Imaging method and device based on battery pack body detection

Abstract

The invention discloses an imaging method and device based on battery pack body detection. The method comprises the following steps: firstly, dividing a plurality of local scanning areas according to the geometric dimension of a to-be-detected battery pack and the size of a single scanning imaging area; secondly, motion data collection is conducted on each local scanning area, and projection data are obtained through the moving tracks of a radiation source and a detector; secondly, performing local data reconstruction on the acquired projection data by adopting a joint algebraic reconstruction algorithm to obtain a three-dimensional image of a local area; and finally, splicing the three-dimensional images of the local areas to form a complete three-dimensional image of the battery pack body. Through a multi-section straight line parallel scanning mode, truncation artifacts can be effectively avoided, and the image quality and the data utilization efficiency are improved.

Description

Technical Field

[0001] The present invention relates to the field of detection technologies, and particularly to an imaging method and device based on battery pack body detection. Background Art

[0002] With the growth of the global demand for clean energy, the new energy vehicle market is expanding rapidly. As one of the core components of new energy vehicles, the safety and reliability of power batteries have become the focus of attention. A power battery is usually composed of a large battery pack (PACK) consisting of multiple battery cells, and the internal structure of these battery packs is complex, containing a variety of materials and components. To ensure the safety of the battery pack, precise non-destructive detection of its interior is required. Industrial Computed Tomography (Industrial CT for short) technology, as a non-invasive detection method, can provide high-resolution three-dimensional images, enabling inspectors to check the internal structure, foreign objects, cracks, etc. of the battery pack without damaging the sample, and evaluate its sealing performance and capacity.

[0003] Traditional industrial CT technology mainly relies on two-dimensional detectors and uses a circular trajectory (CircularTrajectory Cone Beam CT, abbreviated as CBCT) to obtain projection data of the object to be inspected. The basic principle of this technology is to let the X-ray source move in a circular motion around the object to be measured while recording the penetration information at different angles. Traditional circular trajectory CBCT has certain limitations. For example, it may encounter problems with field of view limitations when dealing with large or complex-shaped objects. In addition, since each scan can only cover a limited angular range, for some specific application scenarios, it may not be possible to obtain sufficient information for accurate reconstruction. Summary of the Invention

[0004] Based on this, the embodiments of the present application provide an imaging method and device based on battery pack body detection. In this method, the X-ray source and the detector can move along multiple parallel planes, so as to be able to cover the entire object to be detected. The multi-segment flat scan reconstruction algorithm can perform high-resolution parallel scans in each region, and seamlessly splice the reconstruction results of each independent region through an optimization algorithm. By means of multi-segment linear parallel scans, truncation artifacts can be effectively avoided, and the image quality and data utilization efficiency can be improved.

[0005] In a first aspect, an imaging method based on battery pack body detection is provided, and the method includes:

[0006] S1 Confirm segmented scan regions: Determine the basic geometric dimensions of the battery pack to be detected, and divide multiple local scan regions according to the geometric dimensions of the battery pack to be detected and the size of the single-scan imaging region;

[0007] S2 Local area motion acquisition data: Motion acquisition data is performed on each local scanning area, and projection data is obtained through the movement trajectories of the ray source and the detector;

[0008] S3 Local data reconstruction: The projection data of each local scanning area is reconstructed to obtain a three-dimensional image of the local area;

[0009] S4 Area stitching: The three-dimensional images of each local area are stitched together to form a complete three-dimensional image of the battery pack body.

[0010] Optionally, the step of confirming the segmented scanning area specifically includes:

[0011] Determine the length, width, and thickness of the battery pack to be detected;

[0012] Set the size of the single-scan imaging area, including the length in the X direction and the width in the Y direction;

[0013] Set the size of the overlapping area between adjacent scanning areas, including the overlapping length in the X direction and the overlapping width in the Y direction;

[0014] According to the set parameters, calculate the specific position of each scanning area to ensure that there is an overlapping part between adjacent scanning areas to ensure the continuity of information.

