Geometric correction method and device of three-dimensional image, electronic equipment and storage medium

Abstract

This invention discloses a geometric correction method, apparatus, electronic device, and storage medium for three-dimensional images, belonging to the technical field of image processing. The method includes: sequentially adjusting the position of a flat panel detector according to preset geometric parameters to obtain multiple image reconstruction layers of a preprocessed three-dimensional image; determining target geometric parameters based on the artifact intensities corresponding to the multiple image reconstruction layers; and correcting the preprocessed three-dimensional image using the target geometric parameters to obtain a corrected three-dimensional image. This solution adjusts the position of the flat panel detector using preset geometric parameters, reconstructs the image reconstruction layers of the preprocessed three-dimensional image online in real time, determines the optimal target geometric parameters based on the multiple image reconstruction layers, and uses the target geometric parameters to correct the preprocessed three-dimensional image. Compared with related technologies, this invention eliminates artifacts in the image and avoids deformation through geometric correction without additional physical correction and scanning, thereby significantly improving the quality of the three-dimensional image.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and more particularly to a geometric correction method, apparatus, electronic device, and storage medium for three-dimensional images. Background Technology

[0002] DR (Digital Radiography) refers to the process of converting X-ray information penetrating the human body into digital signals using an X-ray detector under computer control, followed by post-processing and display of the images by the computer. Because DR utilizes digital technology, it can perform various image post-processing techniques according to clinical needs. Furthermore, DR is widely used in physical examinations and medical imaging diagnostics due to its advantages such as low dose, high spatial resolution, short scan time, and low cost. However, DR equipment generally only acquires two-dimensional images.

[0003] To obtain three-dimensional images, related technologies use idealized geometric structures in conjunction with DR equipment. However, the resulting three-dimensional images exhibit severe reconstruction artifacts, affecting diagnosis. Alternatively, offline geometric correction using CT / CBCT can lead to varying degrees of deformation due to the instability of the geometric structure.

[0004] Therefore, a new geometric correction technique for three-dimensional images is needed. Summary of the Invention

[0005] This application provides a geometric correction method, apparatus, electronic device, and storage medium for three-dimensional images, which can solve the technical problems of severe reconstruction artifacts and deformation in three-dimensional images.

[0006] The first aspect of this invention provides a geometric correction method for a three-dimensional image, the method comprising:

[0007] The position of the flat panel detector is adjusted step by step according to the preset geometric parameters to obtain multiple image reconstruction layers of the preprocessed three-dimensional image;

[0008] The target geometric parameters are determined based on the artifact intensities corresponding to the multiple image reconstruction layers;

[0009] The preprocessed 3D image is corrected using the target geometric parameters to obtain the corrected 3D image.

[0010] Optionally, the step of sequentially adjusting the position of the flat panel detector according to preset geometric parameters to obtain multiple image reconstruction layers of the reconstructed 3D image includes:

[0011] Receive X-ray scan data of the object to be detected at different angles on the flat panel detector;

[0012] The X-ray scanning data is sequentially subjected to dark correction, bad pixel / bad line correction and air correction to obtain pre-processed scanning data.

[0013] The pre-corrected scan data is subjected to cone angle weighting and filtering three-dimensional back projection processing to obtain the three-dimensional image.

[0014] Optionally, the geometric parameters include a range of geometric parameter values ​​and a unit geometric parameter. Then, the step of sequentially adjusting the position of the flat panel detector according to the preset geometric parameters to obtain multiple image reconstruction layers of the reconstructed 3D image includes:

[0015] Obtain the preset range of the geometric parameters and the unit geometric parameters;

[0016] Compare the geometric values ​​corresponding to the position of the flat panel detector with the range of geometric parameters;

[0017] If the geometric value corresponding to the position of the flat panel detector is less than the range of geometric parameters, the position of the flat panel detector is adjusted step by step according to the preset unit geometric parameters.

[0018] After adjusting the position of the flat panel detector, the target image layers of the three-dimensional image are reconstructed sequentially to obtain multiple image reconstruction layers.

[0019] Optionally, the step of determining the target geometric parameters based on the artifact intensities corresponding to the plurality of image reconstruction layers includes:

[0020] Extract the image features corresponding to the multiple image reconstruction layers respectively;

[0021] Based on multiple image features, the image reconstruction layer with low artifact intensity is determined as the target image reconstruction layer;

[0022] Obtain the target geometric parameters corresponding to the target image reconstruction layer.

[0023] Optionally, the step of determining the image reconstruction layer with low artifact intensity as the target image reconstruction layer based on multiple image features includes:

[0024] The image information entropy of each of the aforementioned image features is calculated, and the formula for calculating the image information entropy is as follows:

[0025]

[0026] P i,j =f(i,j) / N 2 ,

[0027] Where i represents the pixel grayscale value of the image feature, j represents the neighborhood grayscale value of the image feature, f(i,j) represents the total number of pixels of the image feature, N represents the side length of the image feature, and P i,j This indicates the probability that the image feature contains a pixel;

[0028] Obtain the largest image information entropy among the multiple image information entropies;

[0029] The image reconstruction layer corresponding to the largest image information entropy is determined as the target image reconstruction layer with low artifact intensity.

