Pose calibration method, device and equipment based on Gaussian radiation and storage medium
By constructing a Gaussian radiation model for image matching, the problems of high cost and insufficient flexibility of high-precision calibration phantoms in CT imaging technology are solved, and high-stability and high-precision geometric pose calibration is achieved.
Patent Information
- Application Number
- CN202511591418.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-03
- Publication Date
- 2026-02-03
AI Technical Summary
Existing CT imaging technology suffers from several drawbacks in geometric correction, including high cost, insufficient flexibility, large errors, and poor consistency and stability of calibration results due to external factors.
A Gaussian radiation-based pose calibration method is adopted. By constructing a Gaussian radiation model, image matching is performed on the target phantom image sequence to determine the geometric pose of the target calibration device, thus avoiding the use of high-precision calibration phantoms.
It improves the flexibility and stability of geometric pose calibration, reduces production costs, and enhances the accuracy and reliability of sub-millimeter level measurements.
Smart Images
Figure CN121458804A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] Embodiments of the present application relate to the technical field of computers, which can be applied to the field of industrial detection, and in particular to a pose calibration method and device based on Gaussian radiation, an apparatus, and a storage medium. BACKGROUND
[0002] At present, in the field of industrial technology, the internal structure information of an object is obtained by X-rays emitted by a computed tomography (CT device), and internal defects, cracks, and foreign objects of the object are identified, so as to ensure the integrity and safety of the object, which has become a common non-destructive detection method.
[0003] However, the existing CT imaging technology faces significant challenges in geometric correction. The traditional geometric correction method mainly relies on formulating a standard phantom and then using a direct linear transformation technique to solve the system geometric pose. However, this method has the following defects: 1) high-precision calibration workpieces need to meet strict manufacturing standards, resulting in high production costs and insufficient flexibility in actual applications; 2) the error range of existing calibration workpieces cannot meet the high-precision detection requirements, especially in sub-millimeter measurement scenarios; 3) external factors such as temperature changes and mechanical vibrations easily cause calibration piece deformation or imaging system parameter drift, resulting in inconsistent geometric calibration results and poor long-term reliability. SUMMARY
[0004] The present application provides a pose calibration method and device based on Gaussian radiation, an apparatus, and a storage medium to improve the flexibility and stability of geometric pose calibration by calibration equipment.
[0005] According to an aspect of the present application, a pose calibration method based on Gaussian radiation is provided, which comprises:
[0006] Based on a target calibration device, image acquisition is performed on a target calibration phantom to generate a target phantom image sequence; wherein the target phantom image sequence includes phantom images of the target calibration phantom under at least two acquisition angles;
[0007] Based on a pre-trained Gaussian radiation model, image matching is performed on the phantom images in the target phantom image sequence, and the imaging pose of the matched phantom images in the Gaussian radiation model is used to calibrate the pose of the target calibration device at the acquisition angle corresponding to the phantom images; wherein the Gaussian radiation model is trained based on the volume data of the target calibration phantom and is used to represent the Gaussian radiation representation of the target calibration phantom under each acquisition angle of the target calibration device.
[0008] According to another aspect of the present application, a pose calibration device based on Gaussian radiation is provided, which comprises:
[0009] The image acquisition module is configured to acquire images of the target calibration phantom based on the target calibration device, and generate a target phantom image sequence; wherein the target phantom image sequence comprises phantom images of the target calibration phantom under at least two acquisition perspectives;
[0010] The pose calibration module is configured to perform image matching on the phantom images in the target phantom image sequence based on the pre-trained Gaussian radiation model, and perform pose calibration on the target calibration device under the acquisition perspective corresponding to the phantom image according to the imaging pose of the matched phantom image in the Gaussian radiation model; wherein the Gaussian radiation model is trained based on the volume data of the target calibration phantom, and is used to represent the Gaussian radiation representation of the target calibration phantom under each acquisition perspective of the target calibration device.
[0011] According to another aspect of the present application, an electronic device is provided, which comprises:
[0012] one or more processors;
[0013] a memory for storing one or more programs;
[0014] When the one or more programs are executed by the one or more processors, the one or more processors implement any one of the pose calibration methods based on Gaussian radiation provided by the embodiments of the present application.
