WO2018121325A1

WO2018121325A1 - Multi-angle imaging data processing method and apparatus

Info

Publication number: WO2018121325A1
Application number: PCT/CN2017/116970
Authority: WO
Inventors: 吴珊珊; 王强; 刘晶晶; 廖育华; 张金宇
Original assignee: 同方威视技术股份有限公司
Priority date: 2016-12-29
Filing date: 2017-12-18
Publication date: 2018-07-05
Also published as: CN108267465A

Abstract

A multi-angle imaging data processing method and apparatus. The method comprises: irradiating, using X-ray, and from two or more angles, an object to be checked (120) to obtain multi-angle imaging data (310); extracting, from the multi-angle imaging data, data representing the same region in the object to be checked (120) (320); reconstructing, according to the extracted data, a cross-sectional area (330); processing, on the basis of gradient descent, the reconstructed cross-sectional area to obtain a processed reconstructed cross-sectional area; determining, according to the processed reconstructed cross-sectional area, an effective atomic number and a characteristic density in the region; and determining, on the basis of the determined effective atomic number and characteristic density, whether the region contains a prohibited article.

Description

Multi-view imaging data processing method and device

Reference to related application

The present application claims priority to Chinese Patent Application No. 201611257546.0, filed on Dec. 29, s.

Technical field

The present invention relates to the field of security technologies, and in particular, to a method and apparatus for processing multi-view imaging data.

Background technique

For baggage security inspection, X-ray security technology is currently the most widely used type of non-contact security technology. Most of the X-ray inspection equipment used in the early days were X-ray technology using single-view single-energy. With the continuous development of X-ray technology, dual-energy X-ray technology that can estimate the effective atomic number has emerged. In addition, in order to meet the security requirements of high detection accuracy, X-ray tomography (X-CT) technology has been introduced into the security field. However, the amount of data that X-CT technology needs to process is very large, the time for obtaining images is long, and the manufacturing cost is high.

The multi-view device is a device between a common single-view device and a CT device. For example, the viewing angle is generally 3 or more and 6 or less. Although the multi-view device does not accurately reconstruct the complete data of the 360-degree scan like CT, it can partially reconstruct the 3D shape of the object with a limited number of views, and can be based on partial reconstruction. It can estimate the equivalent atomic number and density of the object, and can partially overcome the problem of mutual occlusion between objects, so as to achieve more accurate detection and identification of prohibited items.

In addition to the advantages of a single-view product, the multi-view device can also more accurately detect and identify dangerous substances in the test object, reduce the dependence on the security inspectors, and the detection speed is higher, and the cost is lower than CT. Channel size and detection speed are not limited as much as CT, and there is a great market prospect.

Summary of the invention

It is an object of the present disclosure, at least in part, to provide a method and apparatus for multi-view imaging data processing that enables better image reconstruction from multi-view imaging data.

According to an aspect of the present disclosure, there is provided a method of processing multi-view imaging data, wherein the multi-view imaging data is obtained by irradiating an object to be inspected from two or more viewing angles by X-rays, the method The method comprises: extracting data representing the same region in the object to be inspected from the multi-view imaging data; performing cross-sectional reconstruction on the region according to the extracted data; processing the reconstructed cross-section based on the gradient descent, and obtaining the reconstructed cross-section after processing; Determine the equivalent atomic number and density of the region based on the reconstructed cross section after processing; and determine whether the region contains contraband based on the determined equivalent atomic number and feature density.

According to an embodiment of the present disclosure, the multi-view imaging data may be obtained by dual energy X-rays. At this time, material properties can be better obtained. For example, extracting data representing the same region in the object to be inspected from the multi-view image data may include obtaining a base material decomposition coefficient from the multi-view image data based on the base material coefficient table.

According to an embodiment of the present disclosure, cross-sectional reconstruction may be performed using an algebraic reconstruction technique (ART). For example, the section to be reconstructed can be discretized into N cell pixels x _j , where j=1, 2, . . . , N, and can be reconstructed according to the following formula:

Where k = 0, 1, ..., K is the number of iterations of ART, x(k) _j is the value of cell x _j after the kth iteration, and x(k+1) _j is the k+1th time The value of the cell x _j after iteration,

x(k+1) _N } represents the cross section reconstructed after the k+1th iteration, λ is the relaxation parameter, p _i is the projection value corresponding to ray i, and the projection coefficient w _ij is the length of the ray i passing through the cell x _j .