[0030] Optionally, the step of correcting the preprocessed 3D image using the target geometric parameters to obtain the corrected 3D image includes:

[0031] The pixel indices of the three-dimensional image are calculated using the target geometric parameters;

[0032] The preprocessed 3D image is then subjected to weighted back projection using the pixel index to obtain the corrected 3D image.

[0033] A second aspect of the present invention provides a geometric correction apparatus for a three-dimensional image, the apparatus comprising:

[0034] The adjustment module is used to sequentially adjust the position of the flat panel detector according to preset geometric parameters to obtain multiple image reconstruction layers of the preprocessed 3D image;

[0035] The determination module is used to determine the target geometric parameters based on the artifact intensities corresponding to the plurality of image reconstruction layers;

[0036] The correction module is used to correct the preprocessed 3D image using the target geometric parameters to obtain the corrected 3D image.

[0037] Optionally, the device further includes:

[0038] The receiving module is used to receive X-ray scanning data of the object to be detected at different angles on the flat panel detector;

[0039] The first processing module is used to sequentially perform dark correction processing, bad spot and bad line correction processing and air correction processing on the X-ray scanning data to obtain pre-processed scanning data.

[0040] The second processing module is used to perform cone angle weighting and filtering three-dimensional back projection processing on the pre-corrected scan data to obtain the three-dimensional image.

[0041] A third aspect of the present invention provides an electronic device, comprising: a memory, a processor, and a communication bus, wherein the communication bus is communicatively connected to the memory and the processor, the memory is coupled to the processor, the memory stores a computer program, and when the processor executes the computer program, it implements the various steps of the geometric correction method for three-dimensional images in the first aspect.

[0042] A fourth aspect of the present invention provides a storage medium, which is a computer-readable storage medium, and a computer program is stored on the computer-readable storage medium. When the computer program is executed by a processor, it implements the various steps of the geometric correction method for three-dimensional images in the first aspect.

[0043] The geometric correction method for three-dimensional images provided by this invention includes: sequentially adjusting the position of a flat panel detector according to preset geometric parameters to obtain multiple image reconstruction layers of a preprocessed three-dimensional image; determining target geometric parameters based on the artifact intensities corresponding to the multiple image reconstruction layers; and correcting the preprocessed three-dimensional image using the target geometric parameters to obtain a corrected three-dimensional image. This method adjusts the position of the flat panel detector using preset geometric parameters, reconstructs multiple image reconstruction layers corresponding to the preprocessed three-dimensional image online in real time, determines the optimal target geometric parameters based on the multiple reconstructed image reconstruction layers, and uses these target geometric parameters to correct the preprocessed three-dimensional image. Compared with related technologies, this invention eliminates artifacts in the image online in real time through geometric correction without additional physical correction or scanning, avoiding deformation and thus significantly improving the quality of the three-dimensional image, demonstrating reliability. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0045] Figure 1 A DR system architecture diagram provided for embodiments of the present invention;

[0046] Figure 2 A schematic diagram of the external rotation angle and internal rotation angle of the flat panel detector provided in this embodiment of the invention;

[0047] Figure 3 A flowchart illustrating the steps of a geometric correction method for three-dimensional images provided in an embodiment of the present invention;

[0048] Figure 4a The three-dimensional image reconstructed when the external rotation angle is accurate, as provided in the embodiments of the present invention;

[0049] Figure 4b This is a reconstructed three-dimensional image provided in an embodiment of the present invention when the external rotation angle is inaccurate;

[0050] Figure 5a A cross-sectional view of the geometric structure provided in an embodiment of the present invention;

[0051] Figure 5b A three-dimensional diagram of the geometric structure provided in the embodiments of the present invention;

[0052] Figure 6 A flowchart illustrating another step of the geometric correction method for three-dimensional images provided in this embodiment of the invention;

[0053] Figure 7 A block diagram of a geometric correction device for three-dimensional images provided in an embodiment of the present invention;

[0054] Figure 8 This is an architectural diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0055] To make the objectives, features, and advantages of this invention more apparent and understandable, the technical solutions of the embodiments of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0056] Due to the serious reconstruction artifacts and deformation problems in existing 3D images.

[0057] To address the aforementioned technical problems, this invention proposes a geometric correction method, apparatus, electronic device, and storage medium for three-dimensional images.

[0058] DR (Digital Radiography) is a digital X-ray imaging technology where X-ray detectors convert the X-ray information penetrating the human body into digital signals, which are then processed and displayed by a computer. DR breaks away from traditional X-ray imaging, realizing the transformation from analog to digital X-ray images. Because DR uses digital technology, it can perform various image post-processing functions according to clinical needs, such as automatic image processing, edge enhancement, magnification and panning, image stitching, region of interest window width and level adjustment, and distance, area, and density measurements. DR technology has a wide dynamic range, high X-ray quantum detection efficiency (DQE), and a wide exposure latitude, allowing for excellent image quality even under slightly poor exposure conditions. Furthermore, DR's low dose, high spatial resolution, short scan time, and low cost make it widely used in physical examinations and medical imaging diagnostics, and it is one of the main pieces of equipment in medical imaging diagnostics.