[0015] According to another aspect of the present application, a computer readable storage medium is provided, which stores a computer program, and the program is executed by a processor to implement any one of the pose calibration methods based on Gaussian radiation provided by the embodiments of the present application.
[0016] According to another aspect of the present application, a computer program product is provided, which comprises a computer program, and the computer program is executed by a processor to implement any one of the pose calibration methods based on Gaussian radiation provided by the embodiments of the present application.
[0017] The present application constructs a Gaussian radiation model of the target calibration device, and uses the trained Gaussian radiation model to perform image matching on the phantom images in the target phantom image sequence, and performs pose calibration on the target calibration device under the acquisition perspective corresponding to the phantom image according to the imaging pose of the matched phantom image in the Gaussian radiation model, thereby avoiding geometric pose calibration of the target calibration device by a high-precision calibration phantom, and improving the flexibility and stability of geometric pose calibration. BRIEF DESCRIPTION OF DRAWINGS
[0018] Figure 1 is a flowchart of a pose calibration method based on Gaussian radiation according to an embodiment of the present application;
[0019] Figure 2 is a flow chart of a pose calibration method based on Gaussian radiation according to Embodiment Two of the present application;
[0020] Figure 3 is a structural schematic diagram of a pose calibration device based on Gaussian radiation according to Embodiment Three of the present application;
[0021] Figure 4 is a structural schematic diagram of an electronic device implementing the pose calibration method based on Gaussian radiation according to Embodiment Four of the present application. DETAILED DESCRIPTION
[0022] In order to enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should fall within the scope of protection of the present application.
[0023] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and do not necessarily indicate a specific order or sequence. It should be understood that the data thus used can be interchanged under appropriate circumstances, so that the embodiments of the present application described herein can be implemented in an order other than that illustrated or described herein. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusion, for example, a process, method, system, product or device including a series of steps or units does not have to be limited to only those steps or units clearly listed, but can include other steps or units not clearly listed or inherent to the process, method, product or device.
[0024] Embodiment One
[0025] Figure 1 is a flow chart of a pose calibration method based on Gaussian radiation according to Embodiment One of the present application. The present embodiment can be applicable to the case of geometric pose calibration of an electronic computed tomography scanner, and can be executed by a pose calibration device based on Gaussian radiation. The pose calibration device based on Gaussian radiation can be implemented in the form of hardware and / or software, and can be configured in a computer device, such as a server. As shown in the figure, the method comprises: Figure 1
[0026] S110, based on a target calibration device, image acquisition is performed on a target calibration phantom to generate a target phantom image sequence.
[0027] The target calibration device can be a CT device to be calibrated in geometry pose; the target calibration phantom can be an object to be imaged by the target calibration device; and the target phantom image sequence can include phantom images of the target calibration phantom under at least two acquisition perspectives. It should be noted that the acquisition perspective can be the geometric pose of X-rays emitted by the target calibration device.
[0028] Specifically, the target calibration phantom can be placed in the scanning range of the target calibration device to be calibrated in geometry pose, the target calibration device is controlled to scan the target calibration phantom, and phantom images under different acquisition perspectives are acquired to generate the target phantom image sequence. It should be noted that the number of phantom images in the target phantom image sequence can be adaptively set by those skilled in the art.
[0029] Optionally, in the embodiment of the present application, the target calibration phantom can include a phantom main body and at least one phantom feature; wherein the phantom feature can be uniformly embedded in the phantom main body, and the attenuation coefficient of the phantom feature is higher than that of the phantom main body.
[0030] For example, the phantom main body of the target calibration phantom can be low-attenuation organic glass, and the embedded phantom feature of the target calibration phantom can be high-purity graphite carbon block; the size of the phantom main body can be set to 200mm x 150mm x 100mm, and the phantom feature can be 5mm x 5mm x 5mm high-purity graphite carbon block, and the number is 9. The nine graphite carbon blocks can be uniformly distributed in the matrix form of 3 x 3 in the base, and the adjacent carbon blocks maintain a spacing of 30mm. Such design not only ensures that the phantom images acquired by the target calibration device have clear contrast, but also ensures the contrast difference in the target calibration phantom during X-ray imaging, which is helpful for subsequent image processing and analysis.
