Construction method of simulation data acquisition platform for digital impression technology research

Abstract

The invention discloses a construction method of a simulation data acquisition platform for digital impression technology research, which comprises the following steps of: (1) importing an acquisition object by using an unreal engine, and constructing a scanning environment; (2) constructing a blueprint class representing an intraoral scanner in an unreal engine, and constructing and designing a novel intraoral scanner for research by taking an existing intraoral scanner as a contrast; (3) formulating control and data acquisition logic according to the designed novel intraoral scanner, and acquiring multi-modal data in combination with an Unreal CV plug-in; and (4) carrying out post-processing on the obtained original data. The simulation platform provided by the invention has the characteristics of low cost and high error-tolerant rate, and the research and experiment investment is remarkably reduced; the quality of data output by the platform is high, and the simulation performance is excellent; the platform may return the true value of the acquisition object to verify the validity of the novel design.

Description

Technical Field

[0001] The invention relates to the field of digital impression technology, and in particular to a method for constructing a simulation data acquisition platform for digital impression technology research. Background Art

[0002] In the field of dental diagnosis and treatment, the methods of obtaining oral and dental model data can be divided into two categories: traditional impression technology and digital impression technology. Digital impression technology uses an intraoral scanner to directly scan the patient's mouth and obtain digital oral data including the three-dimensional morphology and color texture of the soft and hard tissue surfaces in the oral cavity.

[0003] There are many intraoral scanner products at present, but as far as we know, they all use a single invasive optical window to obtain surface color data and depth data, and then stitch the multiple single-frame data obtained to form a complete oral dental model. Although the single-window design has been increasingly widely used, it still has some problems, such as complex scanning trajectories, long time-consuming complete data acquisition, higher requirements for operators, and reduced patient comfort; insufficient scanning accuracy in special cases such as missing teeth, and not recommended for use in clinical applications such as edentulous jaw implants. In order to solve the above problems, it is necessary to study and innovate the design of intraoral scanners. However, before the effectiveness of the newly designed intraoral scanner is determined, the cost of directly manufacturing physical objects or prototypes for research is high and the fault tolerance rate is low. This high risk and high cost have restricted the development of digital impression technology to a certain extent.

[0004] With the rapid development of computer simulation technology, various simulation platforms have been widely used in many high-tech fields, such as autonomous driving, space exploration, and synchronous positioning and composition research. However, in the field of digital impressions, such as the design and research of intraoral scanners, the development of related technologies is relatively lagging behind. Summary of the invention

[0005] In view of the above-mentioned defects of the prior art, the technical problem to be solved by the present invention is that directly manufacturing physical objects or prototypes to study digital impression technology is costly and has a low error tolerance.

[0006] To achieve the above object, the present invention provides a method for constructing a simulation data acquisition platform for digital impression technology research, comprising:

[0007] (1) Use Unreal Engine to import the collected objects and build the scanning environment;

[0008] (2) Construct a blueprint class in Unreal Engine to represent the intraoral scanner, and use the existing intraoral scanners as a reference to construct and design a new intraoral scanner for research;

[0009] (3) Based on the designed new intraoral scanner, develop control and data acquisition logic and combine it with the UnrealCV plug-in to collect multimodal data;

[0010] (4) Post-process the raw data obtained by UnrealCV.

[0011] The present invention utilizes computer simulation technology to carry out prototype design and oral data collection. Various newly conceived impression technologies can be realized as a blueprint class of Unreal Engine; Unreal Engine can capture the color and depth data required for digital impressions after combining with UnrealCV plug-in; different oral models are imported into Unreal Engine to simulate various clinical situations.

[0012] The present invention can design a new intraoral scanner or other impression technology without hardware development, and simulate the acquisition of oral images for related research, thereby improving the fault tolerance rate and reducing the research cost.

[0013] In a preferred embodiment of the present invention, the scanning environment set up in step (1) includes a simulated laboratory scanning environment, an intraoral scanning environment, or an interference environment for intraoral scanning;

[0014] The laboratory scanning environment uses a sky ball and directional light to produce a natural lighting effect;

[0015] The intraoral scanning environment uses a closed sky ball and corresponding directional lighting, and binds the directional light source to the camera component of the subsequent simulated intraoral scanner, and switches and moves synchronously during the data acquisition process;

[0016] The interference environment of the intraoral scanning includes the introduction of the tongue and the buccal mucosa of the oral cavity.