[0059] DR (Digital Radiography) equipment is primarily used to acquire two-dimensional images, but its geometric accuracy and stability are relatively poor. This differs significantly from the design requirements of CT / CBCT (Computed Tomography-Conducting Radiography). Three-dimensional images reconstructed from DR scan data often exhibit severe artifacts. Furthermore, CT / CBCT relies on offline correction, which, due to the instability of the geometric structure, can lead to varying degrees of deformation, hindering the correction or reconstruction of three-dimensional images. To address this, the present invention provides a geometric correction method, apparatus, device, and storage medium for three-dimensional images, enabling online real-time geometric correction and resolving the issues of severe reconstruction artifacts and deformations in three-dimensional images.

[0060] Please see Figure 1 This is a diagram illustrating the architecture of a DR system provided in an embodiment of the present invention. The DR system mainly includes: a 3D image reconstruction terminal 10, a high-voltage generator 20, an X-ray emitter 30, a rotating mechanism 40, an X-ray detector (flat panel detector) 50, a frame 60, and a simulated object under test 70. The high-voltage generator 20 is electrically connected to the X-ray emitter 30. The rotating mechanism 40 has angle measurement and feedback functions and is used to place the object under test 70. The rotating mechanism 40 is positioned between the X-ray emitter 30 and the X-ray detector 50. The X-ray emitter 30 and the X-ray detector 50 are respectively fixedly mounted on the frame 60. The 3D image reconstruction terminal 10 is electrically connected to the X-ray detector 50. It should be noted that the feedback function of the rotating mechanism 40 means that the rotating mechanism 40 can record and transmit the corresponding angle information to the 3D image reconstruction terminal 10.

[0061] Please see Figure 2This diagram illustrates the external and internal rotation angles of the flat panel detector provided in this embodiment of the invention. Specifically, before geometrically correcting the 3D image, this embodiment establishes a 3D coordinate system on the side of the flat panel detector. The xoy rectangular coordinate system is a 3D coordinate system perpendicular to the side of the flat panel detector; specifically, the x-axis is perpendicular to the side of the flat panel detector, meaning the x-axis aligns with the direction of X-rays. Furthermore, the zoy coordinate axes are parallel to the side of the flat panel detector; specifically, the y-axis and z-axis are parallel to the side of the flat panel detector. Further, the geometric parameters involved in the geometric correction include external and internal rotation angles. The external rotation angle includes the angle φ of the flat panel detector's offset along the y-axis in the xoy rectangular coordinate system and the angle σ of the flat panel detector's offset along the z-axis in the xoz rectangular coordinate system. The internal rotation angle includes the angle η of the flat panel detector's offset along the z-axis and y-axis in the yoz rectangular coordinate system. Systematic geometric deviations (detector internal and external rotation angles, etc.) have a significant impact on the image reconstruction effect, with the external rotation angle having a greater impact on the reconstruction or correction of the 3D image. The geometric correction method for three-dimensional images provided in this invention mainly involves online real-time evaluation of the system's external rotation angle, and applying the estimated geometric parameters such as the system's external rotation angle during the reconstruction or correction process to achieve online real-time geometric correction of the three-dimensional image.

[0062] Please see Figure 3 This is a flowchart illustrating the steps of a geometric correction method for a three-dimensional image provided in an embodiment of the present invention. The embodiment of the present invention provides a geometric correction method for a three-dimensional image, which includes the following steps:

[0063] Step S301: Adjust the position of the flat panel detector step by step according to the preset geometric parameters to obtain multiple image reconstruction layers of the preprocessed three-dimensional image.

[0064] In this step, the preprocessed 3D image refers to the projected image that has undergone image correction and filtering in standard FBP (Filtered Backprojection). Specifically, the preprocessed 3D image is used as the input for the reconstructed corrected backprojection. For example, exemplarily, the pixel or resolution size of the 3D image in this system is 1024x1024, and 404 frames are acquired for the object to be detected. The preprocessing process includes, but is not limited to, dark correction, bad pixel / line correction, atmospheric correction, cone angle weighting, and filtered 3D backprojection processing.

