Data processing method, scanning device, computing device, and storage medium

Abstract

The application provides a data processing method, a scanning device, a computing device and a storage medium. The method comprises: obtaining source data generated by a scanning device, wherein the source data at least comprises a point cloud generated when the scanning device scans an object; and in response to generating one or more first analyses of the scanning based on the point cloud, performing desensitization on one or more sensitive first analyses of the object revealed in the one or more first analyses, wherein the desensitization makes the one or more sensitive first analyses no longer completely reveal the object. The above method can improve data processing efficiency.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of data processing, and in particular to a data processing method, a scanning device, a computing device and a storage medium. BACKGROUND

[0002] In the application of three-dimensional scanning technology, it is usually required to analyze, for example, scanning point cloud data and operation logs. However, the results of the above analysis have the problems of privacy leakage and compliance risk. SUMMARY

[0003] A first aspect of the embodiments of the present application provides a data processing method, comprising: obtaining source data generated by a scanning device, wherein the source data at least includes point cloud generated when the scanning device scans an object; in response to generating one or more first analyses of the scanning based on the point cloud, performing desensitization point cloud on one or more sensitive first analyses of the object revealed in the one or more first analyses, wherein the desensitization makes the one or more sensitive first analyses no longer completely reveal the object.

[0004] In some embodiments, the one or more sensitive first analyses include a category analysis of the object, and performing desensitization on the category analysis of the object includes: replacing the category analysis with a desensitized category associated with the category analysis; the one or more sensitive first analyses include a visual effect analysis of the object, and performing desensitization on the visual effect analysis of the object includes: replacing at least part of the point cloud representing one or more visual effects of the category analysis in the point cloud with one or more desensitized point clouds associated with the category analysis, wherein the one or more desensitized point clouds correspond to changing the actual presentation of the one or more visual effects.

[0005] In some embodiments, the category analysis of the object is obtained by using a first neural network to recognize the point cloud, comprising: performing a structured conversion on the point cloud to obtain a converted point cloud; obtaining a plurality of local features by the first neural network based on the converted point cloud; aggregating the plurality of local features to obtain a global feature; and outputting, by the first neural network, a category analysis probability of the object based on the global feature; determining the category analysis of the object based on the category analysis probability.

[0006] In some embodiments, the structured conversion of the point cloud to obtain the converted point cloud comprises: voxelizing the point cloud to obtain a voxelized point cloud; or performing multi-view projection processing on the point cloud to obtain a multi-view projection processed point cloud.

[0007] In some embodiments, the category analysis of the object is obtained using a first neural network, including: constructing a graph structure corresponding to the point cloud; obtaining node features based on the graph structure through the first neural network; performing feature aggregation on the node features to obtain global features; and outputting, by the first neural network, a category analysis probability of the object based on the global features; and determining the category analysis of the object based on the category analysis probability.

[0008] In some embodiments, further comprising: determining, based on the category analysis, the at least partial point cloud representing one or more visual effects of the category analysis from the point cloud; and generating, based on the at least partial point cloud and the category analysis, one or more desensitized point clouds using a second neural network.

[0009] In some embodiments, the source data further includes an operation log generated when the scanning device scans an object, and the method further comprises: in response to generating one or more second analyses related to the operation based on the operation log, performing desensitization on one or more sensitive second analyses that reveal an operator related to the operation in the one or more second analyses, wherein the desensitization causes the one or more sensitive second analyses to no longer reveal the operator.

[0010] In some embodiments, further comprising: being applied to an associated computing device in communication connection with the scanning device, and displaying, on the associated computing device, the one or more first analyses after the desensitization is performed.

[0011] In some embodiments, further comprising: being applied to the scanning device, and displaying, on the scanning device, the one or more first analyses after the desensitization is performed.

[0012] In some embodiments, further comprising: in response to a data transmission operation of the scanning device to an associated computing device, obtaining source data generated by the scanning device; or, in response to an operation of the scanning device requesting generation of a log analysis report, obtaining source data generated by the scanning device; or, in response to an operation of the scanning device requesting generation of a data analysis report, obtaining source data generated by the scanning device.

[0013] In some embodiments, after replacing at least partial point cloud representing one or more visual effects of the category analysis in the point cloud with one or more desensitized point clouds associated with the category analysis, further comprising: in response to a data transmission operation of the scanning device to an associated computing device, obtaining the point cloud after the at least partial point cloud is replaced; sending the replaced point cloud to the associated computing device; and receiving a quality evaluation result of the replaced point cloud from the associated computing device.