[0017] In a preferred embodiment of the present invention, the acquisition object in step (1) includes a dental model with or without teeth, and the scanning environment includes any one of a simulated laboratory scanning environment, an intraoral scanning environment, or an interference environment for intraoral scanning.

[0018] In another preferred embodiment of the present invention, the existing intraoral scanner in step (2) comprises a single-view window intraoral scanner;

[0019] The novel intraoral scanner includes any one or more of a dual-viewing window intraoral scanner, a triple-viewing window intraoral scanner or a tray-type intraoral scanner.

[0020] In another preferred embodiment of the present invention, the control and data acquisition logic in step (3) requires an input device to receive control instructions, and the input device used includes a keyboard, a mouse, or a somatosensory game controller;

[0021] In step (3), the movement of the novel scanner is controlled to perform scanning;

[0022] In step (3), the control logic includes scanner movement, data acquisition instructions or view angle switching;

[0023] In step (3), the collected data includes window pose and image;

[0024] The data acquisition in step (3) uses Python combined with Unreal Engine to record the window posture;

[0025] In step (3), the color depth data is collected by combining the viewport pose and the UnrealCV plug-in.

[0026] In another preferred embodiment of the present invention, the image acquired in step (3) includes color and depth information.

[0027] In another preferred embodiment of the present invention, the novel intraoral scanner in step (2) is a multi-window scanner, and the correspondence between the logic and the instructions in step (3) includes:

[0028] W\S\A\D: front, back, left, and right motion inputs of the new intraoral scanner;

[0029] Horizontal and vertical mouse movements: Angular input of the new intraoral scanner in the coronal and transverse planes;

[0030] Mouse wheel: Angle input of the new intraoral scanner in the sagittal plane;

[0031] V: Switches between the first-person novel intraoral scanner and the third-person operator's perspective;

[0032] Left mouse button: switch the angle control object, control the new intraoral scanner or the third-person camera;

[0033] C: Start and stop video-based continuous data acquisition;

[0034] R: Photographic single-frame data acquisition;

[0035] F: End the entire data collection process and store the pose results.

[0036] In another preferred embodiment of the present invention, the novel intraoral scanner in step (2) is a tray-type intraoral scanner, and the logic and instruction list in step (3) includes:

[0037] V: view switching between multiple cameras;

[0038] R: Photographic single-frame data acquisition;

[0039] F: End the entire data collection process and store the pose results.

[0040] In another preferred embodiment of the present invention, the post-processing in step (4) comprises:

[0041] (A) Depth conversion;

[0042] (B) Camera intrinsic parameter calculation;

[0043] (C) Camera pose ground truth conversion.

[0044] (D) Depth noise addition

[0045] In another preferred embodiment of the present invention, the formula used for the depth conversion in step (A) is as follows:

[0046]

[0047] Among them, d plane is the plane depth, d ray is the depth obtained by UnrealCV, d pixel is the pixel distance from the midpoint of the field of view to the optical center in the imaging result, and f is the focal length of the camera;

[0048] The formula of the internal reference in step (B) without considering the error is as follows:

[0049]

[0050] Where f is the focal length of the camera, C x is the optical center position on the x-axis of the image, C y is the optical center position on the y-axis of the image, W is the image pixel width, H is the image pixel height, FOV h is the horizontal field of view of the camera.

[0051] In another preferred embodiment of the present invention, the pose truth value is converted from the left-handed lower and Euler angle representation in the Unreal Engine to the right-handed system and transfer matrix representation commonly used for reconstruction, and the pose truth value is converted into the motion sensor output.

[0052] In another preferred embodiment of the present invention, in step (C), simulating the output of the intraoral scanner motion sensor comprises the following steps:

[0053] (a) Calculate linear and angular velocities from the true pose output;

[0054] (b) simulate accelerometer data, calculate linear acceleration from the linear velocity obtained in (11) and add Gaussian noise or fixed bias of the sensor;

[0055] (c) Simulated gyroscope data. The gyroscope directly measures the angular velocity and adds Gaussian noise or a fixed offset of the sensor to the angular velocity obtained in (a).

[0056] In another preferred embodiment of the present invention, the method of adding the depth noise in step (D) includes:

[0057] (i) Add random xy offsets to the input depth image and calculate the disparity map from the depth map based on the camera parameters;

[0058] (j) Using the speckle image of the Kinect camera as a mask, a sliding window is used to traverse the disparity image and the mask image;

[0059] (k) If the pixel mask of the center of the window is greater than 0 and the number of pixels with a mask greater than 0 in the window exceeds the threshold thresh1, the processing flow is entered;

[0060] (l) Calculate the mean of the valid pixels in the window, and calculate the difference between each pixel and the mean and weight them. The farther away from the center, the lower the weight. Mark the points whose difference is less than the threshold thresh2 as inliers.