[0015] Optionally, the step of local area motion acquisition data specifically includes:

[0016] Define the distance from the detector to the sample center and the distance from the ray source to the sample center;

[0017] Determine the movement trajectories of the ray source and the detector so that the ray source and the detector move in the diagonal direction in the X direction to cover the entire local scanning area while keeping the position in the Y direction unchanged.

[0018] Optionally, in the step of local data reconstruction, the jointly algebraic reconstruction algorithm is used to reconstruct the acquired projection data, including:

[0019] Calculate the angle between the ray source detector and the plane at each position. Without placing the battery pack, acquire the empty-field data at each angle, place the battery pack on the sample stage, and acquire the projection data at each moving position;

[0020] Set the initial state of the reconstructed object and set the initial value of all voxels to 0;

[0021] At each angle, calculate the system matrix, and the elements of this matrix represent the intersection line length of the ray and the voxel;

[0022] Calculate the residual data at each angle and perform iterative updates until all angles are updated to obtain the final three-dimensional image of the local area.

[0023] Optionally, calculate the angle between the ray source detector and the plane at each position, specifically including according to the formula:

[0024]

[0025] Calculate the angle θ between the ray source detector and the plane at each position, where old represents the distance from the detector to the sample center, sod represents the distance from the ray source to the sample center, and x s (k) represents the ray source position in the X direction, and x d (k) represents the detector center position in the X direction.

[0026] Optionally, collect the empty field data at each angle, place the battery pack on the sample stage, and collect the projection data at each moving position, specifically including according to the formula:

[0027]

[0028] Calculate the projection data P θ , where I 0θ represents the empty field data at each angle, and I θ represents the projection data at each moving position.

[0029] Optionally, calculate the residual data at each angle and perform iterative updates, specifically including through the formula:

[0030]

[0031] Calculate the residual data at each angle θ i , where the pixel value of each corresponding point in P θ is P θ (i,j), where i and j are the detector pixel coordinates respectively. Set X = (X1, X2,..., X N ) as the reconstructed object, N is the total number of voxels, initialize all elements of X to 0, and W θ (q,n) is the system matrix;

[0032] Update X at each angle θ i :

[0033]

[0034] After traversing and updating all angles θ i , obtain the final reconstruction result X.

[0035] Optionally, the step of region stitching includes:

[0036] For the pixel points in the non-overlapping part, directly use the pixel values of the pixel points in the non-overlapping part in the corresponding local region image;

[0037] For the overlapping part in the X direction, take the average of the pixel values at the corresponding positions in the left and right adjacent scanning regions as the final pixel value;

[0038] For the overlapping part in the Y direction, take the average of the pixel values at the corresponding positions in the upper and lower adjacent scanning regions;

[0039] For the overlapping part at the four corners, which involves four adjacent scanning regions, take the average of the pixel values at the corresponding positions in these four regions.

[0040] Optionally, the method further includes:

[0041] Replace the scanning trajectory of each segment in an equidistant or equiangular manner;

[0042] And perform local data reconstruction through algebraic reconstruction technique, simultaneous iterative reconstruction technique or conjugate gradient least squares method.

[0043] In a second aspect, an imaging device based on battery pack body detection is provided. The device includes:

[0044] A region confirmation module, configured to determine the basic geometric dimensions of the battery pack to be detected, and divide a plurality of local scanning regions according to the geometric dimensions of the battery pack to be detected and the size of a single scan imaging region;

[0045] A data acquisition module, configured to perform motion acquisition of data for each local scanning region, and obtain projection data through the moving trajectories of a ray source and a detector;

[0046] A data reconstruction module, configured to reconstruct the projection data of each local scanning region to obtain a three-dimensional image of the local region;

[0047] A region stitching module, configured to stitch the three-dimensional images of each local region to form a complete three-dimensional image of the battery pack body.