[0014] In some embodiments, after replacing at least part of the point cloud representing the one or more visual effects of the category analysis in the point cloud with the one or more desensitized point clouds associated with the category analysis, the method further comprises, in response to the operation of the scanning device requesting generation of a data analysis report, obtaining the point cloud in which the at least part of the point cloud has been replaced; analyzing global features of the replaced point cloud using a regression model to obtain a quality assessment result; and generating the data analysis report based on the quality assessment result.

[0015] In some embodiments, the desensitization causes the one or more sensitive first analyses to no longer fully reveal the object, including:

[0016] The one or more desensitized first analyses reveal one aspect of the object after desensitization, wherein the one aspect includes one or more of a material, a size, a color, and a weight of the object.

[0017] A second aspect of the embodiments of the present application provides a scanning device, comprising a processor and a memory, wherein the processor is configured to implement the data processing method according to any one of the first aspect.

[0018] A third aspect of the embodiments of the present application provides a computing device, comprising a processor and a memory, wherein the memory stores a computer program, and the processor is configured to implement the data processing method according to any one of the first aspect when executing the computer program.

[0019] A fourth aspect of the embodiments of the present application provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program, and the computer program is configured to implement the data processing method according to any one of the first aspect when executed by a processor.

[0020] In the data processing method provided by the present application, source data generated by a scanning device is obtained to provide a data basis for subsequent data processing, wherein the source data at least includes a point cloud generated when the scanning device scans an object. In response to generating one or more first analyses of the object based on the point cloud, desensitization is performed on one or more sensitive first analyses of the one or more first analyses that reveal the object, wherein the desensitization causes the one or more sensitive first analyses to no longer fully reveal the object. The purpose of avoiding privacy data leakage caused by analyzing the point cloud is achieved. BRIEF DESCRIPTION OF DRAWINGS

[0021] Figure 1 is an exemplary application scenario diagram of the data processing method provided by an embodiment of the present application.

[0022] Figure 2is a structural schematic diagram of a scanning device provided by an embodiment of the present application.

[0023] Figure 3 is a flowchart of a data processing method provided by an embodiment of the present application.

[0024] Figure 4 is a schematic diagram of a pre-de-identification analysis interface of an object based on point cloud provided by an embodiment of the present application.

[0025] Figure 5 is a schematic diagram of a post-de-identification analysis interface of an object based on point cloud provided by an embodiment of the present application.

[0026] Figure 6 is a schematic diagram of a de-identified point cloud provided by an embodiment of the present application.

[0027] Figure 7 is a schematic diagram of a comprehensive report provided by an embodiment of the present application.

[0028] Figure 8 is a detection flowchart of category analysis of an object provided by an embodiment of the present application.

[0029] Figure 9 is a detection flowchart of category analysis of an object provided by another embodiment of the present application.

[0030] Figure 10 is a structural diagram of a computer device provided by an embodiment of the present application. DETAILED DESCRIPTION

[0031] For the convenience of understanding, exemplary descriptions of some concepts related to the embodiments of the present application are given for reference.

[0032] It should be noted that “at least one” in the present application means one or more, and “multiple” means two or more than two. “And / or” describes the association relationship of the associated objects, which means that there can be three kinds of relationships, for example, A and / or B can represent the following three cases: A exists alone, A and B exist together, and B exists alone, where A and B can be singular or plural. The terms “first”, “second”, “third”, “fourth” and the like (if any) in the description and claims of the present application and the drawings are used to distinguish similar objects, and are not used to describe a specific order or sequence.

[0033] In the application of three-dimensional scanning technology, it is usually required to analyze, for example, scanning point cloud data, operation logs, etc. However, the above analysis results may contain relevant information of the scanning object and / or the operator, and there is a problem of privacy leakage and compliance risk.

[0034] The embodiment of the present application provides a data processing method, a scanning device, a computing device and a storage medium, which can analyze point cloud, desensitize sensitive analysis generated to reveal a scanning object, so that the analysis result no longer reflects the scanning object, and leakage of the scanning object is avoided. First, an exemplary application scenario of the data processing method of the present application is described.

[0035] Figure 1 is an exemplary application scenario diagram of the data processing method provided by one embodiment of the present application. The data processing method provided by the embodiment of the present application can be applied to the scanning device 10, and can also be applied to the computer device 20 (also referred to as an associated computing device herein) in communication connection with the scanning device 10. It can be understood that, Figure 1 The examples do not limit the scanning device 10 to be connected and used with the computer device 20. The scanning device 10 can also be used as a scanning and processing device alone, without the help of any computer device 20.

[0036] The communication connection mode can include a wired communication connection mode and a wireless communication connection mode. The wired communication connection mode can include one or more of a universal serial bus (USB), a controller area network bus (CAN), and the like. The wireless communication connection mode can include one or more of wireless fidelity (Wi-Fi), Bluetooth (BT), a mobile (cellular) communication network, frequency modulation (FM), near field communication (NFC), infrared technology (IR), and the like.