[0031] In S120, the phantom images in the target phantom image sequence are matched based on the pre-trained Gaussian radiation model, and the imaging pose of the target calibration phantom in the Gaussian radiation model corresponding to the matched phantom images is used to calibrate the pose of the target calibration device under the acquisition perspective corresponding to the phantom image.
[0032] The Gaussian radiation model can be trained based on the volume data of the target calibration phantom, and is used to represent the Gaussian radiation representation of the target calibration phantom under each acquisition perspective of the target calibration device, i.e. the response characteristics of the X-rays emitted by the target calibration device entering the target calibration phantom. It should be noted that in the embodiment of the present application, the content of the pose calibration can be the imaging geometric pose of the X-rays emitted by the target calibration device, i.e. the rotation angle and displacement angle in three directions.
[0033] In the embodiment of the present application, the object phantom images in the object phantom image sequence are images captured by the target calibration device without pose calibration, that is, the pose state corresponding to the object phantom images in the object phantom image sequence is not verified and is unknown. The trained Gaussian radiation model can be used to represent the imaging representation of the target calibration device for image capturing of the target calibration phantom, so that the object phantom images can be extracted from the object phantom image sequence, matched with the imaging representation of the trained Gaussian radiation model, and the object phantom image is calibrated according to the imaging pose corresponding to the imaging representation of the Gaussian radiation model matched with the object phantom image.
[0034] The embodiment of the present application constructs the Gaussian radiation model of the target calibration device, and uses the trained Gaussian radiation model to perform image matching on the object phantom images in the object phantom image sequence, and calibrates the capturing angle of the target calibration device corresponding to the object phantom image according to the imaging pose of the object phantom image in the Gaussian radiation model, thereby avoiding geometric pose calibration of the target calibration device by using high-precision calibration phantom, and improving the flexibility and stability of geometric pose calibration.
[0035] Embodiment two
[0036] Figure 2 is a flowchart of a pose calibration method based on Gaussian radiation according to the second embodiment of the present application. Based on the technical solutions of the above-mentioned embodiments, the present embodiment further limits the "performing image matching on the object phantom images in the object phantom image sequence based on the pre-trained Gaussian radiation model, and calibrating the capturing angle of the target calibration device corresponding to the object phantom image according to the imaging pose of the object phantom image in the Gaussian radiation model". It should be noted that the parts not described in detail in the embodiment of the present application can be referred to the related descriptions of other embodiments. For example, as shown in the figure, the method comprises the following steps. Figure 2
[0037] S210, capturing the target calibration phantom by the target calibration device to generate the target phantom image sequence.
[0038] S220, sequentially taking the object phantom images in the object phantom image sequence as the target phantom image, matching the target phantom image with the standard phantom image generated by the Gaussian radiation model, and taking the matched standard phantom image as the reference phantom image.
[0039] Optionally, the matching of the target phantom image with the standard phantom image generated by the Gaussian radiation model comprises: performing first image matching by determining the mean loss of image pixels between the target phantom image and the standard phantom image; performing second image matching on the target phantom image and the standard phantom image by a preset local feature matching algorithm; and taking the standard phantom image that simultaneously satisfies the first image matching and the second image matching as the matched image.
[0040] Optionally, the matching process can be represented by the following formula:
[0041] ;
[0042] wherein β is an empirical parameter, which can be generally set as 0.2, MSE() is a mean square error loss function for determining the image, which can be represented as the first image matching process, LoFTR() is a preset local feature matching algorithm, is the target phantom image, i can refer to the position of the target phantom image in the target phantom image sequence, and I is the standard phantom image.
[0043] The image matching in the standard phantom image is determined by the image pixel matching and the local feature matching, so that the matching range and the matching accuracy of the image are improved.
[0044] S230, taking the geometric pose of the ray imaging corresponding to the reference phantom image as the geometric pose of the ray imaging corresponding to the target phantom image.