[0048] By adopting the method of multi-segment linear parallel scanning, the present invention can effectively avoid the truncation artifacts caused by the field of view limitation in traditional circular trajectory scanning. This makes the reconstructed image clearer and more accurate, reducing edge blurring and distortion. For large or irregularly shaped objects (such as large battery packs), it is difficult for the traditional circular trajectory scanning method to fully cover all their areas. However, the method of the present invention allows the X-ray source and detector to move along multiple parallel planes, so as to comprehensively capture all details of the object to be detected and ensure the integrity of information. By performing multiple scans at different angles and planes, more abundant projection data can be obtained. After being processed by the optimized algorithm, these data can better reflect the true structure inside the object and improve the utilization efficiency of the data. Since there is no need to be limited to a fixed circular trajectory, the system of the present invention can flexibly adjust the scanning path and plane position according to the actual situation. This provides greater convenience and adaptability for dealing with various different application scenarios. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only exemplary, and for those of ordinary skill in the art, without creative efforts, other implementation drawings can also be obtained according to the provided drawings.

[0050] Figure 1 It is a flowchart of the steps of an imaging method for detecting a battery pack body provided by an embodiment of the present application;

[0051] Figure 2 It is a mechanical system diagram of the implementation scenario provided by an embodiment of the present application;

[0052] Figure 3 It is a schematic diagram of segmented regions provided by an embodiment of the present application;

[0053] Figure 4 It is a regional scanning trajectory diagram provided by an embodiment of the present application;

[0054] Figure 5 It is a regional reconstruction diagram provided by an embodiment of the present application;

[0055] Figure 6 It is a splicing result diagram provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0056] In order to make the purpose, technical solutions and advantages of the present application clearer, the following will further elaborate on the present application in conjunction with the drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0057] In the description of the present invention, the terms "include", "have" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units does not necessarily have to be limited to those steps or units that are clearly listed, but may also include other steps or units that are inherent to these processes, methods, products or devices although not clearly listed, or steps or units added based on further optimization schemes conceived in the present invention.

[0058] In the prior art, for the FDK algorithm based on circular trajectory scanning reconstruction, first, the X-ray source rotates around the object to be detected for one or more circles, and the projection data on the detector is recorded at each angular position. The collected raw projection data usually needs to go through some preprocessing steps, such as correction, normalization, and removal of bad pixels, etc., to ensure the quality of subsequent reconstruction. Then, a filter is applied to the projection data at each view angle, and finally, the filtered projection data is back-projected into the three-dimensional space to form the final reconstructed image.

[0059] The FDK algorithm assumes that the X-ray source rotates around the object to be detected along a circular trajectory, and the detector can cover the entire region of interest (ROI). However, in practical applications, especially when dealing with large or irregularly shaped objects, it may not be possible to place the detector within a complete circle, resulting in loss of some information. Due to the field of view limitation, if the object exceeds the effective imaging range of the detector, truncation artifacts will occur. These artifacts will affect the image quality, making the edges of the reconstructed image blurred, and even serious distortion may occur.

[0060] Since large objects such as large battery packs exceed the effective imaging range of the detector, using the traditional circular trajectory scanning method will result in missing some data, causing serious truncation artifacts. For irregularly shaped objects, especially large battery packs containing multiple materials and components inside, scanning within a single angular range is difficult to provide sufficient information for accurate reconstruction. To address the above problems, the present invention designs an imaging method for CT detection of battery packs. The X-ray source and the detector can move along multiple parallel planes, so as to be able to cover the entire object to be detected. The multi-segment flat scanning reconstruction algorithm can perform high-resolution parallel scanning in each region, and seamlessly splice the reconstruction results of each independent region through an optimization algorithm to form a complete and accurate data.

[0061] Please refer to Figure 1 , which shows a flowchart of an imaging method for battery pack body detection provided by an embodiment of the present application, and may include the following steps:

[0062] S1 Confirm segmented scanning areas: Determine the basic geometric dimensions of the battery pack to be detected. Divide multiple local scanning areas based on the geometric dimensions of the battery pack to be detected and the size of the single-scan imaging area.