[0037] The scanning device 10 can include, but is not limited to, an oral scanning device, a facial scanning device, a CT (Computed Tomography, electronic computed tomography) scanning device or a CBCT (Cone beam Computer Tomography, oral and maxillofacial cone beam CT) scanning device, a professional scanner, an industrial scanner, and the like. The oral scanning device includes an intraoral scanner and an extraoral scanner. The scanning device 10 can be a handheld scanning device or a fixed scanning device. The scanning device 10 can realize three-dimensional reconstruction of teeth, faces, bodies, industrial products, industrial equipment, cultural relics, artworks, prostheses, medical devices, buildings and the like. The present application does not make specific limitations.

[0038] Figure 2A structural diagram of an exemplary scanning device 10 is shown. In the depicted embodiment, the scanning device 10 includes a frame structure 20 of the scanning device 10, imaging modules 30 located at the frame structure 20, and a display screen 40. The imaging modules 30 can be arranged staggered to each other such that the field of view of each imaging module at least partially overlaps. In some embodiments, the imaging modules 30 can include five cameras, i.e., a first camera 31, a second camera 32, a third camera 33, a fourth camera 34, and a fifth camera 35. The imaging modules 30 can also include two groups of light projectors, i.e., a first group of light projectors 36 and a second group of light projectors 37, each of which can include one or more light projectors, e.g., including three light projectors. Each light projector can include a light source configured to emit a light beam to a collimating lens, the collimated light beam propagating to a diffractive optical element, and the diffractive optical element replicating the focused light beam to form a speckle pattern or a fringe pattern and projecting to a scanning object / region, the reflected speckle pattern or fringe pattern from the scanning object / region being collected by the imaging modules 30 to further obtain three-dimensional information of the scanning object / region by a processing chip (not shown in the figure). In other embodiments, the light projectors can be image projectors, such as digital micromirror devices, liquid crystal display projectors, or organic electroluminescent display projectors, etc.

[0039] In some embodiments, the light projectors of the first group of light projectors 36 can include a single same light source, e.g., a light source emitting one of infrared light, white light, blue light, or other visible monochromatic light. In other embodiments, the light projectors of the first group of light projectors 36 are all configured to emit light with a wavelength between 405 nm and 1100 nm, e.g., all configured to emit near-infrared light with a wavelength between 780 nm and 1100 nm. In other embodiments, the light projectors of the first group of light projectors 36 can include multiple different light sources, e.g., one or two light projectors including a light source emitting infrared light, e.g., one or two light projectors including a light source emitting near-infrared light with a wavelength between 780 nm and 1100 nm, and another light projector including a light source emitting blue light or white light. For example, a first light projector of the first group of light projectors 36 includes a first light source emitting near-infrared light, a second light projector of the first group of light projectors 36 includes a second light source emitting white light, a third light projector of the first group of light projectors 36 includes a third light source emitting blue light, or a first light projector of the first group of light projectors 36 includes a first light source emitting near-infrared light, a second light projector of the first group of light projectors 36 includes a second light source emitting near-infrared light, and a third light projector of the first group of light projectors 36 includes a third light source emitting white light / blue light. In some examples, the multiple projectors of the first group of light projectors 36 can be implemented as separate units, or as separate units in one area.

[0040] In some embodiments, the light projectors of the second light projector group 37 can comprise a single identical light source, e.g. a light source emitting infrared light, white light, blue light or one of the other visible monochromatic lights. In other embodiments, the light projectors of the second light projector group 37 are all configured to emit light having a wavelength between 405 nm and 1100 nm, e.g. all configured to emit blue light having a wavelength between 435 nm and 480 nm. In other embodiments, the light projectors of the second light projector group 37 can comprise a plurality of different light sources, e.g. one or two light projectors comprising a light source emitting blue light, e.g. one or two light projectors comprising a light source emitting blue light having a wavelength between 435 nm and 480 nm, and further light projectors comprising a light source emitting near infrared light or white light. For example, a first light projector of the second light projector group 37 comprises a first light source emitting blue light, a second light projector of the second light projector group 37 comprises a second light source emitting white light, a third light projector of the second light projector group 37 comprises a third light source emitting near infrared light, or a first light projector of the second light projector group 37 comprises a first light source emitting blue light, a second light projector of the second light projector group 37 comprises a second light source emitting blue light, and a third light projector of the second light projector group 37 comprises a third light source emitting white / near infrared light. In some examples, the plurality of projectors of the second light projector group 37 can be implemented as separate units, or as separate units in one area.