[0065] The target layer in the preprocessed 3D image is reconstructed using preset set parameters. This step involves reconstructing the same target layer in the 3D image based on multiple preset geometric parameters or a unit geometric parameter within a preset range of geometric parameter values. Specifically, the 3D image may contain multiple layers. Preferably, the intermediate layer of the 3D image is reconstructed using preset geometric parameters. Since the region of interest is generally located in the center of the 3D image, a clearer region of interest can be obtained by reconstructing the intermediate layer. These geometric parameters are not limited to external rotation angle and internal rotation angle. Taking the external rotation angle as an example, the preset external rotation angle can be 0.1°, or within a preset external rotation angle range of -3° to +3°, with a step external rotation angle (unit external rotation angle) of 0.1°. The position of the flat panel detector is adjusted according to this external rotation angle or the step external rotation angle, such as adjusting the external rotation of the flat panel detector by 0.1 degrees in the forward or reverse direction. After adjustment, the target layer of the 3D image is reconstructed, such as the reconstruction of the intermediate layer of the 3D image. It should be noted that, with the xoy coordinate axis as a reference, the clockwise direction of the flat panel detector on the xoy coordinate axis is the positive direction, and the counterclockwise rotation direction is the negative direction. The external rotation angle of the flat panel detector in both directions should fall within the external rotation angle range [-3°, 3°], as this angle range is more conducive to obtaining a clearer target layer. Furthermore, within the preset external rotation angle range, the external rotation angle of the flat panel detector is adjusted step-by-step according to a preset unit external rotation angle (0.1°). After each adjustment of the flat panel detector, the intermediate layer (target layer) of the 3D image is reconstructed or corrected, resulting in the intermediate layer (target layer) of the reconstructed 3D image corresponding to each adjustment of the flat panel detector, thus obtaining multiple set parameters. This embodiment is not limited to the target layer of the reconstructed 3D image being the intermediate layer; it can also be other layers. The reconstruction process of other layers is consistent with the reconstruction method of this embodiment, and this embodiment will not elaborate further on this.

[0066] It should be noted that reconstructing a 3D image using different geometric parameters will yield different results. Using incorrect geometric parameters will result in noticeable artifacts in the reconstructed image, as shown in Figure 4. Figure 4a and Figure 4b , Figure 4a This is a reconstructed three-dimensional image provided by an embodiment of the present invention when the external rotation angle is accurate. Figure 4b The three-dimensional image reconstructed when the external rotation angle is inaccurate, as provided in the embodiments of the present invention, may have artifacts near the phalanx; in the medical field, this manifests as bone edge fracture and edge blurring, and the presence of object information in other layers can lead to image disorder.

[0067] In one embodiment, step S101 includes: obtaining a preset geometric parameter range and a unit geometric parameter; comparing the geometric value corresponding to the position of the flat panel detector with the geometric parameter range; if the geometric value corresponding to the position of the flat panel detector is less than the geometric parameter range, then adjusting the position of the flat panel detector step by step according to the preset unit geometric parameter; after adjusting the position of the flat panel detector, reconstructing the target image layer of the three-dimensional image in sequence to obtain multiple image reconstruction layers.

[0068] Specifically, the preset geometric parameters can include a geometric parameter range and a unit geometric parameter. The unit geometric parameter is the unit value used to adjust the rotation or displacement of the flat panel detector, such as the unit outward rotation angle of the flat panel detector. After obtaining the preset set parameter range and unit geometric parameter, the geometric value corresponding to the position of the flat panel detector is compared with the preset geometric parameter range. For example, the current position information or physical position information of the flat panel detector is obtained, and the corresponding geometric value is generated based on the position information. By comparing this geometric value with the preset geometric parameter range, it is determined whether the position of the flat panel detector is within the position range defined by the preset geometric parameter range. If the geometric value corresponding to the position of the flat panel detector is less than the boundary value of the geometric parameter range, or if the geometric value belongs to the geometric parameter range, then the position of the flat panel detector is determined to be within the position range defined by the preset geometric parameter range. Furthermore, the position of the flat panel detector is adjusted according to the preset unit geometric parameter, and a target layer in three dimensions is reconstructed as required, such as the intermediate layer of a three-dimensional image. By setting the range and unit geometric parameters, the target layer of the 3D image can be reconstructed according to the preset range and unit geometric parameters. This reduces the amount of computation, speeds up the acquisition of multiple image reconstruction layers, and improves the efficiency of the reconstruction layer process by using an iterative correction method based on the image domain.

[0069] Step S302: Determine the target geometric parameters based on the artifact intensities corresponding to multiple image reconstruction layers.

[0070] After obtaining multiple image reconstruction layers corresponding to the flat panel detector under different geometric parameters, feature information of each image reconstruction layer is extracted to characterize the artifact intensity of each image reconstruction layer. During feature extraction, the methods that can be used include typical feature extraction methods such as gradient information extraction and edge information extraction. Image entropy methods or skin texture and bone texture detection methods can also be used for feature extraction. Based on the extracted feature information, the artifact intensity of each image reconstruction layer is determined. The geometric parameters corresponding to the image reconstruction layer with low artifact intensity are obtained as target geometric parameters. These target geometric parameters are the optimal geometric parameters. The preprocessed 3D image is then corrected using these geometric parameters to eliminate artifacts, deformations, or disturbances in the 3D image.

[0071] In one embodiment of this step, the steps include: extracting image features corresponding to multiple image reconstruction layers respectively; determining the image reconstruction layer with low artifact intensity as the target image reconstruction layer based on the multiple image features; and obtaining the target geometric parameters corresponding to the target image reconstruction layer.