[0063] Among them, determine the length, width, and thickness of the battery pack to be detected; set the dimensions of the single-scan imaging area, including the length in the X direction and the width in the Y direction; set the size of the overlapping area between adjacent scanning areas, including the overlapping length in the X direction and the overlapping width in the Y direction; calculate the specific positions of each scanning area according to the set parameters to ensure that there is an overlapping part between adjacent scanning areas to guarantee the continuity of information.

[0064] S2 Collect data by local area movement: Collect data by movement for each local scanning area, and obtain projection data through the movement trajectories of the radiation source and the detector.

[0065] Among them, define the distance from the detector to the sample center and the distance from the radiation source to the sample center;

[0066] Determine the movement trajectories of the radiation source and the detector so that the radiation source and the detector move in the diagonal direction in the X direction to cover the entire local scanning area while keeping the position in the Y direction unchanged.

[0067] S3 Reconstruct local data: Reconstruct the projection data of each local scanning area to obtain a three-dimensional image of the local area.

[0068] Among them, calculate the angle between the radiation source detector and the plane at each position. Without placing the battery pack, collect the empty-field data at each angle. Place the battery pack on the sample stage and collect the projection data at each moving position; set the initial state of the reconstructed object and set the initial value of all voxels to 0; at each angle, calculate the system matrix, and the elements of this matrix represent the intersection line length between the ray and the voxel; calculate the residual data at each angle and perform iterative updates until all angles are updated to obtain the final three-dimensional image of the local area.

[0069] S4 Stitch regions: Stitch the three-dimensional images of each local area to form a complete three-dimensional image of the battery pack body.

[0070] Among them, for the pixel points in the non-overlapping part, directly use the pixel value of this point in the corresponding local area image;

[0071] For the overlapping part in the X direction, take the average value of the pixel values at the corresponding positions in the left and right adjacent scanning areas as the final pixel value;

[0072] For the overlapping part in the Y direction, take the average value of the pixel values at the corresponding positions in the upper and lower adjacent scanning areas;

[0073] For the overlapping parts at the four corners, which involve four adjacent scanning regions, the average value of the pixel values at the corresponding positions in these four regions is taken.

[0074] The present invention relates to a novel multi-segment straight-line parallel scanning imaging system for CT detection of battery packs. This system aims to solve the problems of field-of-view limitation and truncation artifacts existing in traditional circular trajectory scanning, and can provide more comprehensive and accurate internal structure information, especially for objects with irregular shapes and large sizes (such as power battery packs).

[0075] As Figure 2 shown, a schematic diagram of the application scenario of the above method is given. The system consists of the following parts:

[0076] X-ray source: used to generate X-rays to penetrate the battery pack to be detected.

[0077] Detector: receives the X-rays after penetrating the battery pack and converts them into electrical signals for recording.

[0078] X-ray source moving axis S x , S y : allows the X-ray source to move independently in the X and Y directions to cover the entire detection area.

[0079] Detector moving axis D x , D y : enables the detector to move in the X and Y directions to ensure precise correspondence with the X-ray source and data collection.

[0080] The following gives a specific implementation process of this method:

[0081] In S1 to confirm the segmented scanning area, first, determine the basic geometric dimensions of the battery pack to be detected. Let the actual length of the battery pack be L, the width be W, and the thickness be T. Set the maximum imaging area size that can be covered by a single scan. Assume that the length of the single-scan imaging area in the X direction is L x , and the width in the Y direction is W y . To ensure the continuity and consistency of information between adjacent scanning areas, an overlapping area needs to be set. Assume that the overlapping length in the X direction is o x , and the overlapping width in the Y direction is o y .