[0041] In some embodiments, the first light projector group 36 and the second light projector group 37 can be configured to be selectively turned on or off in adaptation to different scanning modes of the scanning device 10.

[0042] The type of camera of the first camera 31, the second camera 32, the third camera 33 and the fourth camera 34 is typically a monochrome (e.g. black and white) camera, and depends on the type of light source(s) used in the first light projector group 36 and the second light projector group 37. In some embodiments, the first camera 31, the second camera 32, the third camera 33 and the fourth camera 34 can be monochrome, visible color spectrum or near infrared cameras, and the light projectors of the first light projector group 36 are infrared light projectors or near infrared light projectors, and the light projectors of the second light projector group 37 are blue light projectors.

[0043] In some embodiments, a first camera 31 and a second camera 32 are used together as a first camera combination, and a third camera 33 and a fourth camera 34 are used together as a second camera combination. The first camera 31 and the second camera 32 are configured to have the same focal length, e.g., a closer focal length or a farther focal length, and the third camera 33 and the fourth camera 34 are configured to have the same focal length, e.g., a farther focal length or a closer focal length, and the focal lengths of the first camera combination and the second camera combination are not the same. For example, the first camera combination or the second camera combination can be selected to capture the reflections on the object from either a set of light projectors based on the adapted focal length, and the light projector set used can also be selected based on, for example, the scanning speed, scanning accuracy required by the environment, and the characteristics of the object.

[0044] In some embodiments, the first camera 31, the second camera 32, the third camera 33, and the fourth camera 34 may use any suitable shutter technology, including but not limited to rolling shutters, global shutters, mechanical shutters, and optical liquid crystal display (LCD) shutters. In some embodiments, the fifth camera 35 may be a color camera (also known as a texture camera). The texture camera may use any suitable shutter technology, including but not limited to rolling shutters, global shutters, mechanical shutters, and optical liquid crystal display (LCD) shutters. In some embodiments, the first camera 31, the second camera 32, the third camera 33, the fourth camera 34, and the fifth camera 35 may have similar configurations to improve matching confidence and speed. In some embodiments, the imaging module 30 may also be able to use fewer projector groups, such as only one projector group, and fewer cameras, such as only two cameras, to complete the acquisition of reflected light and color textures, for example, omitting the second projector group 37 (first projector group 36), and omitting the third camera 33 (first camera 31) and the fourth camera 34 (second camera 32).

[0045] like Figure 1 As shown, the first camera 31, the second camera 32, and the first light projector group 36 can be located side-by-side on one surface of the frame structure 20, with the first camera 31, the second camera 32, and the first light projector group 36 spaced apart from each other and all facing directly forward of the surface. The third camera 33, the fourth camera 34, and the second light projector group 37 can be located side-by-side on one surface of the frame structure 20, with the third camera 33, the fourth camera 34, and the second light projector group 37 spaced apart from each other and all facing directly forward of the surface.

[0046] The data connection (such as a serial communication connection) between the scanning device 10 and one or more computer processors (not shown in the figure) can allow the transfer of data collected by the first camera 31, the second camera 32, the third camera 33, the fourth camera 34 and the fifth camera 35 so that it can be processed to derive 3D measurements of the surface of the object / object being scanned. The one or more computer processors can be implemented in a remote computing system (electronic device) or, alternatively, can be part of the scanning device 10 itself.

[0047] Exemplarily, each light projector group can comprise a single light projector or can also have two or more (e.g. three) light projectors. The light projectors can be configured to project visible or non-visible light, coherent or non-coherent light. In some embodiments, the light projectors can comprise one or more light sources consisting of lasers (e.g. Vertical Cavity Surface Emitting Lasers (VCSELs), Edge Emitting Lasers (EELs), solid-state lasers and semiconductor lasers) and / or one or more LEDs (or OLEDs).

[0048] The light projectors can be configured to project a structured light pattern consisting of a plurality of light patches arranged side by side to each other, which can appear as elongated light fringe bands when the light patches are projected onto the surface of the object. The elongated light fringe bands are non-intersecting elongated light fringe bands and, in some implementations, the elongated light fringe bands can be substantially parallel to each other. In some embodiments, the light projectors can be programmable light projector units, which are capable of projecting more than one light pattern. For example, the light projectors can be configured to project different structured line patterns. In some embodiments, the light projectors can emit light having a wavelength between 405 nm and 1100 nm.

[0049] In some examples, the two images of the object can be captured simultaneously using the first camera 31, the second camera 32 or using the third camera 33, the fourth camera 34, an image processing can be applied to a computational method implemented, for example, by one or more processors, or to a computational method implemented, for example, by an electronic device to derive 3D measurements of the surface of the object / object being scanned.