According to an embodiment of the present disclosure, the reconstructed cross section may be processed using a Total Variation (TV) algorithm. For example, the reconstructed section can be processed according to the following formula:

Where l=0,1,...,L is the number of iterations of TV, and x(0,k+1) is the section reconstructed after the k+1th iteration in ART

x(L,k+1) as the reconstructed section after processing

And as the initial value of the k+2th iteration in ART, α is the variable that controls the rate of gradient descent, and d is the difference between the reconstructed sections before and after the k+1th iteration of ART, expressed as

Is a normalized TV gradient, ie

According to an embodiment of the present disclosure, the method may further include performing 2-average classification on the pixels in the reconstructed cross section after processing.

According to another aspect of the present disclosure, there is provided an imaging data processing apparatus for processing multi-view imaging data, wherein the object to be inspected is irradiated from two or more angles of view by X-rays to obtain the plurality of Viewing imaging data, the imaging data processing comprising: a communication interface for receiving multi-view imaging data; a memory for storing the received multi-view imaging data; and a processor for: extracting the representative from the multi-view imaging data The data of the same area in the object; according to the extracted data, the section is reconstructed; the reconstructed section is processed based on the gradient descent, and the reconstructed section is processed; the area is determined according to the reconstructed section after processing Effect atomic number and feature density; and determining whether the region contains contraband based on the determined equivalent atomic number and feature density.

According to an embodiment of the present disclosure, the imaging data processing device may further include: a display for displaying a reconstruction result.

According to an embodiment of the present disclosure, the processor may perform cross-sectional reconstruction using Algebraic Reconstruction Technique (ART) as described above, and may utilize a Total Variation (TV) algorithm to process the reconstructed cross-section.

According to still another aspect of the present disclosure, there is provided a computer readable recording medium having stored thereon executable instructions that, when executed by one or more processors, cause the one or more processors to execute The following operation: extracting data representing the same region in the object to be inspected from the multi-view imaging data; performing cross-section reconstruction on the region according to the extracted data; processing the reconstructed cross-section based on the gradient descent, and obtaining the reconstructed cross-section after processing Determining the equivalent atomic number and feature density of the region according to the reconstructed cross section after processing; and determining whether the region contains contraband based on the determined equivalent atomic number and feature density.

According to an embodiment of the present disclosure, the reconstructed cross section is processed based on the gradient descent when the cross section is reconstructed. In this way, the pixels can be made smoother.

DRAWINGS

The above and other objects of the present disclosure, the description of the embodiments of the present disclosure, Features and advantages will be more clear, in the drawing:

FIG. 1 is a perspective view schematically showing a security check system according to an embodiment of the present disclosure;

2(a) is a perspective view schematically showing a radiation source and detector layout in accordance with an embodiment of the present disclosure;

2(b) is a plan view schematically showing a radiation source and detector layout in accordance with an embodiment of the present disclosure;

FIG. 3A is a flowchart schematically illustrating a multi-view imaging data processing method according to an embodiment of the present disclosure; FIG.

FIG. 3B is a flow chart schematically showing a cross-section reconstruction method according to an embodiment of the present disclosure; FIG.

4 is a schematic diagram showing the principle of an Algebraic Reconstruction Technique (ART);

FIG. 5 is a block diagram schematically showing a multi-view imaging data processing apparatus according to an embodiment of the present disclosure.

Throughout the drawings, the same or similar reference numerals indicate the same or similar components.

detailed description

Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. It should be understood, however, that the description is only illustrative, and is not intended to limit the scope of the disclosure. In addition, descriptions of well-known structures and techniques are omitted in the following description in order to avoid unnecessarily obscuring the concept of the present disclosure.