[0082] According to the above parameters, the specific positions of each scanning area can be automatically calculated. Here, we define the lower-left corner coordinates of a scanning area as (x i , y j ), where i and j represent the indices in the X and Y directions respectively. In the X direction, the starting point x0 of the first scanning area is 0. The starting points of subsequent scanning areas can be obtained through a recurrence formula: xi+1 = x i + L x - o x 。Similarly, in the Y direction, the starting point y0 of the first scan area is 0. The starting points of subsequent scan areas can be obtained through the recurrence formula: y j+1 = y j + w y - o y 。By traversing all possible values of i and j until x i + L x > L or y j + W y > W. The calculated area structure is as Figure 3 shown.

[0083] In the S2 local area motion data acquisition, the local scan area is as Figure 4 shown. Define the following parameters:

[0084] odd: The distance from the detector to the sample center.

[0085] sod: The distance from the radiation source to the sample center.

[0086] W i : The width of the scan area, where W i = L x 。

[0087] H i : The height of the scan area, where H i = W y 。

[0088] o = (x0, y0): The coordinates of the center point of the scan area, where

[0089] n: The sampling points of the local scan area.

[0090] Then the movement trajectories of the detector and the radiation source are as follows:

[0091] Detector initial position:

[0092] Radiation source initial position:

[0093] For each movement, the radiation source and the detector only move diagonally in the X direction until the entire scan area is covered. Let the current step number be k, then:

[0094] Radiation source position:

[0095] X direction:

[0096] Y direction: always remains as y0

[0097] Center position of the detector:

[0098] X direction:

[0099] Y direction: also remains as y0

[0100] In the local data reconstruction of S3, the simultaneous algebraic reconstruction technique (SART) is used to reconstruct the data.

[0101] 1. According to the motion trajectory, calculate the angle between the ray source detector and the plane at each position:

[0102]

[0103] 2. According to the motion trajectory, first place no battery pack and collect the empty-field data I at each angle 0θ .

[0104] 3. Then place the battery pack body on the sample stage and collect the data I at each moving position θ .

[0105] 4. Calculate the projection data P θ .

[0106]

[0107] where the pixel value of each point corresponding in P θ is P θ (i,j), where i and j are the detector pixel coordinates respectively.

[0108] 5. Set X = (X1, X2,..., X N ) as the reconstructed object, where N is the total number of voxels. Initialize all elements of X to 0, and denote X n as the pixel value of the nth pixel in X.

[0109] 6. At each angle θ i , denote the straight line from the ray source to the point (i,j) of the detector as l(q), the total number of straight lines as Q, and calculate the system matrix W θ (q,n), where:

[0110]

[0111] l(q,n) is the intersection line length of l and X n .

[0112] 7. Calculate the residual data at each angle θ i :

[0113]

[0114] 8. For each angle θ i calculate and update X as follows:

[0115]

[0116] 9. After traversing and updating all angles θ i the final reconstructed result X is obtained. Figure 5 The reconstructed result diagram of the local scan is shown.

[0117] In the S4 area splicing, after the reconstruction of each segment area scan is completed, the coordinates of each area are spliced. Figure 6 The splicing result diagram is shown. Let I ij (x, y) represent the pixel value at the (x, y) coordinate in the scan area of the i-th row and j-th column.

[0118] For non-overlapping parts, if a pixel point is not in any overlapping area, the pixel value of this point is directly used:

[0119] P(x, y) = I ij (x - x i , y - y j ) Here, P(x, y) represents the pixel value in the final spliced image, and I ij is the image matrix corresponding to a specific scan area.

[0120] For the overlapping part in the X direction, if a pixel is in the overlapping area in the X direction, its position can be expressed as: the position in the left scan area A is (x A , y), where the range of x A is [x i + W i - o x , x i + W i . The position in the adjacent right scan area B is (x B , y), where the range of x B is [x i+1 , x i+1 + o x . At this time, these two positions are actually the same physical position, so the average value needs to be taken to calculate the final pixel value:

[0121]

[0122] For the overlapping part in the Y direction, if a pixel is in the overlapping area in the Y direction, its position can be expressed as: the position in the lower scan area C is (x, y C), where y C ranges from [y j +H j -o y , y j +H j .