[0045] Optionally, in the embodiment of the present application, the training process of the Gaussian radiation model comprises: determining an initial Gaussian radiation model of a target calibration device according to the volume data of a target calibration phantom; uniformly dividing the target calibration phantom into a preset number of sampling points according to the volume data of the target calibration phantom; constructing a target loss function of the initial Gaussian radiation model by using the initial Gaussian radiation model, the sampling imaging result of the sampling points and the simulation image result corresponding to the sampling points, and iteratively training the initial Gaussian radiation model to generate the Gaussian radiation model.
[0046] wherein the target loss function can comprise the average absolute error loss of the pixels between the sampling imaging result and the simulation image result, and the structural similarity loss between the sampling imaging result and the simulation image result.
[0047] It should be noted that the sampling point can refer to a small cuboid after the target calibration phantom is divided. Optionally, the number of sampling points can be adaptively set by those skilled in the art.
[0048] In an optional implementation of the embodiment of the present application, a CT device that has been pose calibrated can be used to scan the target calibration phantom to obtain volume data for representing the target calibration phantom, which can be used to represent the length, width and height information of the target calibration phantom; the volume data of the target calibration phantom is normalized to ensure that the center position of the target calibration phantom is at the origin position, and the volume data is normalized to a preset coordinate range. The center coordinates of the sampling points are input to the initial Gaussian radiation model as the sampling point coordinates, and the initial Gaussian radiation model is used to construct a target loss function of the initial Gaussian radiation model for the sampling imaging results of the sampling points and the corresponding simulation image results of the sampling points, and the initial Gaussian radiation model is iteratively trained. It should be noted that the simulation image result is a credible image formed by simulating the attenuation of X-rays penetrating the sampling points through digital reconstruction radiography technology.
[0049] Optionally, the initial Gaussian radiation model can be represented by the following formula:
[0050] ;
[0051] wherein G represents a group of N Gaussian point distribution representations, N is the number of sampling points, represents the center coordinates of the jth sampling point, refers to the covariance matrix corresponding to the jth sampling point, refers to the opacity corresponding to the jth sampling point, refers to the X-ray response intensity corresponding to the jth sampling point.
[0052] Optionally, the covariance matrix can be further determined as a combination of a rotation matrix R and a scaling matrix S, i.e.:
[0053] .
[0054] Optionally, the target loss function can be represented by the following formula:
[0055] ;
[0056] wherein N is the number of sampling points; is an empirical parameter; is a mean absolute error loss function; is a structural similarity loss function; is a simulation image result generated by using digital reconstruction radiography technology for the jth sampling point; is a sampling imaging result generated by the Gaussian radiation model for the jth sampling point in the training process.
[0057] By constructing a 3D Gaussian representation under an X-ray scene, the 3D Gaussian representation field is optimized through machine learning, and the 3D Gaussian radiation representation of the target calibration phantom is accurately calibrated, which is used to generate the sampling imaging result of the target calibration phantom under a new view angle.
[0058] The embodiment of the present application determines the geometric pose of the X-ray emitted by the target calibration device by adopting the matching degree of the feature points and the image pixels, improves the repeatability and efficiency of the geometric pose calibration, and simultaneously improves the matching accuracy of the image matching by adopting a composite loss function in the image matching process.
[0059] Embodiment three
[0060] Figure 3 is a structure schematic diagram of a pose calibration device based on Gaussian radiation according to the embodiment three of the present application, which can be applied to the geometric pose calibration of the electronic computed tomography scanner. The pose calibration device based on Gaussian radiation can be realized in the form of hardware and / or software, and can be configured in a computer device, such as a server. As shown in the figure, the device comprises: Figure 3
[0061] The image acquisition module 310 is configured to acquire images of the target calibration phantom based on the target calibration device, and generate a target phantom image sequence. The target phantom image sequence comprises phantom images of the target calibration phantom under at least two acquisition view angles.