The terminology used herein is for the purpose of describing the particular embodiments, The words "a", "an", "the" and "the" In addition, the terms "including", "comprising", etc., are used to indicate the presence of the described features, steps, operations and/or components, but do not exclude the presence or addition of one or more other features, steps, operations or components .

All terms (including technical and scientific terms) used herein have the meaning commonly understood by one of ordinary skill in the art, unless otherwise defined. It should be noted that the terms used herein are to be interpreted as having a meaning consistent with the context of the present specification and should not be interpreted in an ideal or too rigid manner.

Some block diagrams and/or flowcharts are shown in the drawings. It will be understood that some blocks or combinations of the block diagrams and/or flowcharts can be implemented by computer program instructions. These computer program instructions can be provided to a general purpose computer, a special purpose computer or other programmable data processing device processor, such that Instructions, when executed by the processor, can create means for implementing the functions/operations described in these block diagrams and/or flowcharts.

Thus, the techniques of this disclosure may be implemented in the form of hardware and/or software (including firmware, microcode, etc.). Additionally, the techniques of this disclosure may take the form of a computer program product on a computer readable medium storing instructions for use by or in connection with an instruction execution system. In the context of the present disclosure, a computer readable medium can be any medium that can contain, store, communicate, propagate or transport the instructions. For example, a computer readable medium can include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, device, or propagation medium. Specific examples of the computer readable medium include: a magnetic storage device such as a magnetic tape or a hard disk (HDD); an optical storage device such as a compact disk (CD-ROM); a memory such as a random access memory (RAM) or a flash memory; and/or a wired /Wireless communication link.

FIG. 1 is a perspective view schematically showing a security check system according to an embodiment of the present disclosure.

As shown in FIG. 1, the security system 100 according to this embodiment may include a security channel 110. The object to be inspected 120 can pass through the security inspection channel 110, for example, by a transport device (for example, a conveyor belt). A radiation source and a detector may be disposed in the security channel 110. When the object to be inspected 120 travels through the security inspection channel 110, radiation emitted by the radiation source (usually X-rays) is irradiated onto the object to be inspected 120, and the detector can detect radiation transmitted through the object to be inspected 120 (optionally, The radiation scattered by the object to be inspected 120 can be detected).

According to an embodiment of the present disclosure, a plurality of (eg, two or more) radiation sources and corresponding plurality of detectors may be disposed to align the object to be inspected 120 from a plurality of (eg, two or more) viewing angles. Scan imaging is performed. Hereinafter, the arrangement of the radiation source and the detector will be described in further detail.

The security system 100 can also include a control device 130. A control device 130 (eg, a computer) can be connected (wired or wirelessly) to the security channel 110 to control components in the security channel, such as transmission devices, radiation sources, and detectors, as well as to process data detected by the detector, etc. . According to an embodiment of the present disclosure, the control device 130 may process the multi-view imaging data and reconstruct an image of the object 120 to be inspected (or a portion thereof). The reconstructed image can be displayed on an output device (eg, a display) for viewing by security personnel. In addition, if contraband items such as explosives, weapons, etc. are identified, an alarm can be output (for example, highlighted on the display or a beep).

The security inspection system 100 can be applied to stations, airports, customs, etc. to check items. Prevent places where contraband is entered.

2(a) is a perspective view schematically showing a radiation source and detector layout according to an embodiment of the present disclosure, and FIG. 2(b) is a schematic diagram showing a radiation source and a detector according to an embodiment of the present disclosure. The floor plan of the layout.

As shown in Figures 2(a) and 2(b), within the security channel 110, radiation sources V1, V2, V3 and corresponding detectors 112-1, 112-2 and 112-3 are disposed.

Radiation sources V1, V2 and V3 may be X-ray generators and include corresponding drive circuits. In addition, detectors 112-1, 112-2, and 112-3 may be provided corresponding to the respective radiation sources V1, V2, and V3. These detectors can for example be L-shaped detectors. Each radiation source can emit a fan beam having a certain angle of incidence, and a corresponding detector can be opposite the fan beam to at least partially receive the transmitted radiation. Thus, the radiation source, together with the corresponding detector, defines the viewing angle of the radiation source. The viewing angles of all of the radiation sources disposed within the security channel 110 together achieve a substantially 360° coverage of the object being inspected.