[0123] The position in the adjacent upper scanning area D is (x, y D ), where y D ranges from [y j+1 , y j+1 +o y . It is necessary to average the pixel values of these overlapping areas:

[0124]

[0125] For the overlapping parts at the four corners, if a pixel is simultaneously within the overlapping areas in both the X and Y directions, then it will be involved in four adjacent scanning areas. In this case, the same averaging method is adopted:

[0126]

[0127] In an alternative embodiment of the present application, in the segmented scanning imaging method of the present invention, the scanning trajectory of each segment can be equally spaced and replaced by equal angles θ. The local reconstruction method SART can be replaced by ART, SIRT, CGLS. This scanning method is applicable not only to battery packs but also to the imaging of other large irregularly shaped objects.

[0128] In summary, it can be seen that by adopting the method of multi-segment straight parallel scanning, the present invention can effectively avoid the truncation artifacts caused by the field of view limitation in traditional circular trajectory scanning. This makes the reconstructed image clearer and more accurate, reducing edge blurring and distortion. For large or irregularly shaped objects (such as large battery packs), it is difficult for traditional circular trajectory scanning methods to fully cover all their areas. However, the method of the present invention allows the X-ray source and detector to move along multiple parallel planes, thereby being able to comprehensively capture all details of the object to be detected and ensuring the integrity of information. By performing multiple scans at different angles and planes, more abundant projection data can be obtained. After being processed by an optimized algorithm, these data can better reflect the true internal structure of the object, improving the utilization efficiency of the data. Since there is no need to be limited to a fixed circular trajectory, the system of the present invention can flexibly adjust the scanning path and plane position according to the actual situation. This provides greater convenience and adaptability for dealing with various different application scenarios.

[0129] An imaging device based on battery pack detection provided by an embodiment of the present application may include:

[0130] An area confirmation module, configured to determine the basic geometric dimensions of the battery pack to be detected, and divide a plurality of local scanning areas according to the geometric dimensions of the battery pack to be detected and the size of the single-scan imaging area;

[0131] A data acquisition module, configured to perform motion acquisition data on each local scanning area, and obtain projection data through the moving trajectories of the ray source and the detector;

[0132] A data reconstruction module, configured to reconstruct the projection data of each local scanning area to obtain a three-dimensional image of the local area;

[0133] An area stitching module, configured to stitch the three-dimensional images of the local areas to form a complete three-dimensional image of the battery pack body.

[0134] For the specific limitations of the imaging device based on battery pack body detection, reference can be made to the limitations of the imaging method based on battery pack body detection in the above text, which will not be elaborated here. Each module in the above imaging device based on battery pack body detection can be implemented in whole or in part by software, hardware and their combination. The above modules can be embedded in the processor of the computer device in hardware form or independent of it, or stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to the above modules.

[0135] The technical features of the above-described embodiments can be combined arbitrarily. For the sake of brevity of description, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, it should be considered as the scope described in this specification.

[0136] The above-described embodiments only represent several implementation manners of the present application, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the patent application. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present application, several modifications and improvements can still be made, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent of the present application should be subject to the appended claims.

Claims

1. An imaging method based on battery pack body detection, characterized in that, The method includes: S1 Confirm segmented scanning areas: Determine the basic geometric dimensions of the battery pack to be detected, and divide multiple local scanning areas according to the geometric dimensions of the battery pack to be detected and the size of the single-scan imaging area; S2 Motion data acquisition for local areas: Perform motion data acquisition for each local scanning area, and obtain projection data through the moving trajectories of the ray source and the detector; S3 Local data reconstruction: Reconstruct the projection data of each local scanning area to obtain a three-dimensional image of the local area; S4 Area stitching: Stitch the three-dimensional images of each local area to form a complete three-dimensional image of the battery pack body.

2. The imaging method according to claim 1, characterized in that, The steps of confirming segmented scanning areas specifically include: Determine the length, width, and thickness of the battery pack to be detected; Set the size of the single-scan imaging area, including the length in the X direction and the width in the Y direction; Set the size of the overlapping area between adjacent scanning areas, including the overlapping length in the X direction and the overlapping width in the Y direction; According to the set parameters, calculate the specific positions of each scanning area to ensure that there is an overlapping part between adjacent scanning areas to guarantee the continuity of information.