[0061] (m) Verify the number of inliers in the window in step (l). If the number exceeds the threshold thresh3, it is considered credible and the input disparity of the window center point is updated to the output;

[0062] (n) Fill the output disparity of the invalid pixels in step (m), and the criterion is to select the input disparity of the nearest valid center point as the output;

[0063] (o) Add Gaussian noise to the disparity obtained in step (n) and recalculate the depth map using the same camera parameters as in step (i).

[0064] In another preferred embodiment of the present invention, the depth noise pattern in step (D) complies with actual imaging principles, and the imaging principles include active structured light, confocal or active wavefront sampling.

[0065] Technical Effects

[0066] The simulation platform provided by the present invention has the characteristics of low cost and high fault tolerance, which significantly reduces the investment in research and experiments; the data output by the platform is of high quality and the simulation performance is excellent; the platform can return the true value of the collected object. These performance indicators enable researchers to design and develop new digital impression technologies through the platform and verify their effectiveness. It has important application value in scientific research.

[0067] The concept, specific structure and technical effects of the present invention will be further described below in conjunction with the accompanying drawings to fully understand the purpose, characteristics and effects of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 It is a flow chart of a method for constructing a simulation data acquisition platform for digital impression technology research proposed by the present invention;

[0069] Figure 2 It is a digital model of three edentulous jaw implant scans of a preferred embodiment of the present invention;

[0070] Figure 3 It is a preferred embodiment of the present invention, which includes a blueprint class of a scanner in a dual-view window and an Unreal Engine interface of a dental jaw model of a collection object;

[0071] Figure 4 is an example diagram of the motion logic of a preferred embodiment of the present invention;

[0072] Figure 5 This is an example diagram of the posture recording logic during the scanning process of a preferred embodiment of the present invention;

[0073] Figure 6 is a flow chart of a construction method of an embodiment of the present invention in which a user interface is added;

[0074] Figure 7 is a user interface of a preferred embodiment of the present invention;

[0075] Figure 8 This is the depth conversion calculation principle of a preferred embodiment of the present invention.

[0076] Among them, 1-construction of acquisition object and scanning environment, 2-construction of intraoral scanner object, 3-implementation of control and data acquisition logic, 4-post-processing of raw data, 5-scanning body with notch, 6-gear-shaped scanning body, 7-scanning body with extension rod, 8-camera component representing the main window, 9-camera representing sub-windows arranged at a certain angle, 10-third-person lens, 11-dental model, 12-event distributor, 13-timer node, 14-construction of acquisition object and scanning environment, 15-construction of intraoral scanner object, 16-implementation of control and data acquisition logic, 17-user interface design, 18-post-processing of raw data, 19-view observed by the third-person lens, 20-data collected by the sub-windows on both sides of the teeth, 21-text information prompt area, 22-scanner shell, 23-plane depth, 24-depth obtained by UnrealCV, 25-pixel distance, 26-camera focal length. DETAILED DESCRIPTION

[0077] The following describes several preferred embodiments of the present invention with reference to the drawings in the specification, so that the technical content is clearer and easier to understand. The present invention can be embodied in many different forms of embodiments, and the protection scope of the present invention is not limited to the embodiments mentioned in the text.

[0078] In the drawings, components with the same structure are indicated by the same numerical reference numerals, and components with similar structures or functions are indicated by similar numerical reference numerals. The size and thickness of each component shown in the drawings are arbitrarily shown, and the present invention does not limit the size and thickness of each component. In order to make the illustration clearer, the thickness of the components is appropriately exaggerated in some places in the drawings.

[0079] like Figure 1 As shown, the method for constructing a simulation data acquisition platform for digital impression technology research proposed by the present invention includes:

[0080] (1) Use Unreal Engine to import the collected objects and build the scanning environment;

[0081] (2) Construct a blueprint class in Unreal Engine to represent the intraoral scanner, and use the existing intraoral scanners as a reference to construct and design a new intraoral scanner for research;

[0082] (3) Based on the designed new intraoral scanner, develop control and data acquisition logic and combine it with the UnrealCV plug-in to collect color depth data;

[0083] (4) Post-process the raw data obtained by UnrealCV.

[0084] Import the acquired object to import the digital model of the dental jaw or the digital model of the edentulous jaw implant scan into Unreal Engine.