[0072] Specifically, after obtaining multiple image reconstruction layers, image features corresponding to each layer are extracted. These features contain image information, such as pixel size or pixel density. Based on these features, the image reconstruction layer with low artifact intensity is determined as the target image reconstruction layer. The process for determining the image reconstruction layer with low artifact intensity is as follows:

[0073] In one embodiment, the method includes: after extracting image features corresponding to multiple image reconstruction layers, calculating the image information entropy of each of the multiple image features, wherein the formula for calculating the image information entropy is as follows:

[0074]

[0075] P i,j =f(i,j) / N 2 ,

[0076] Where i represents the pixel grayscale value of the image feature, j represents the neighborhood grayscale value of the image feature, f(i,j) represents the total number of pixels of the image feature, N represents the side length of the image feature, and P i,j This indicates the probability that an image feature contains a pixel.

[0077] Furthermore, the maximum image information entropy among multiple image information entropies is obtained. Specifically, since image information entropy reflects the amount of information contained in an image, the greater the amount of information, the greater the information entropy, and the less image artifact intensity. Therefore, the image information entropy of all obtained image reconstruction layers is calculated by comparing the image information entropies corresponding to multiple image features, i.e., comparing the calculated image information entropies, and the maximum image information entropy is selected.

[0078] Furthermore, the image reconstruction layer corresponding to the largest image information entropy is determined as the target image reconstruction layer with low artifact intensity; the target geometric parameters corresponding to the target image reconstruction layer are obtained, and the target set geometric parameters are the geometric parameters corresponding to the position of the flat panel detector.

[0079] Step S303: Correct the preprocessed 3D image using the target geometric parameters to obtain the corrected 3D image.

[0080] After obtaining the optimal target geometric parameters, the preprocessed 3D image is corrected using these parameters to obtain a 3D image free of artifacts. The online real-time geometric correction process is as follows:

[0081] In one embodiment, this step includes: calculating the pixel index of the three-dimensional image using the target geometric parameters.

[0082] Specifically, the geometric structure of the three-dimensional image differs from that of a typical CBCT scan in that, before the image reconstruction layer of the three-dimensional image is reconstructed, i.e., before the three-dimensional image is established, the flat panel detector remains stationary when acquiring X-ray scan data. The object to be detected is rotated by rotating the rotating structure 40. The clockwise direction in the xoy coordinate system of the aforementioned three-dimensional coordinate system is taken as the positive direction of rotation, as shown in Figure 5. Figure 5a and Figure 5b , Figure 5a A cross-sectional view (top view) of the geometric structure provided in an embodiment of the present invention. Figure 5b This is a three-dimensional diagram of the geometric structure provided in an embodiment of the present invention; assuming the initial position of a point on the object to be detected is (x0, y0, z0), when the object to be detected rotates clockwise by an angle α, the position a1 in the three-dimensional coordinate system is (x1, y1, z1), where the coordinate relationship between the initial position and the position a1 is:

[0083]

[0084] z1 = z0;

[0085] Where Rot is the rotation matrix, expressed as:

[0086]

[0087] In the xoz coordinate plane of the three-dimensional coordinate system

[0088]

[0089]

[0090] Where η represents the inward rotation angle of the system, that is, the angle by which the flat panel detector rotates along the x-axis in the yoz coordinate plane; φ and σ represent the outward rotation angle of the system, where φ represents the angle by which the flat panel detector rotates along the z-axis in the xoy coordinate plane, and σ represents the angle by which the flat panel detector rotates along the y-axis in the xoz coordinate plane; z1=z0, U=R+y1.

[0091] On the plane of the flat panel detector facing the rotating mechanism (yoz coordinate plane), see... Figure 5b The uov coordinate plane representation of the flat panel detector is as follows:

[0092]

[0093]

[0094] The projected coordinate system is then transformed into the pixel index of the flat panel detector, as follows:

[0095]

[0096]

[0097] Where Row represents the pixel index in the v direction, Chm represents the pixel index in the u direction, Rm represents the target image layer index, and in this embodiment, the target image layer is an intermediate layer, taking a flat panel detector as the plane and a perspective matrix as an example. Therefore, when calculating the intermediate layer index, the length of the flat panel detector is divided by 2, or the resolution in the length direction is divided by 2. Cm represents the layer channel index. In this embodiment, if the target image layer is an intermediate layer, taking a flat panel detector as the plane and a perspective matrix as an example, when calculating the layer channel index, the width of the flat panel detector is divided by 2, or the resolution in the width direction is divided by 2. Furthermore, D... v D represents the pixel size in the v direction. u This indicates the pixel size in the u direction.

[0098] The preprocessed 3D image is then subjected to weighted backprojection using pixel indexing to obtain the corrected 3D image. The calculation process for the weighted backprojection is as follows:

[0099]

[0100] Where V(x0,y0,z0) Δ I(Row,Chm) represents the corrected time voxel, and I(Row,Chm) represents the pixel index on the plane of the flat panel detector.

[0101] It should be noted that in the process of online real-time geometric correction of 3D images, it is necessary to traverse all voxels in the 3D image and perform a weighted back projection. The voxels after weighted back projection are then loaded into the corresponding image positions to finally output the corrected 3D image within the field of view. Compared with the preprocessed 3D image, the corrected 3D image has eliminated artifacts, distortions, and blurring.