3. The imaging method according to claim 1, characterized in that The steps of motion data acquisition for local areas specifically include: Define the distance from the detector to the sample center and the distance from the ray source to the sample center; Determine the moving trajectories of the ray source and the detector, so that the ray source and the detector move diagonally in the X direction to cover the entire local scanning area while keeping the position in the Y direction unchanged.

4. The imaging method according to claim 1, wherein In the steps of local data reconstruction, the jointly algebraic reconstruction algorithm is used to reconstruct the acquired projection data, including: Calculate the angle between the ray source detector and the plane at each position. Without placing the battery pack, acquire the empty-field data at each angle. Place the battery pack on the sample stage and acquire the projection data at each moving position; Set the initial state of the reconstructed object, and set the initial value of all voxels to 0; At each angle, calculate the system matrix, and the elements of this matrix represent the intersection line length between the ray and the voxel; Calculate the residual data at each angle and perform iterative updates until the updates of all angles are completed to obtain the final three-dimensional image of the local area.

5. The imaging method according to claim 4, wherein Calculating the angle between the ray source detector and the plane at each position specifically includes according to the formula: Calculate the angle θ between the ray source detector and the plane at each position, where odd represents the distance from the detector to the sample center, sod represents the distance from the ray source to the sample center, and x s (k) represents the position of the ray source in the X direction, and x d (k) represents the center position of the detector in the X direction.

6. The imaging method according to claim 4, wherein Acquiring the empty-field data at each angle. Placing the battery pack on the sample stage and acquiring the projection data at each moving position specifically includes according to the formula: Calculate the projection data P θ , where I 0θ represents the open-field data at each angle, and I θ represents the projection data at each moving position.

7. The imaging method according to claim 4, characterized in that Calculating the residual data at each angle and performing iterative updates specifically includes through the formula: Calculate each angle θ i of the residual data, where the pixel value of each corresponding point in P θ is P θ (i, j), where i and j are the detector pixel coordinates respectively. Set X = (X1, X2,..., X N ) as the reconstructed object, N is the total number of voxels, initialize all elements of X to 0, and W θ (q, n) is the system matrix; Each angle θ i Calculate and update X as follows: After traversing and updating all angles θ i Finally, the final reconstruction result X is obtained.

8. The imaging method according to claim 1, characterized in that, The steps of area stitching include: For the pixel points in the non-overlapping part, directly use the pixel values of the pixel points in the corresponding local area image of the non-overlapping part; For the overlapping part in the X direction, take the average value of the pixel values at the corresponding positions in the left and right adjacent scanning areas as the final pixel value; For the overlapping part in the Y direction, take the average value of the pixel values at the corresponding positions in the upper and lower adjacent scanning areas; For the overlapping part at the four corners, which involves four adjacent scanning areas, take the average value of the pixel values at the corresponding positions in these four areas.

9. The imaging method according to claim 1, wherein The method further includes: Replace the scanning trajectory of each segment in an equally spaced or equally angled manner; and local data reconstruction is performed by algebraic reconstruction technique, simultaneous iterative reconstruction technique or conjugate gradient least squares method.

10. An imaging device based on battery pack body detection, characterized in that, The device includes: a region confirmation module, configured to determine the basic geometric dimensions of the battery pack to be detected, and divide a plurality of local scanning regions according to the geometric dimensions of the battery pack to be detected and the size of a single scan imaging region; a data acquisition module, configured to perform motion acquisition of data for each local scanning region, and obtain projection data through the movement trajectories of a ray source and a detector; a data reconstruction module, configured to reconstruct the projection data of each local scanning region to obtain a three-dimensional image of the local region; a region splicing module, configured to splice the three-dimensional images of the local regions to form a complete three-dimensional image of the battery pack body.

Images