[0062] The pose calibration module 320 is configured to perform image matching on the phantom images in the target phantom image sequence based on the pre-trained Gaussian radiation model, and calibrate the pose of the target calibration device at the acquisition view angle corresponding to the phantom image according to the imaging pose of the matched phantom image in the Gaussian radiation model. The Gaussian radiation model is trained based on the volume data of the target calibration phantom, and is used to represent the Gaussian radiation representation of the target calibration phantom under each acquisition view angle of the target calibration device.
[0063] The embodiment of the application constructs a Gaussian radiation model of a target calibration device, and performs image matching on a phantom image in a target phantom image sequence by using the trained Gaussian radiation model, and calibrates the pose of the target calibration device at a collection angle corresponding to the phantom image according to the imaging pose of the matched phantom image in the Gaussian radiation model, thereby avoiding geometric pose calibration of the target calibration device by using a high-precision calibration phantom, and improving the flexibility and stability of geometric pose calibration.
[0064] Optionally, the image acquisition module 310 comprises:
[0065] an image matching unit, configured to sequentially take a phantom image in the target phantom image sequence as a target phantom image, and match the target phantom image with a standard phantom image generated by the Gaussian radiation model, and take the standard phantom image after the matching as a reference phantom image;
[0066] a pose calibration unit, configured to take a geometric pose of ray imaging corresponding to the reference phantom image as a geometric pose of ray imaging corresponding to the target phantom image; wherein the geometric pose of ray imaging is used to represent a rotation angle and a displacement angle of a ray emitted by the target calibration device.
[0067] Optionally, the image matching unit can be specifically configured to:
[0068] perform first image matching by determining a mean loss of image pixels between the target phantom image and the standard phantom image;
[0069] perform second image matching on the target phantom image and the standard phantom image by using a preset local feature matching algorithm;
[0070] take the standard phantom image that satisfies both the first image matching and the second image matching as a matched image.
[0071] Optionally, the training process of the Gaussian radiation model comprises: determining an initial Gaussian radiation model of the target calibration device according to volume data of the target calibration phantom; dividing the target calibration phantom into a preset number of sampling points uniformly according to the volume data of the target calibration phantom; constructing a target loss function of the initial Gaussian radiation model by using a sampling imaging result of the sampling points and a simulation image result corresponding to the sampling points in the initial Gaussian radiation model, and iteratively training the initial Gaussian radiation model.
[0072] Optionally, the target loss function comprises a mean absolute error loss of pixels between the sampling imaging result and the simulation image result, and a structural similarity loss between the sampling imaging result and the simulation image result.
[0073] Optionally, the target calibration phantom comprises a phantom main body and at least one phantom feature; wherein the phantom feature is uniformly embedded in the phantom main body, and the attenuation coefficient of the phantom feature is higher than that of the phantom main body.
[0074] The pose calibration device based on Gaussian radiation provided by the embodiments of the present application can execute the pose calibration method based on Gaussian radiation provided by any of the embodiments of the present application, and has the corresponding function modules and beneficial effects of executing each pose calibration method based on Gaussian radiation.
[0075] According to the embodiments of the present application, the present application further provides an electronic device, a readable storage medium and a computer program product.
[0076] Embodiment Four
[0077] Figure 4 is a structural schematic diagram of an electronic device 410 for implementing the pose calibration method based on Gaussian radiation according to the embodiments of the present application. The electronic device is intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular telephones, smart phones, wearable devices (such as helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are merely examples and are not intended to limit the implementations described and / or claimed in this document to the examples described herein.
[0078] As shown in Figure 4 The electronic device 410 includes at least one processor 411, and a memory, such as a read-only memory 412, a random access memory 413, etc., which is communicatively connected to the at least one processor 411, wherein the memory stores a computer program that can be executed by the at least one processor. The processor 411 can execute various appropriate actions and processes according to the computer program stored in the read-only memory 412 or loaded from the storage unit 418 to the random access memory 413. In the random access memory 413, various programs and data required for the operation of the electronic device 410 can also be stored. The processor 411, the read-only memory 412 and the random access memory 413 are connected to each other through a bus 414. An input / output interface 415 is also connected to the bus 414.