[0085] In some embodiments, the collection object is an oral dental model collected clinically and provided by a hospital. The model is usually in a common three-dimensional format, for example, a three-dimensional model in an obj format and texture data in a jpg format.

[0086] In some embodiments, clinical data on edentulous teeth are relatively scarce. To conduct relevant research, standard models of edentulous jaws and scanned bodies can be downloaded from the Internet to simulate the production of acquisition objects in the case of edentulous teeth.

[0087] like Figure 2 As shown, three digital models for edentulous jaw implant scanning were designed for research, corresponding to a scanning body 5 with a notch, a gear-shaped scanning body 6, and a scanning body 7 with an extension rod.

[0088] Scan the environment and build it. Unreal Engine provides a wealth of lighting objects and objects to present realistic effects.

[0089] In some embodiments, a laboratory scanning environment is simulated, and a sky ball and directional light are used to produce a natural lighting effect.

[0090] In some embodiments, the environment of intraoral scanning is simulated, the sky ball and the corresponding directional lighting are turned off, and the directional light source is bound to each camera component of the subsequent simulated intraoral scanner, and the switching and movement are synchronized during the data acquisition process.

[0091] In some embodiments, the tongue and buccal mucosa are introduced to simulate some interferences that may occur in actual intra-oral scanning.

[0092] Intraoral scanner construction. In order to verify the effectiveness of the new intraoral scanner, you first need to build a single-view window intraoral scanner as a control. Specifically, create a blueprint class and introduce a camera component to it to represent a single window used to collect data. In order to better observe the acquisition process, you can also introduce another camera to represent the third-person lens of the operator's perspective.

[0093] New intraoral scanner object construction. This is just an example of a possible new design approach and does not limit the applicability of the simulation platform.

[0094] like Figure 3 As shown, a dual-view window scanner includes the design blueprint class and the interface of the acquisition object jaw model 11. In the blueprint class of the single-view window scanner, a camera 9 is introduced with a certain angle to the existing camera component 8. The third-person lens 10 can be used to observe the posture of the scanner during data acquisition.

[0095] In some embodiments, a three-view window scanner is provided, based on a two-view window scanner, with an additional camera assembly. Three windows at certain angles can simultaneously collect data on the occlusal surface, lingual side, and buccal side of the tooth. The simulation platform proposed in the present invention can help determine the angles between multiple windows and design the scanning trajectory of the multi-view window scanner.

[0096] In some embodiments, a tray-type intraoral scanner is provided. Cameras at several angles are placed around the dental model to collect data synchronously. The positions of the cameras are obtained by pre-calibration, and after the data is collected, the calibrated positions are directly used for splicing and reconstruction. The design further shortens the acquisition time and improves the model accuracy. The simulation platform proposed in the present invention can help determine the number of cameras and the arrangement angles that are ultimately required.

[0097] For the multi-view window scanner in step 2, researchers need to control its movement to scan in order to obtain the complete data required for reconstruction. Optionally, use the keyboard and mouse as input devices, and use the nodes provided by Unreal Engine in the event graph of the blueprint class to complete it, such as Figure 4The following is an example of motion logic. The Get Control Rotation and Get World Rotation nodes are used to get the orientation of the data acquisition camera and the third-person camera, and motion input is added based on the orientation to get a more intuitive control experience.

[0098] The UnrealCV plug-in can move the camera to a specified position and output data such as color images and depth images. However, the speed of acquiring and storing data by the plug-in is limited. Therefore, when scanning, Python programming is used to combine the Unreal Engine to first record and store the camera position, and then the data is collected in combination with the UnrealCV plug-in according to the stored position. Multiple frames of color data in png format and depth data in npy format are obtained.

[0099] like Figure 5 As shown, an example of the pose recording logic during the scanning process, the event distributor 12 is responsible for recording the camera pose of the current frame. Use the Unreal Engine Python plug-in to write a python class and bind it to the blueprint object, write the camera pose recording and storage function and bind it to the event distributor 12. Use the timer node 13 to collect data at fixed time intervals.

[0100] The corresponding list of all commands and logic is as follows. Because the scanner needs to be inserted into the mouth, it is generally not rotated on the sagittal plane of the human body, so the mouse wheel is selected as its input.