Each radiation source and detector pair (eg, (V1, 112-1), (V2, 112-2), (V3, 112-3)) may be in a plane (the detection field of view is also in the plane). The planes of the respective radiation sources and detector pairs may be parallel to each other and spaced apart from each other. For example, the planes may be planes (x-y planes) substantially orthogonal to the traveling direction of the object to be inspected 120 (the longitudinal direction of the security inspection channel 110, that is, the z-axis), and are spaced apart from each other along the z-axis direction.

As shown in FIGS. 2(a) and 2(b), in this example, the radiation source V1 may be disposed at the lower left of the security inspection channel 110, and the corresponding detector 112-1 may be disposed at the right side wall of the security inspection channel 110. And the top wall; the radiation source V2 may be disposed at the lower right side of the security inspection channel 110, and the corresponding detector 112-2 may be disposed on the left side wall and the top wall of the security inspection channel 110; the radiation source V3 may be disposed in the security inspection channel 110 On the right side, and corresponding detectors 112-3 may be disposed on the left and top walls of the security channel 110. Of course, the radiation source and detector can be arranged in different ways.

According to an embodiment of the present disclosure, dual energy scanning can be achieved. For example, each of the radiation sources V1, V2, and V3 can radiate both high and low energy X-rays, and the respective detectors 112-1, 112-2, and 112-3 can include a detection array that detects the corresponding X-rays. Dual energy scanning helps identify material information.

The detectors 112-1, 112-2, and 112-3 can detect radiation transmitted from the object to be inspected 120, and can convert the detected radiation into an electrical signal. This detection signal can be sent to the control device 130 for processing. Due to scanning imaging at multiple viewing angles, this detection signal is referred to herein as "multi-view imaging number" According to the inspection object 120 traveling through the security inspection channel 110, multi-view scan imaging data at different sections of the object to be inspected 120 can be obtained. According to these multi-view scan imaging data, the cross sections of the object to be inspected 120 can be reconstructed. And the three-dimensional contour of the object to be inspected 120 can be reconstructed therefrom.

In this example, three radiation sources V1, V2, and V3 (and corresponding detectors 112-1, 112-2, and 112-3) are shown to achieve scanning imaging at three viewing angles. However, the present disclosure is not limited thereto. For example, scanning imaging at two or more viewing angles (eg, 4-6 viewing angles) is also possible.

FIG. 3A is a flow chart schematically illustrating a multi-view imaging data processing method according to an embodiment of the present disclosure. For example, the multi-view imaging data processing method can be performed by the control device 130.

As shown in FIG. 3A, the multi-view imaging data processing method 300 in accordance with this embodiment can include obtaining multi-view imaging data at operation 310. For example, as described above, the multi-view imaging data is obtained by irradiating X-rays emitted from the radiation source onto the object to be inspected 120 and detecting the transmitted X-rays by the detector when the object to be inspected 120 travels through the security inspection channel 110. . Control device 130 can receive multi-view imaging data over a wired or wireless connection.

At operation 320, image segmentation may be performed on the multi-view imaging data, for example, into foreground and background. There are numerous techniques in the art for image segmentation. For example, the contour or edge of each item within the object under inspection 120 (eg, the trunk) can be detected by edge detection techniques to obtain a foreground image (and thus the background can be rejected). For the prospects obtained, the possible dangerous areas can be identified. For example, in the case of dual energy scanning imaging, material properties can be identified based on the difference between the absorption of high energy X-rays by the material and the absorption of low energy X-rays. For example, based on the base material coefficient table, the base material decomposition of the multi-view imaging data can be performed to obtain the base material decomposition coefficient, and the possible dangerous regions are identified based on the base material decomposition coefficient. The following processing can be performed for these identified potentially dangerous areas in order to improve processing efficiency.

In the data for these potentially dangerous areas, data for the same area at different viewing angles (for example, data decomposed by the base material) can be extracted. For example, if a certain area under one view has a certain matching relationship with an area under another view (for example, in terms of geometry, collected high and low energy data, etc.), then the two areas can be considered to be in fact It is the performance of the same region from different perspectives.