[0102] This invention provides a geometric correction method for three-dimensional images, comprising: sequentially adjusting the position of a flat panel detector according to preset geometric parameters to obtain multiple image reconstruction layers of a preprocessed three-dimensional image; determining target geometric parameters based on the artifact intensities corresponding to the multiple image reconstruction layers; and correcting the preprocessed three-dimensional image using the target geometric parameters to obtain a corrected three-dimensional image. This method adjusts the position of the flat panel detector using preset geometric parameters, reconstructs multiple image reconstruction layers corresponding to the preprocessed three-dimensional image online in real time, determines the optimal target geometric parameters based on the multiple reconstructed image reconstruction layers, and uses these target geometric parameters to correct the preprocessed three-dimensional image. Compared with related technologies, this invention eliminates artifacts in the image online in real time through geometric correction without performing additional physical correction and scanning, avoiding deformation and thus significantly improving the quality of the three-dimensional image, demonstrating reliability.

[0103] Please see Figure 6 This is a flowchart illustrating another step in the geometric correction method for three-dimensional images provided in this embodiment of the invention. Specifically, the process is as follows:

[0104] Step S601: Receive X-ray scan data of the object to be detected at different angles on the flat panel detector.

[0105] The object to be examined can be a person or animal, etc. It can stand or be fixed on a rotatable body equipped with angle measurement and transmission functions. By rotating the object with these functions, X-ray scan data at different angles can be obtained. The rotation speed can be adjusted as needed, and the object must remain stationary during rotation. The X-ray scan data acquired by the flat panel detector is used to create a three-dimensional image.

[0106] Step S602: Perform dark correction, bad pixel / bad line correction, and air correction on the X-ray scan data in sequence to obtain preprocessed scan data.

[0107] In this step, dark correction processing can avoid the problem of uneven data acquisition by the flat panel detector itself. Specifically, dark correction processing is to subtract the pixel value of the corresponding pixel in the dark correction image from the pixel value of each pixel in the received X-ray scan data.

[0108] Furthermore, the purpose of performing dead pixel and bad line correction is to remove dead pixels and bad lines present in the X-ray detector (flat panel detector). These dead pixels and bad lines are pixels with a value of 0 or other outliers in the image. These outliers are fixed values ​​that do not change with scanning conditions. Without correction, this can cause filtering abnormalities, resulting in no image or severe ring artifacts in the image. Specifically, the dead pixel and bad line correction process involves: setting a preset pixel range; determining whether a pixel in the dark-corrected scan data falls within the preset pixel range; if not, using the pixel value of a neighboring pixel within the preset pixel range as the pixel value of the first pixel; if yes, the pixel value of the first pixel remains unchanged.

[0109] Furthermore, the purpose of air correction processing is to correct the inhomogeneity and afterimage problems of X-ray detector reception. This is achieved by calculating the incident X-ray intensity and performing a ln operation on the scan data to obtain the final air scan data usable for 3D reconstruction. This air scan data represents the data obtained when X-rays penetrate only the air and are received by the detector. Specifically, the air correction process involves: collecting air scan data under different voltage and current conditions to obtain an air correction table; calculating the incident light intensity under the preset voltage and current conditions based on the air correction table; performing a ln logarithmic operation on the incident light intensity, the scan data corrected for bad pixels and lines, and the air scan data to obtain the logarithmic incident light intensity, logarithmic scan data, and logarithmic air scan data; and subtracting the difference between the logarithmic scan data and the logarithmic air scan data from the logarithmic incident light intensity to obtain the corrected scan data. In this embodiment, the incident light intensity can be calculated as the average value of all pixels in the air correction table, or the average value of any one or more pixels can be selected as the incident light intensity.

[0110] After the above multiple different pre-correction processes, the pre-processed scan data is obtained.

[0111] Step S603: Perform cone angle weighting and filtering three-dimensional back projection processing on the pre-corrected scanning data to obtain the pre-processed three-dimensional image.

[0112] Specifically, the corrected scan data is weighted using a cone angle cosine weighting method. DR equipment's X-ray detectors have a large X-ray cone angle, which can cause artifacts due to a mismatch between the scan data and the actual X-rays. Therefore, it's necessary to reduce the impact of the cone angle on the scan data; a larger cone angle results in a lower weight, and a smaller cone angle results in a higher weight. Furthermore, linear or Gaussian weights can also be used to weight the scan data.

[0113] Furthermore, the filtered 3D backprojection processing specifically involves applying a filter to the cone-angle-weighted scan data for 3D backprojection. Since the backprojection operation requires reconstructing data from different projection angles into the 3D reconstructed image, this process enhances the low-frequency components in the reconstructed image, leading to blurring and unevenness. To suppress this phenomenon, a filter is needed to suppress the low-frequency regions of the image while enhancing high-frequency information. The scan data is filtered to enhance its boundary information. After the above preprocessing, the final 3D image is obtained.

[0114] Step S604: Determine the target step size of the target user's traffic data based on a preset total cost function using the remaining traffic volume, traffic rate, and target time.