[0101] (1) W\S\A\D: Scanner motion input, front, back, left, and right;

[0102] (2) Horizontal and vertical movement of the mouse: input of the angle of the scanner in the coronal and transverse planes;

[0103] (3) Mouse wheel: Scanner angle input in the sagittal plane;

[0104] (4) V: Switches between the first-person scanner and third-person operator perspectives;

[0105] (5) Left mouse button: switch between angle control objects, control the scanner or the third-person camera;

[0106] (6) C: Start and stop video-based continuous data acquisition (recording camera position);

[0107] (7) R: Photographic single-frame data acquisition (recording camera position);

[0108] (8)F: End the entire data collection process and store the pose results in a json file.

[0109] If the tray-type intraoral scanner described in step 2 is studied, the above motion-related control logic is not required and the instruction list is as follows.

[0110] (1) V: view switching between multiple cameras;

[0111] (2) R: Photographic single-frame data acquisition (recording camera position);

[0112] (3)F: End the entire data collection process and store the pose results in a json file.

[0113] In some embodiments, a somatosensory game controller PICO is used as an input device. After installing the PICO UnrealIntegration SDK toolkit, configuring the input mapping of the controller in the Unreal Engine Editor can better simulate the feel of controlling the movement of the intraoral scanner.

[0114] like Figure 6 As shown in a preferred embodiment, after the control and data acquisition logic 16 is implemented, a user interface 17 can be designed to better perform data acquisition.

[0115] like Figure 7 As shown, an embodiment of a user interface is a schematic diagram of the user interface of the scanner in the three-view window described in step 2, which is made using the Widget blueprint component of the Unreal Engine. The left side is the user interface in the first-person scanner perspective, and the main body of the screen is the data collected by the camera on the occlusal surface of the teeth. The upper left window 19 displays the view observed by the third-person camera, and the upper right window 20 displays the data collected by the cameras on both sides of the teeth. The txt text prompt area 21 on the left indicates the current status of the system, such as showing which perspective the current angle input acts on, whether video-style continuous data collection is currently being performed, etc. The right side is the user interface after switching to the third-person operator perspective, and the theme of the screen is the view observed by the third-person camera. At this time, the scanner housing 22 will be displayed on the screen, and the operator can observe whether the posture of the scanner is appropriate during the scanning process.

[0116] In some embodiments, the post-processing in step (4) includes:

[0117] (1) Depth conversion. Figure 8 As shown, the depth 24 obtained by UnrealCV is the distance from the midpoint of the field of view to the optical center of the camera, which needs to be converted into the commonly used plane depth 23. The formula used is as follows:

[0118]

[0119] where d plane The required plane depth is 23, d rayThe depth 24 obtained by UnrealCV, d pixel is the pixel distance from the midpoint of the field of view to the optical center in the imaging result 25, and f is the focal length of the camera 26.

[0120] (2) Calculation of camera internal parameters. The internal parameters of the virtual camera are obtained by the following formula without considering the error:

[0121]

[0122] Where f is the focal length of the camera, C x is the optical center position on the x-axis of the image, C y is the optical center position on the y-axis of the image, W is the image pixel width, H is the image pixel height, FOV h is the horizontal field of view of the camera.

[0123] (3) Camera pose true value conversion. The coordinate system in Unreal Engine is a left-handed system and the camera pose is represented by coordinate position and Euler angle. The reconstruction algorithm generally uses a right-handed system and a transfer matrix to represent the camera pose. Use Python's scipy library to convert Euler angles into rotation matrices. The axis correspondence from left-handed to right-handed coordinate systems is x to z, y to x, and z to -y.

[0124] In some embodiments, the output of the intraoral scanner motion sensor may be simulated by the actual camera pose, including the following steps:

[0125] (a) Calculate linear and angular velocities from the true pose output;

[0126] (b) simulated accelerometer data, calculating the linear acceleration from the linear velocity obtained in (a) and adding Gaussian noise or the fixed offset of the sensor;

[0127] (c) Simulated gyroscope data. The gyroscope directly measures the angular velocity and adds Gaussian noise or a fixed offset of the sensor to the angular velocity obtained in (a).

[0128] (4) Noise addition. In some embodiments, a depth noise of an active structured light pattern is added to the original depth map. Specifically, the noise pattern is similar to that of a Kinect structured light camera, and its main feature is that data at the edge of an object with a large depth difference cannot be obtained. This is accomplished by the following steps:

[0129] (a) adding a random xy offset to the input depth image, the pattern of the random offset may be normally distributed, and calculating a disparity map from the depth map according to the camera parameters;

[0130] (b) Using the speckle image of the Kinect camera as a mask, a sliding window is used to traverse the disparity image and the mask image;

[0131] (c) If the pixel mask of the center of the window is greater than 0 and the number of pixels with a mask greater than 0 in the window exceeds the threshold thresh1, the processing flow is entered;