Then, at operation 330, a cross-sectional reconstruction of the region may be based on the extracted data at different perspectives for the same region. Next, the section reconstruction will be described in further detail in conjunction with FIG. 3B. Reason.

After reconstructing the cross section, at operation 340, various processes can be performed based on the reconstructed cross section. For example, a three-dimensional contour of the region can be constructed from a plurality of reconstructed sections of the same region (eg, by splicing the various sections). In addition, the equivalent atomic number and density at the corresponding region can be calculated from the reconstructed cross section. Whether the area contains contraband can be determined based on the determined atomic number and density determined. For example, atomic numbers and densities in multiple cross sections can be averaged. If the average falls into the contraband area in the contraband database, it can be determined that the area contains contraband and can be alarmed; if it falls into other areas outside the contraband area, it can be determined that the area is a safe product, not Give an alarm.

FIG. 3B is a flow chart that schematically illustrates a cross-section reconstruction method in accordance with an embodiment of the present disclosure.

As shown in FIG. 3B, the cross-section reconstruction method 330 according to this embodiment may include reconstructing pixel values in the cross-section according to multi-view imaging data (data for different viewing angles of the same region as described above) in operation 331. Here, the reconstruction may be referred to as "preliminary reconstruction" and the reconstructed pixel values may be referred to as "preliminary reconstruction (pixel) values."

There are a variety of cross-section reconstruction techniques in the art. According to an embodiment of the present disclosure, an algebraic reconstruction technique (ART) may be used. In ART, pixel values are reconstructed by multiple iterations.

Specifically, referring to FIG. 4, for the section 421 to be reconstructed, it can be discretized into N cell pixels x _j (j = 1, 2, ..., N). For each cell, its reconstruction value can be calculated by the following iteration:

Where k = 0, 1, ..., K is the number of iterations of ART (may be specified in advance, or can be determined adaptively), x(k) _j is the value of cell x _j after kth iteration, x (k+1) _j is the value of the cell x _j after the k+1th iteration,

x(k+1) _N } represents the cross section reconstructed after the k+1th iteration, and λ is the relaxation parameter (can be selected according to experience, for example, it can be evaluated in advance according to the actual situation, such as λ=0.5), p _i is the ray i The corresponding projection value (that is, the signal detected by the detector 412 after the ray i is transmitted through the object to be inspected may be the data after the decomposition of the base material described above), and the projection coefficient w _ij is the ray i passing through the cell The length of x _j . These rays i are the individual X-rays emitted by the radiation source, and their angle (and relative to the detector) is determined by the physical layout.

According to an embodiment of the present disclosure, at operation 333, the preliminary reconstruction values may be processed based on the gradient descent to smooth between pixels. For example, you can use the Total Variation (TV) algorithm to This kind of processing.

Specifically, the TV algorithm can be performed for each iteration result in the preliminary reconstruction process. For example, the cross section obtained after the k+1th iteration of the initial reconstruction

It can be used as the initial value in the TV algorithm. The value obtained by the TV algorithm can be used as the initial value of the k+2th iteration in the initial reconstruction process, and so on.

In the case of initial reconstruction using ART as described above, the TV algorithm can be performed as follows:

Where l = 0, 1, ..., L is the number of iterations of the TV (can be specified in advance, or can be determined adaptively), x (0, k + 1) is the reconstruction after the k+1th iteration in ART Cross section

x(L,k+1) as the reconstructed section after processing

And can be used as the initial value of the k+2th iteration in ART (ie, the reconstructed section after processing)

The value x(k+1) _j of each cell can replace the corresponding preliminary reconstruction value x(k+1) _j ) calculated in ART above, and α is a variable that controls the rate of gradient decrease (in general, the value of α is more Small, the TV smoothing effect is weaker, and the larger the α value, the stronger the TV smoothing effect; on the other hand, the relationship between the α value and the maximum iteration number L of the TV has a balance relationship, so the α value can be selected according to the demand. , for example, α = 0.2), where d represents the difference between the reconstructed sections before and after the k+1th iteration of ART, expressed as