[0050] In some examples, a texture of the object can be captured using the second camera 32 simultaneously to the capture of the object images by the first camera 31, the second camera 32 or by the third camera 33, the fourth camera 34, and the texture can be applied to a computational method implemented, for example, by one or more processors, or to a computational method implemented, for example, by an electronic device to map on the 3D measurements of the surface of the object / object being scanned.

[0051] In some examples, using a film / membrane with band-pass filter function fixed on the lens of the camera (e.g., the first camera 31, the second camera 32, the third camera 33, and the fourth camera 34) can match the wavelength of the projector(s), which can help reduce light source interference from ambient light and other projectors.

[0052] In some examples, the intrinsic and extrinsic parameters of the first camera 31, the second camera 32, the third camera 33, the fourth camera 34, and the fifth camera 35 are measured using a calibration board with a geometric distance true value measured in advance by a high-precision method such as photogrammetry. The measurement process usually includes a series of continuous measurements using the calibration camera of the scanning device 10 after adjusting the calibration board to different positions, and calculating the spatial position and orientation of the first camera 31, the second camera 32, the third camera 33, the fourth camera 34, and the fifth camera 35 by identifying the positions of the reference (marker) points / areas / lines in the calibration images, thereby completing the calibration of the intrinsic and extrinsic parameters of the first camera 31, the second camera 32, the third camera 33, the fourth camera 34, and the fifth camera 35.

[0053] In some examples, the display screen 40 is configured to display a three-dimensional model of the object captured by the imaging module 30, and is further configured to display the analysis results of the source data generated by the captured object. The display screen 40 can be arranged opposite to the imaging module 30, and can include, for example, a touch function or physical buttons / touch areas attached to the surrounding area of the display screen 40, which can be used to touch and move, zoom, or turn on / off the three-dimensional model of the object displayed on the display screen 40. In some examples, the display screen 40 can be a liquid crystal display (LCD), an organic light-emitting diode (OLED), an active-matrix organic light-emitting diode (AMOLED), a flexible light-emitting diode (FLED), a Mini led, a Micro Led, a Micro-oLed, a quantum dot light emitting diode (QLED), etc.

[0054] The computer device 20 can be a mobile phone, a tablet computer, a smart wearable device, an Augmented Reality (AR) / Virtual Reality (VR) device, a notebook computer, a netbook, etc. The embodiments of the present application do not limit the specific type of the computer device 20. In one example, the method of analyzing the source data described in the embodiments of the present application is executed by the scanning device 10.

[0055] In another example, the computer device 20 is configured to receive the source data sent by the scanning device 10 and analyze the source data. In some examples, the method of analyzing the source data can be executed by the scanning device 10 and the computer device 20. In addition, the computer device 20 can also send control instructions to the scanning device 10 so that the scanning device 10 can perform corresponding operations. The present application does not limit the interaction between the scanning device 10 and the computer device 20. The specific structure of the computer device 20 can be referred to the following Figure 10

[0056] It can be understood that the structure illustrated in the embodiments of the present application does not constitute a specific limitation on the scanning device 10. In other embodiments of the present application, the scanning device 10 can include more or fewer components than those illustrated, or combine certain components, or split certain components, or different component arrangements. The illustrated components can be implemented in hardware, software, or a combination of software and hardware. For example, the scanning device 10 can also include a power supply, a control circuit, etc.

[0057] Figure 3 is a flowchart of the data processing method provided by the embodiments of the present application, which is applied in a scanning device (for example, the scanning device 10 of Figure 2 ). According to different requirements, the order of the steps in the flowchart can be changed.

[0058] In step S301, source data generated by the scanning device is obtained.

[0059] In some embodiments of the present application, the source data can include point cloud generated when the scanning device scans an object, operation log generated when the scanning device is operated, etc. The object can be any object or living body that needs to be measured for geometric shape, for example, a car, a medical device, a product part, a tooth, a human model, etc. The present application does not limit the type and shape of the object scanned by the scanning device. In some examples, in order to accurately detect the object to be scanned, a suitable amount of marker points can be pasted on the object before scanning the object. The marker points can be reflective marker points or coded marker points, etc.

[0060] ​Step S302, in response to one or more first analyses generated based on point cloud scanning, desensitization is performed on one or more sensitive first analyses that reveal objects in one or more first analyses.