[0079] A number of components in the electronic device 410 are connected to the input / output interface 415, including: an input unit 416, e.g., a keyboard, a mouse, etc.; an output unit 417, e.g., various types of displays, speakers, etc.; a storage unit 418, e.g., a magnetic disk, an optical disk, etc.; and a communication unit 419, e.g., a network card, a modem, a wireless communication transceiver, etc. The communication unit 419 allows the electronic device 410 to exchange information / data with other devices through a computer network, such as the Internet, and / or various telecommunication networks.
[0080] The processor 411 can be various general and / or special purpose processing components having processing and computing capabilities. Some examples of the processor 411 include, but are not limited to, central processing units, graphics processing units, various specialized artificial intelligence computing chips, various processors running machine learning model algorithms, digital signal processors, and any appropriate processors, controllers, microcontrollers, etc. The processor 411 performs various methods and processes described above, e.g., the Gaussian radiation based pose calibration method.
[0081] In some embodiments, the Gaussian radiation based pose calibration method can be implemented as a computer program tangibly embodied in a computer readable storage medium, e.g., the storage unit 418. In some embodiments, part or all of the computer program can be loaded and / or installed onto the electronic device 410 via the read-only memory 412 and / or the communication unit 419. When the computer program is loaded onto the random access memory 413 and executed by the processor 411, one or more steps of the Gaussian radiation based pose calibration method described above can be performed. Alternatively, in other embodiments, the processor 411 can be configured as the Gaussian radiation based pose calibration method by any other appropriate means, e.g., by means of firmware.
[0082] The various implementations of the systems and techniques described above can be realized in digital electronic circuitry, integrated circuitry, specially designed application specific integrated circuits, application specific standard products, chips, microprocessors, computer hardware, firmware, software, and / or combinations thereof. These various implementations can include implementation in one or more computer programs that are executable and / or interpretable on a programmable system including at least one programmable processor, which can be special or general purpose, coupled to receive data and instructions from, and to transmit data and instructions to, a storage system, at least one input device, and at least one output device.
[0083] A computer program for implementing the methods of the present application can be written in any combination of one or more programming languages. The computer program can be provided to a general purpose computer, special purpose computer, or other programmable apparatuses that can be programmed to implement the functions / operations specified in the flowcharts and / or block diagrams. The computer program can be executed entirely on a machine, partially on a machine, partially on a machine as a standalone software package and partially on a remote machine, or entirely on a remote machine or server.
[0084] In the context of the present application, a computer-readable storage medium can be a tangible medium that can contain or store the computer program for use by or in connection with an instruction execution system, apparatus, or device. Computer-readable storage media can include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium can be a machine-readable signal medium. More specific examples of a machine-readable storage medium will include one or more lines of electrical connections, portable computer disks, hard disk drives, random access memory, read-only memory, erasable programmable read-only memory, optical fibers, portable compact disc read-only memories, optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.
[0085] To provide for interaction with a user, the systems and techniques described here can be implemented on an electronic device having a display device (e.g., a CRT or LCD monitor) for displaying information to the user and a keyboard and a pointing device (e.g., a mouse or a trackball) by which the user can provide input to the electronic device. Other kinds of devices can be used to provide for interaction with a user as well; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form, including acoustic, speech, or tactile input.
[0086] The systems and techniques described herein can be implemented in a computing system that includes a back end component, e.g., as a data server, or that includes a middleware component, e.g., an application server, or that includes a front end component, e.g., a user computer having a graphical user interface or a Web browser through which a user can interact with an implementation of the systems and techniques described herein, or any combination of such back end, middleware, or front end components. The components of the system can be interconnected by any form or medium of digital data communication, e.g., a communication network. Examples of communication networks include a local area network ("LAN"), a wide area network ("WAN"), a blockchain network, and the Internet.
[0087] The computing system can include clients and servers. A client and server are generally remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. A server can be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system, and solves the defects of large management difficulty and weak business scalability in traditional physical host and virtual private server.
[0088] It should be understood that the various forms of flow shown above can be re-ordered, added to, or deleted from without departing from the scope of the present disclosure. For example, the steps recited in the present disclosure can be executed in parallel, executed in sequence, or executed in different orders, as long as the desired results of the technical solutions of the present disclosure can be achieved, and the present disclosure is not limited herein.