Is a normalized TV gradient, ie (where ||x(l,k+1)|| _TV represents the total variation of x(l,k+1))

It should be noted that the above cross-section reconstruction assumes that the density inside the object is substantially uniform. That is, within the scope of the object, each pixel has substantially the same value; and outside of the object range, the pixel value is substantially zero (0). That is, the pixel value of the area where the first object is located on the image is the same first positive value, and the pixel value of the area where the second object is located may be the same second positive value (may be different from the first positive value); The pixel value between the blank area (background) can be 0. Under this assumption, after the cross-section reconstruction, the pixel values can be classified by 2-means (for example, classified into 0 pixels and 1 pixel).

As shown in FIG. 5, the processing device 530 according to this embodiment (for example, the above-described control device 130) A bus 531, a processor 532, a memory 533, an input/output (I/O) interface 534, a display 535, and a communication interface 536 can be included. In various embodiments of the present disclosure, at least one of the above elements may be omitted, or other elements may be added to the processing device 530.

The bus 531 can connect the above-described components 532 to 536 to each other and communicate between the above components (for example, control messages and/or data).

The processor 532 may include, for example, a central processing unit (CPU), an application processor (AP), or a communication processor (CP) or the like. Processor 532 can perform data processing or operations related to communication and/or control of at least one other component of processing device 530.

Memory 533 can include volatile memory and/or non-volatile memory. Memory 533 can store instructions or data related to at least one other component in processing device 530. According to an embodiment of the present disclosure, the memory 533 may store software and/or programs. Processor 532 can perform the corresponding operations by executing software and/or programs stored in memory 533, such as those described above in connection with Figures 3A and 3B. In addition, a table of base material coefficients may also be stored in the memory 533.

I/O interface 534 can be used to communicate instructions or data input from a user or other external device to other components of processing device 530. Additionally, I/O interface 534 can output instructions or data received from other components of processing device 530 to a user or other external device. According to an embodiment, the I/O interface 534 may alert via an output device, such as an audio device that processes an audio signal, or an audio device (eg, in the presence of a contraband).

Display 535 can include, for example, a liquid crystal display (LCD), a light emitting diode (LED) display, an organic light emitting diode (OLED) display, and the like. Display 535 can present various content (eg, text, images, video, etc.) to, for example, a user. Display 535 can include a touch screen and can receive touch input. As noted above, display 535 can present the reconstruction results to the user.

Communication interface 536 can communicate between processing device 530 and other devices (e.g., detectors). For example, communication interface 536 can be coupled to a network, such as a computer network (eg, a local area network (LAN), wide area network (WAN), the Internet) via wireless communication or wired communication to communicate with other devices.

Wireless communication may use cellular communication protocols such as Long Term Evolution (LET), Code Division Multiple Access (CDMA), Wideband CDMA (WCDMA), or Global System for Mobile Communications (GSM). Wireless communication may include, for example, short-range communication such as WiFi, Bluetooth, Near Field Communication (NFC), and the like. Wired communication can be packaged Including Universal Serial Bus (USB).

The embodiments of the present disclosure have been described above. However, the examples are for illustrative purposes only and are not intended to limit the scope of the disclosure. Although the various embodiments have been described above, this does not mean that the measures in the various embodiments are not advantageously used in combination. The scope of the disclosure is defined by the appended claims and their equivalents. Numerous alternatives and modifications may be made by those skilled in the art without departing from the scope of the present disclosure.

Claims

A method of processing multi-view imaging data, wherein the multi-view imaging data is obtained by irradiating an object to be inspected from two or more viewing angles by X-rays, the method comprising:

Extracting data representing the same region in the object to be inspected from the multi-view imaging data;

Section reconstruction of the area based on the extracted data;

The reconstructed section is processed based on the gradient descent, and the reconstructed section after the treatment is obtained;

Determine the equivalent atomic number and density of the region based on the reconstructed cross section after processing;

Based on the determined equivalent atomic number and feature density, it is determined whether the area contains contraband.
The method of claim 1 wherein said multi-view imaging data is obtained by dual energy X-rays.
The method according to claim 2, wherein extracting data representing the same region in the object to be inspected from the multi-view image data comprises:

