the measured terahertz wave; The terahertz extended phase extraction method uses a mixing circuit to mix the terahertz wave signal after interaction with the sample and the reference signal, and performs analog-to-digital conversion. The real and imaginary parts of the mixed signal are extracted through digital mixing technology, and the unextended phase is obtained by numerical calculation, and the extended phase is obtained through phase expansion technology. The terahertz extended phase data preprocessing includes projection offset correction processing and Gaussian smoothing noise reduction processing, wherein the projection offset correction is used to correct the projection misalignment of the collected terahertz extended phase data, thereby obtaining high-precision projection data, and the Gaussian smoothing noise reduction is used to perform Gaussian beam model correction on the terahertz extended phase data to make it conform to the projection data collection method under the ray model, thereby forming terahertz extended phase data based on the Gaussian model; The terahertz extended phase tomography image reconstruction includes filtered back-projection image reconstruction and iterative image reconstruction, wherein the filtered back-projection image reconstruction performs nearest neighbor interpolation back-projection on the terahertz extended phase data after terahertz extended phase data preprocessing, and the iterative image reconstruction performs a finite number of iterations on the terahertz extended phase data after terahertz extended phase data preprocessing to achieve image reconstruction.

2. The electronic terahertz continuous wave extended phase tomography image reconstruction method according to claim 1, characterized in that: The terahertz extended phase data acquisition device is used to collect the unexpanded phase data of the terahertz wave, which is a necessary prerequisite for the reconstruction of the terahertz extended phase tomography image. It includes an electronic terahertz continuous wave radiation source, a terahertz radiation antenna, a first reflecting mirror of the terahertz beam, a terahertz object three-dimensional rotation displacement console, a second reflecting mirror of the terahertz beam, a terahertz receiving antenna, a phase shift extraction circuit and a computer; the terahertz extended phase data acquisition device is used to realize the interaction between the terahertz beam and the target object, and record the phase offset data of the target object carried by the terahertz beam after the interaction, which is the unexpanded phase data of the terahertz wave.

3. The electronic terahertz continuous wave extended phase tomography image reconstruction method according to claim 1, characterized in that: The terahertz extended phase extraction method utilizes phase expansion technology to expand the unexpanded phase data of the measured terahertz wave acquired by a terahertz extended phase data acquisition device, thereby obtaining the extended phase data of the terahertz beam after the target object interacts with the terahertz beam.

4. The electronic terahertz continuous wave extended phase tomography image reconstruction method according to claim 1, characterized in that: This method can be used to achieve structural imaging and electromagnetic property testing of target objects. It can be used for positioning and detecting medical lesions in the field of medical imaging to achieve three-dimensional structural imaging of lesions. It can also be used for defect detection in the field of non-destructive testing of electromagnetic materials and to measure the refractive index of materials.

Images