[0061] For example, see Appendix Figure 4 The diagram illustrates a pre-analysis interface 50a for point cloud-based object generation according to an embodiment of this application. The analysis interface 50a includes one or more first analyses 501 generated from the point cloud for scanning, such as effective scan rate analysis and data integrity analysis. It includes a sensitive first analysis 501a that correctly discloses, reveals, or indicates the object. The analysis interface 50a can be displayed on, for example, the display screen 40 of the scanning device 10, or on an associated computing device XX, leading to leakage of object information. It is understood that the scan used to generate one or more first analyses 501 can refer to the current scan round, all completed scan rounds, or any one or more scan rounds selected from all scan rounds. One or more analyses 501 can be analyses of the scan from multiple directions / dimensions, such as data integrity reflecting scan quality, effective scan rate, rescan rate, etc., which are not limited herein.

[0062] This application embodiment performs desensitization processing on the sensitive first analysis 501a, which will prevent the sensitive first analysis 501a from completely revealing the object. For example, see the appendix. Figure 5 The diagram illustrates a desensitized analysis interface 50b for point cloud-based object generation according to an embodiment of this application. By desensitizing and replacing the results of the sensitive first analysis 501a, the sensitive first analysis 501a will no longer be able to fully reveal the object, but will only reveal a certain type of feature of the object, such as revealing the object's material or other more generalized representations.

[0063] In some embodiments of this application, one or more sensitive first analyses include object category analysis, which is obtained by recognizing the point cloud using a first neural network. The first neural network may be one or a combination of several of the following: Convolutional Neural Network (CNN), Graph Neural Network (GNN), Fully Connected Neural Network (FCN / DNN), Recurrent Neural Network (RNN), Transformer network (such as Visual Transformer (ViT), a network with only an encoder structure, or a network with only a decoder structure).

[0064] The category analysis of the object is used to replace the associated desensitization category, the one or more sensitive first analysis includes visual effect analysis of the object, and the visual effect analysis of the scanned object is desensitized. In an embodiment, one or more desensitization point clouds associated with the category analysis replace at least part of the point cloud representing one or more visual effects of the category analysis in the point cloud, wherein the one or more desensitization point clouds correspond to changing the actual presentation of the one or more visual effects.

[0065] In an example, the first neural network is used to identify features in the point cloud that can present visual effect analysis, such as structural features and texture features. Structural features can include planar features, curved surface features, edge features, etc., and texture features can be color features.

[0066] By identifying the structural features and / or texture features of the point cloud, the category analysis of the object is obtained, which is used to distinguish the attributes of the object. For a detailed description of using the first neural network to identify the point cloud to obtain the category analysis, please refer to the following Figure 8 and Figure 9 Embodiments shown.

[0067] In some embodiments of the present application, the one or more desensitization point clouds correspond to changing the actual presentation of the one or more visual effect analysis, such as changing the structure, color, texture, etc. According to the category analysis determined above, at least part of the point cloud representing one or more visual effects of the category analysis is screened from the point cloud. For example, the category analysis is a vase, and at least part of the point cloud representing one or more visual effects of the vase is screened from the point cloud. Wherein the one or more desensitization point clouds correspond to changing the actual presentation of the one or more visual effects.

[0068] Based on the at least part of the point cloud representing one or more visual effects of the vase, a second neural network associated with the category analysis is called to process the at least part of the point cloud representing one or more visual effects of the vase, to generate one or more desensitization point clouds. Wherein the second neural network corresponding to each category analysis can be pre-trained, for example, if the category analysis is determined to be a vase, the second neural network corresponding to the vase is called. The second neural network can be one or more of a generative adversarial network (GAN), a conditional diffusion model, an algorithm model optimized for a specific task (such as a beta-variational autoencoder (β-VAE)).

[0069] In one example, taking vase point cloud data, at least partially representing a vase, as an example, the vase point cloud data, labeled as "vase" in category analysis, is input into a conditional diffusion model. Strong denoising is applied to the vase point cloud data, destroying textures (such as color) and local structures (such as sculpted patterns and embossed textures). Then, based on the category analysis condition constraints corresponding to the category analysis labels, the basic geometric contour of the vase is restored. Finally, a featureless vase with a clear geometric contour is generated as a desensitized point cloud. Combined with... Figure 6 As shown, for ease of description, the vase point cloud data and the anonymized point cloud are presented in a visual manner. Figure 6 (a) represents the 3D model corresponding to the vase point cloud data, such as Figure 6 (b) represents the 3D model of the desensitized point cloud. The conditional diffusion model is used to... Figure 6 After processing (a), the result is as follows: Figure 6 (b)

[0070] In another example, taking vase point cloud data, at least partially representing a vase, as an example, the vase point cloud data with category analysis labels of "vase" is input into a β-variable autoencoder. Based on the category analysis labels, the corresponding category analysis template is called to remove local features representing blue-and-white patterns, relief carvings, and surface scratches from the vase point cloud data. The vase's curves are then modified to match the standard spline curves in the category analysis template, finally generating a standard vase as the desensitized point cloud. Combined with... Figure 6 As shown, for ease of description, the vase point cloud data and the anonymized point cloud are presented in a visual manner. Figure 6 (c) represents the desensitized point cloud for a 3D model. A β-variable autoencoder is used to... Figure 6 After processing (a), the result is as follows: Figure 6 (c)

[0071] The above embodiments can avoid the leakage of privacy data due to direct analysis of point clouds, thereby improving the security of the scanned data.