[0132] (d) Calculate the mean of the valid pixels in the window, and calculate the difference between each pixel and the mean and weight them. The farther away from the center, the lower the weight. Mark the points whose difference is less than the threshold thresh2 as inliers;

[0133] (e) Verify the number of inliers in the window of step (d). If the number exceeds the threshold thresh3, it is considered credible and the input disparity of the center point of the window is updated to the output;

[0134] (f) Fill the output disparity of the invalid pixels in step (c), and the criterion is to select the input disparity of the nearest valid center point as the output;

[0135] (g) Add Gaussian noise to the disparity obtained in step (f) and recalculate the depth map using the same camera parameters as in step (a).

[0136] In some embodiments, when scanning edentulous implants, the gingival mucosal tissue between the implants lacks splicing features. Stronger noise can be added to the red gingival tissue, which can be achieved by: increasing the variance of the normal distribution in step (a); reducing the threshold thresh2 in step (d); increasing the variance of the Gaussian noise in step (g), etc.

[0137] The preferred specific embodiments of the present invention are described in detail above. It should be understood that ordinary technicians in the field can make many modifications and changes based on the concept of the present invention without creative work. Therefore, all technical solutions that can be obtained by technicians in the technical field based on the concept of the present invention through logical analysis, reasoning or limited experiments on the basis of the prior art should be within the scope of protection determined by the claims.

Claims

1. A method for constructing a simulation data acquisition platform for digital impression technology research, characterized in that: include: (1) Use Unreal Engine to import the collected objects and build the scanning environment; (2) Construct a blueprint class in Unreal Engine to represent the intraoral scanner, and use the existing intraoral scanners as a reference to construct and design a new intraoral scanner for research; (3) Based on the designed new intraoral scanner, develop control and data acquisition logic and combine plug-ins to collect multimodal data; (4) Post-process the obtained raw data.

2. The method for constructing a simulation data acquisition platform for digital impression technology research according to claim 1, characterized in that: The acquisition object in step (1) includes a dental model with or without teeth, and the scanning environment includes any one of a simulated laboratory scanning environment, an intraoral scanning environment, or an intraoral scanning interference environment.

3. The method for constructing a simulation data acquisition platform for digital impression technology research according to claim 1, characterized in that: The existing intraoral scanner in step (2) includes a single-view window scanner; The novel intraoral scanner includes any one or more of a dual-viewing window intraoral scanner, a triple-viewing window intraoral scanner or a tray-type intraoral scanner.

4. The method for constructing a simulation data acquisition platform for digital impression technology research according to claim 1, characterized in that: Step (3) of formulating the control and data acquisition logic requires the input device to accept the control instructions; In step (3), the control logic includes scanner movement, data acquisition instructions or view angle switching; In step (3), the collected data includes window pose and image.

5. The method for constructing a simulation data acquisition platform for digital impression technology research according to claim 1, characterized in that: The image collected in step (3) includes color and depth information.

6. The method for constructing a simulation data acquisition platform for digital impression technology research according to claim 1, characterized in that: The post-processing in step (4) includes: (A) Depth conversion; (B) Camera intrinsic parameter calculation; (C) Pose truth conversion; (D) Depth noise addition.

7. The method for constructing a simulation data acquisition platform for digital impression technology research according to claim 6, characterized in that: The formula used for the depth conversion in step (A) is as follows: Among them, d plane is the plane depth, d ray is the depth obtained, d pixel is the pixel distance from the midpoint of the field of view to the optical center in the imaging result, and f is the focal length of the camera.

8. The method for constructing a simulation data acquisition platform for digital impression technology research according to claim 6, characterized in that: The formula of the internal reference in step (B) without considering the error is as follows: Where f is the focal length of the camera, C x is the optical center position on the x-axis of the image, C y is the optical center position on the y-axis of the image, w is the image pixel width, H is the image pixel height, FOV h is the horizontal field of view of the camera.

9. The method for constructing a simulation data acquisition platform for digital impression technology research according to claim 6, characterized in that: The pose truth value in step (C) includes the conversion of the left-hand system and Euler angle representation in the Unreal Engine to the right-hand system and transfer matrix representation commonly used for reconstruction; In step (C), the true pose value is converted into motion sensor output.

10. The method for constructing a simulation data acquisition platform for digital impression technology research according to claim 6, characterized in that: The depth noise pattern in step (D) complies with actual imaging principles, which include active structured light, confocal or active wavefront sampling.

Images