[0115] Step S605: Determine the target geometric parameters based on the artifact intensities corresponding to multiple image reconstruction layers.

[0116] Step S606: Correct the preprocessed 3D image using the target geometric parameters to obtain the corrected 3D image.

[0117] Specifically, the method steps described in steps S604-S605 are similar to or close to the method steps in steps S301 to S303. The description of this part of the process is consistent with the description of steps S301 to S303, and this embodiment will not elaborate further on it.

[0118] Please see Figure 7 , Figure 7 This is a block diagram of a geometric correction device for a three-dimensional image according to an embodiment of the present invention. The geometric correction device corresponds to a processor executing a geometric correction method for a three-dimensional image. The device 700 includes:

[0119] The adjustment module 701 is used to sequentially adjust the position of the flat panel detector according to preset geometric parameters to obtain multiple image reconstruction layers of the preprocessed three-dimensional image.

[0120] The determination module 702 is used to determine the target geometric parameters based on the artifact intensities corresponding to multiple image reconstruction layers;

[0121] The correction module 703 is used to correct the preprocessed 3D image using the target geometric parameters to obtain the corrected 3D image.

[0122] Furthermore, the device 700 also includes:

[0123] The receiving module 704 is used to receive X-ray scanning data of the object to be detected at different angles on the flat panel detector;

[0124] The first processing module 705 is used to sequentially perform dark correction processing, bad spot and bad line correction processing and air correction processing on the X-ray scanning data to obtain pre-processed scanning data.

[0125] The second processing module 706 is used to perform cone angle weighting and filtering three-dimensional back projection processing on the pre-corrected scanning data to obtain a three-dimensional image.

[0126] This invention provides a geometric correction device for three-dimensional images, comprising: an adjustment module 701, a determination module 702, and a correction module 703. Specifically, the adjustment module 701 sequentially adjusts the position of a flat panel detector according to preset geometric parameters to obtain multiple image reconstruction layers of a preprocessed three-dimensional image. The determination module 702 determines target geometric parameters based on the artifact intensities corresponding to the multiple image reconstruction layers. The correction module 703 uses the target geometric parameters to correct the preprocessed three-dimensional image, obtaining a corrected three-dimensional image. This solution adjusts the position of the flat panel detector using preset geometric parameters, reconstructs multiple image reconstruction layers corresponding to the preprocessed three-dimensional image online in real time, determines the optimal target geometric parameters based on the multiple reconstructed image reconstruction layers, and uses these target geometric parameters to correct the preprocessed three-dimensional image. Compared with related technologies, this invention eliminates artifacts in the image online in real time through geometric correction without additional physical correction or scanning, avoiding deformation and thus significantly improving the quality of the three-dimensional image, demonstrating reliability.

[0127] It should be noted that the geometric correction device for three-dimensional images provided in this embodiment is the device item corresponding to the aforementioned geometric correction method for three-dimensional images. The technical features of this part are similar or close to the aforementioned method steps. For the description of the technical features of this device, please refer to the description of the geometric correction method for three-dimensional images in the aforementioned embodiment. This embodiment will not elaborate further on this.

[0128] This invention provides an electronic device; please refer to [link / reference]. Figure 8 The above is an architecture diagram of an electronic device provided in an embodiment of the present invention. The electronic device includes: a memory 801, a processor 802 and a communication bus 803. The communication bus 803 is communicatively connected to the memory 801 and the processor 802 respectively. The memory 801 is coupled to the processor 802. The memory 801 stores a computer program. When the processor 802 executes the computer program, it implements each step in the geometric correction method for three-dimensional images of any of the above-mentioned items.

[0129] For example, the computer program for the geometric correction method of a 3D image mainly includes: sequentially adjusting the position of the flat panel detector according to preset geometric parameters to obtain multiple image reconstruction layers of the preprocessed 3D image; determining target geometric parameters based on the artifact intensity corresponding to the multiple image reconstruction layers; and correcting the preprocessed 3D image using the target geometric parameters to obtain the corrected 3D image. Additionally, the computer program can be divided into one or more modules, one or more of which are stored in memory and executed by a processor to complete the invention. One or more modules can be a series of computer program instruction segments capable of performing specific functions, which describe the execution process of the computer program in a computing device. For example, the computer program can be divided into... Figure 7 The adjustment module 701, determination module 702, correction module 703, receiving module 704, first processing module 705, and second processing module 706 are shown.

[0130] The processor 802 can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor can be a microprocessor or any conventional processor.

[0131] The present invention also provides a storage medium, which is a computer-readable storage medium, and stores a computer program thereon. When the computer program is executed by a processor, it implements each step of the geometric correction method for the three-dimensional image described above.

[0132] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple modules or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or modules may be electrical, mechanical, or other forms.

[0133] The modules described as separate components may or may not be physically separate. Similarly, the components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0134] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The integrated modules described above can be implemented in hardware or as software functional modules.