[0072] In other embodiments of this application, if the scanning device receives a data request from the associated computing device, or if the scanning device detects a user-triggered request to transmit data to the associated computing device, then the scanning device responds to the data transmission operation to the associated computing device by performing the above-described actions. Figure 3 The embodiment shown. Alternatively, after acquiring the point cloud, the scanning device can directly execute the above-described embodiment. Figure 3 In the illustrated embodiment, then in response to a data transfer operation to the associated computing device, the above-mentioned data is directly obtained. Figure 3 The final result of the illustrated embodiment is to obtain the point cloud corresponding to the point cloud that has been replaced at least partially.

[0073] The scanning device sends the replaced point cloud to the associated computing device, i.e., the at least part of the point cloud representing one or more visual effects of the object category analysis is replaced by one or more desensitization point clouds. The scanning device receives the quality evaluation result of the replaced point cloud from the associated computing device, which can be, but is not limited to, completeness, noise level, hole rate, etc.

[0074] In other embodiments of the present application, the scanning device includes a display screen on which a plurality of virtual controls can be displayed, and in addition, the scanning device can also be provided with a plurality of physical buttons. The user can click any virtual control to perform the corresponding function, or operate any physical button to perform the corresponding operation. When the scanning device detects the user's operation on the virtual control representing the generation of the data analysis report, the above-mentioned embodiments as shown in Figure 3 Or, the scanning device directly executes the above-mentioned embodiments as shown in Figure 3 after collecting the point cloud, and then directly obtains the final result of the above-mentioned embodiments as shown in Figure 3

[0075] The scanning device calls the regression model to analyze the global features of the replaced point cloud to obtain the quality evaluation result. The regression model can be one or a combination of linear regression, random forest regression, support vector regression (SVR), and neural network regression.

[0076] In other embodiments of the present application, the source data also includes an operation log generated when the scanning device scans the object. The operation log includes, but is not limited to, a timestamp (recording the exact time of each operation), a device state (such as start / stop, calibration result, power or temperature warning), a scanning parameter (resolution, laser power, exposure time, scanning range, etc.), a user operation (start / pause scanning, mode switching, file save path), environmental data (temperature, humidity, lighting conditions), positioning information (device position, attitude angle, positioning target state), data quality feedback (real-time point cloud density, missing point rate, scanning interruption reason), error code (hardware failure, communication exception, motion overrun), metadata (scanning target identification, object name, operator identification), and personal data of the operator using the scanning device, etc.

[0077] ​In addition to the replacement of the point cloud, the data in the operation log can also be replaced. In response to generating one or more second analyses related to the operation based on the operation log, desensitization is performed on one or more sensitive second analyses that reveal the operator generating the operation-related one or more second analyses, wherein after the desensitization is performed, the one or more sensitive second analyses no longer reveal the operator. In the case that the data recorded in the operation log has a specified format, the scanning device can call a pre-set keyword to find data matching the keyword in the operation log. In an example, the called keyword includes personal data, and the position of the personal data in the operation log is located and the value carried thereafter is obtained. For example, the keyword includes the name, and the position of the name in the operation log is located, and the value carried thereafter is "Zhang San".

[0078] In the case that the data recorded in the operation log does not have a specified format, or in the case that the format cannot be known, the operation log can be identified by using a machine learning algorithm. The machine learning algorithm can include one or more combinations of a bidirectional long short-term memory-conditional random field (BiLSTM-CRF), a bidirectional encoder representation from transformers, and a support vector machine (SVM).

[0079] In some embodiments of the present application, one or more desensitization data for performing desensitization on one or more sensitive second analyses that reveal the operator generating the operation-related one or more second analyses can be pre-set in the scanning device, and the one or more sensitive second analyses are replaced by using the one or more desensitization data. An association relationship between the personal data and the one or more desensitization data is established in the scanning device. The desensitization data can be one or more combinations of numbers, characters, letters, codes, specified words, and special characters, and the present application does not limit the number format and data type of the desensitization data.

[0080] In an example, the desensitization data is AAA, and the association relationship between AAA and the personal data is established, and after the personal data representing "name: Zhang San" is identified, it is replaced by "name: AAA".