[0089] The specific implementation described above does not constitute a limitation on the scope of protection of the present application. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present application shall be included in the scope of protection of the present application.
Claims
1. A pose calibration method based on Gaussian radiation, characterized in that, The method comprises: based on the target calibration device, image acquisition is performed on the target calibration phantom to generate a target phantom image sequence; wherein the target phantom image sequence comprises phantom images of the target calibration phantom under at least two acquisition angles; based on a pre-trained Gaussian radiation model, image matching is performed on the phantom images in the target phantom image sequence, and the imaging pose of the matched phantom images in the Gaussian radiation model is used to calibrate the pose of the target calibration device at the acquisition angle corresponding to the phantom image; wherein the Gaussian radiation model is trained based on the volume data of the target calibration phantom and is used to represent the Gaussian radiation representation of the target calibration phantom under each acquisition angle of the target calibration device.
2. The method of claim 1, wherein, The method comprises: the target phantom image sequence is sequentially taken as a target phantom image, and the target phantom image is matched with a standard phantom image generated by the Gaussian radiation model, and the matched standard phantom image is taken as a reference phantom image; the geometric pose of the ray imaging corresponding to the reference phantom image is taken as the geometric pose of the ray imaging corresponding to the target phantom image; wherein the geometric pose of the ray imaging is used to represent the rotation angle and displacement angle of the rays emitted by the target calibration device.
3. The method of claim 2, wherein, The method comprises: first image matching is performed by determining the mean loss of image pixels between the target phantom image and the standard phantom image; second image matching is performed on the target phantom image and the standard phantom image by using a pre-set local feature matching algorithm; the standard phantom image that simultaneously satisfies the first image matching and the second image matching is taken as the matched image.
4. The method of claim 1, wherein, The training process of the Gaussian radiation model comprises: determining an initial Gaussian radiation model of the target calibration device according to the volume data of the target calibration phantom; dividing the target calibration phantom into a pre-set number of sampling points according to the volume data of the target calibration phantom; using the sampling imaging results of the sampling points and the corresponding simulation image results of the sampling points to construct a target loss function of the initial Gaussian radiation model, and iteratively training the initial Gaussian radiation model.
5. The method of claim 4, wherein, The target loss function comprises the average absolute error loss of pixels between the sampling imaging results and the simulation image results, and the structural similarity loss between the sampling imaging results and the simulation image results.
6. The method of claim 1, wherein, The target calibration phantom comprises a phantom main body and at least one phantom feature body; wherein the phantom feature body is uniformly embedded in the phantom main body, and the attenuation coefficient of the phantom feature body is higher than that of the phantom main body.
7. A pose calibration apparatus based on Gaussian radiation, characterized by, The method comprises: An image acquisition module configured to acquire images of the target calibration phantom based on the target calibration device, and generate a target phantom image sequence, wherein the target phantom image sequence comprises phantom images of the target calibration phantom under at least two acquisition viewpoints; A pose calibration module configured to perform image matching on the phantom images in the target phantom image sequence based on a pre-trained Gaussian radiance model, and calibrate a pose of the target calibration device under an acquisition viewpoint corresponding to a phantom image according to an imaging pose of the phantom image in the Gaussian radiance model, wherein the Gaussian radiance model is trained based on volume data of the target calibration phantom, and is used to represent a Gaussian radiance representation of the target calibration phantom under each acquisition viewpoint of the target calibration device.
8. An electronic device, comprising: comprising: one or more processors; memory for storing one or more programs; when the one or more programs are executed by the one or more processors, the one or more processors implement the Gaussian radiance based pose calibration method according to any one of claims 1-6.
9. A computer-readable storage medium having stored thereon a computer program, characterized in that, The program is executed by the processor to implement the Gaussian radiance based pose calibration method according to any one of claims 1-6.
10. A computer program product comprising a computer program which, when executed by a processor, implements the Gaussian radiance based pose calibration method according to any one of claims 1-6.