Based on the base material coefficient table, the base material decomposition coefficient is obtained from the multi-view image data.
The method of claim 1 wherein the cross-sectional reconstruction is performed using an algebraic reconstruction technique ART.
The method of claim 4, wherein

Discretizing the cross section to be reconstructed into N cell pixels x j , where j=1, 2,..., N,

And rebuild according to the following formula:

Where k = 0, 1, ..., K is the number of iterations of ART, x(k) j is the value of cell x j after the kth iteration, and x(k+1) j is the k+1th time The value of the cell x j after iteration,

Representing the section reconstructed after the k+1th iteration, λ is the relaxation parameter, p i is the projection value corresponding to the ray i, and the projection coefficient w ij is the length of the ray i passing through the cell x j .
A method according to claim 4 or 5, wherein the reconstructed cross section is processed using a fully variable TV algorithm.
The method of claim 6 wherein the reconstructed cross section is processed according to the following formula:

Where l=0,1,...,L is the number of iterations of TV, and x(0,k+1) is the section reconstructed after the k+1th iteration in ART
x(L,k+1) as the reconstructed section after processing
And as the initial value of the k+2th iteration in ART, α is the variable that controls the rate of gradient descent, and d is the difference between the reconstructed sections before and after the k+1th iteration of ART, expressed as

Is a normalized TV gradient, ie
The method of claim 1 further comprising: performing a 2-mean classification on the pixels in the reconstructed cross section after processing.
An imaging data processing apparatus for processing multi-view imaging data, wherein the multi-view imaging data is obtained by irradiating an object to be inspected from two or more viewing angles by X-rays, the imaging data processing comprising:

a communication interface for receiving multi-view imaging data;

a memory for storing the received multi-view imaging data;

Processor for:

Extracting data representing the same region in the object to be inspected from the multi-view imaging data;

Section reconstruction of the area based on the extracted data;

The reconstructed section is processed based on the gradient descent, and the reconstructed section after the treatment is obtained;

Determining the equivalent atomic number and characteristic density of the region based on the reconstructed cross section after processing;

Based on the determined equivalent atomic number and feature density, it is determined whether the area contains contraband.
The imaging data processing device of claim 9, further comprising:

A display that displays the reconstruction results.
The imaging data processing apparatus according to claim 9, wherein the processor performs the cross-section reconstruction using an algebraic reconstruction technique ART.
The imaging data processing device according to claim 11, wherein

The processor discretizes the cross section to be reconstructed into N cell pixels x j , where j=1, 2, . . . , N,

And rebuild according to the following formula:

Where k = 0, 1, ..., K is the number of iterations of ART, x(k) j is the value of cell x j after the kth iteration, and x(k+1) j is the k+1th time The value of the cell x j after iteration,

Representing the section reconstructed after the k+1th iteration, λ is the relaxation parameter, p i is the projection value corresponding to the ray i, and the projection coefficient w ij is the length of the ray i passing through the cell x j .
The imaging data processing apparatus according to claim 11 or 12, wherein the processor processes the reconstructed section using a total variation TV algorithm.
The imaging data processing apparatus according to claim 13, wherein the processor processes the reconstructed section according to the following formula:

Where l=0,1,...,L is the number of iterations of TV, and x(0,k+1) is the section reconstructed after the k+1th iteration in ART
x(L,k+1) as the reconstructed section after processing
And as the initial value of the k+2th iteration in ART, α is the variable that controls the rate of gradient descent, and d is the difference between the reconstructed sections before and after the k+1th iteration of ART, expressed as

Is a normalized TV gradient, ie
A computer readable recording medium having stored thereon executable instructions that, when executed by one or more processors, cause the one or more processors to:

Extracting data representing the same region in the object to be inspected from the multi-view imaging data;

Section reconstruction of the area based on the extracted data;

The reconstructed section is processed based on the gradient descent, and the reconstructed section after the treatment is obtained;

Determining the equivalent atomic number and characteristic density of the region based on the reconstructed cross section after processing;

Based on the determined equivalent atomic number and feature density, it is determined whether the area contains contraband.