[0135] If the integrated module is implemented as a software functional module and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0136] It should be noted that, for the sake of simplicity, the foregoing method embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, because according to the present invention, some steps can be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0137] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0138] The above is a description of a geometric correction method, apparatus, electronic device and storage medium for three-dimensional images provided by the present invention. For those skilled in the art, based on the ideas of the embodiments of the present invention, there will be changes in the specific implementation and application scope. Therefore, the content of this specification should not be construed as a limitation of the present invention.

Claims

1. A geometric correction method for a three-dimensional image, characterized in that, The method includes: The position of the flat panel detector is adjusted step by step according to preset geometric parameters to obtain multiple image reconstruction layers of the preprocessed three-dimensional image; wherein, the geometric parameters include external rotation angle and internal rotation angle; Extract the image features corresponding to the multiple image reconstruction layers respectively; The image information entropy of each of the aforementioned image features is calculated, and the formula for calculating the image information entropy is as follows: ， ， in, The pixel grayscale value representing the image feature. The grayscale value representing the neighborhood of the image feature. This represents the total number of pixels representing the image feature. The side length of the image feature is represented. This indicates the probability that the image feature contains a pixel; Obtain the largest image information entropy among the multiple image information entropies; The image reconstruction layer corresponding to the largest image information entropy is determined as the target image reconstruction layer with low artifact intensity; Obtain the target geometric parameters corresponding to the target image reconstruction layer; The preprocessed 3D image is corrected using the target geometric parameters to obtain the corrected 3D image.

2. The geometric correction method for three-dimensional images according to claim 1, characterized in that, The step of sequentially adjusting the position of the flat panel detector according to preset geometric parameters to obtain multiple image reconstruction layers of the preprocessed 3D image includes: Receive X-ray scan data of the object to be detected at different angles on the flat panel detector; The X-ray scanning data is sequentially subjected to dark correction, bad pixel / bad line correction and air correction to obtain pre-processed scanning data. The preprocessed scan data is subjected to cone angle weighting and filtering three-dimensional back projection processing to obtain the three-dimensional image.

3. The geometric correction method for three-dimensional images according to claim 1, characterized in that, The geometric parameters include a range of geometric parameter values ​​and a unit geometric parameter. The step of sequentially adjusting the position of the flat panel detector according to the preset geometric parameters to obtain multiple image reconstruction layers of the preprocessed 3D image includes: Obtain the preset range of the geometric parameters and the unit geometric parameters; Compare the geometric values ​​corresponding to the position of the flat panel detector with the range of geometric parameters; If the geometric value corresponding to the position of the flat panel detector is less than the range of geometric parameters, the position of the flat panel detector is adjusted step by step according to the preset unit geometric parameters. After adjusting the position of the flat panel detector, the target image layers of the three-dimensional image are reconstructed sequentially to obtain multiple image reconstruction layers.

4. The geometric correction method for three-dimensional images according to claim 1, characterized in that, The step of correcting the preprocessed 3D image using the target geometric parameters to obtain the corrected 3D image includes: The pixel indices of the three-dimensional image are calculated using the target geometric parameters; The preprocessed 3D image is then subjected to weighted back projection using the pixel index to obtain the corrected 3D image.

5. A geometric correction device for a three-dimensional image, characterized in that, include: An adjustment module is used to sequentially adjust the position of the flat panel detector according to preset geometric parameters to obtain multiple image reconstruction layers of the preprocessed three-dimensional image; wherein, the geometric parameters include external rotation angle and internal rotation angle; The determination module is used to extract image features corresponding to the multiple image reconstruction layers respectively; and to calculate the image information entropy of the multiple image features respectively, wherein the calculation formula of the image information entropy is as follows: ， ， in, The pixel grayscale value representing the image feature. The grayscale value representing the neighborhood of the image feature. This represents the total number of pixels representing the image feature. The side length of the image feature is represented. The probability that the image feature contains a pixel is represented; the largest image information entropy among multiple image information entropies is obtained; the image reconstruction layer corresponding to the largest image information entropy is determined as the target image reconstruction layer with low artifact intensity; the target geometric parameters corresponding to the target image reconstruction layer are obtained. The correction module is used to correct the preprocessed 3D image using the target geometric parameters to obtain the corrected 3D image.

6. The geometric correction device for three-dimensional images according to claim 5, characterized in that, The device further includes: The receiving module is used to receive X-ray scanning data of the object to be detected at different angles on the flat panel detector; The first processing module is used to sequentially perform dark correction processing, bad spot and bad line correction processing and air correction processing on the X-ray scanning data to obtain pre-processed scanning data. The second processing module is used to perform cone angle weighting and filtering three-dimensional back projection processing on the preprocessed scan data to obtain the three-dimensional image.

7. An electronic device, comprising: The system comprises a memory, a processor, and a communication bus, wherein the communication bus is communicatively connected to the memory and the processor, and the memory is coupled to the processor. The memory stores a computer program, and when the processor executes the computer program, it implements each step of the geometric correction method for three-dimensional images according to any one of claims 1 to 4.

8. A storage medium, said storage medium being a computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, implements the steps of the geometric correction method for a three-dimensional image as described in any one of claims 1 to 4.

Images