[0081] In other embodiments of the present application, other data in the operation log other than the personal data can also be desensitized, for example, the name of the object recorded in the operation log can be desensitized. In the scanning device, the desensitization data associated with the name of the object is set in advance.

[0082] In the scanning device, the association between the name of the object and the desensitization data associated with the name of the object is established. The desensitization data associated with the name of the object can be one or a combination of multiple of numbers, characters, letters, codes, specified words, and special characters, and the present application does not limit the data format and data type of the desensitization data associated with the name of the object.

[0083] In an example, the desensitization data associated with the name of the object is CCC, and the association between CCC and the name of the object is established. After identifying that the name of the object represents "object name: vase", it is replaced with "object name: CCC".

[0084] The above replacement of the personal data and the name of the object in the operation log is only an example, and the present application does not limit this. Other data in the operation log can also be replaced, for example, the scanning parameters.

[0085] Through the above embodiments, it can be ensured that the replaced operation log does not contain private data, and the privacy security of the data is improved.

[0086] In other embodiments of the present application, in response to the operation of the scanning device requesting to generate a log analysis report, the source data generated by the scanning device is obtained, so that the operation of desensitizing the operation log in the source data is performed. When the scanning device detects the operation of the user on the virtual control representing the generation of the log analysis report, the operation of desensitizing the operation log in the source data is performed. Further, the operation log in which the personal data and / or the name of the object have been replaced is obtained. The operation type and the operation time of each sub-operation in the replaced operation log are determined, for example, the operation type includes the scanning operation, and the operation time of each time the scanning operation is performed is obtained. Based on the operation time of each operation type, the analysis result of the operation continuity of each operation type is obtained. For example, the analysis result of the scanning operation continuity is obtained according to the sorting of the operation time.

[0087] In addition, the operation time efficiency value can be obtained according to the set of operation times of all operation types and the set of historical operation times of the scanning device scanning the same object. For example, the set of current obtained operation times is compared with the set of historical operation times to obtain a comparison result. If the comparison result indicates that the set of current obtained operation times is greater than the set of historical operation times, the operation time efficiency value is negative, indicating that the current efficiency is lower than the historical efficiency. If the comparison result indicates that the set of current obtained operation times is less than the set of historical operation times, the operation time efficiency value is positive, indicating that the current efficiency is higher than the historical efficiency. The scanning device can generate a log analysis report according to the analysis result of operation continuity and the operation time efficiency value.

[0088] In other embodiments of the present application, the scanning device receives a data request sent by the associated computing device, or the scanning device detects a user-triggered demand for data transmission to the associated computing device, and the scanning device performs the operation of de-sensitizing the operation log in response to the data transmission operation to the associated computing device. Alternatively, the scanning device directly performs the operation of de-sensitizing the operation log after generating the operation log, and then directly obtains the operation log with the replaced personal data and object name in response to the data transmission operation to the associated computing device. The replaced operation log is sent to the associated computing device, so that the associated computing device performs data analysis on the replaced operation log, reduces the computing power of the scanning device, and improves the running efficiency. The scanning device receives the log analysis result of the associated computing device on the replaced operation log.

[0089] In other embodiments of the present application, the scanning device can send the point cloud with the replaced at least part of the point cloud and the operation log with the replaced personal data and the object name to the associated computing device at the same time in response to the data transmission operation to the associated computing device, so that the associated computing device simultaneously analyzes the point cloud with the replaced at least part of the point cloud and the operation log with the replaced personal data and the object name, and generates a comprehensive report.

[0090] In addition, the scanning device detects the operation of generating a data analysis report and the operation of generating a log analysis report, and can simultaneously obtain and analyze the point cloud with the replaced at least part of the point cloud and the operation log with the replaced personal data and the object name, and generate a comprehensive report.

[0091] The comprehensive report will be described below. Figure 7 The comprehensive report is an analysis report with replaced sensitive data, such as Figure 7As shown, the comprehensive report includes a scan habit score, a scan frequency, an average scan duration, a scan time, an effective scan rate, a total data volume, a rescan rate, a scan object (may also be referred to as an object name), data integrity, a sharing frequency, a scan posture recommendation, a parameter configuration, and the like. The parameter configuration includes a type of the object, a resolution, a brightness, and a stitching manner.

[0092] Figure 8 is a detection flowchart of the category analysis of the object provided by the embodiments of the present application. As shown in Figure 8 the first neural network can be a convolutional neural network, and the category analysis of the object is obtained by recognizing the point cloud by using the convolutional neural network, including the following steps:

[0093] In step S801, the point cloud is subjected to a structured conversion to obtain a converted point cloud.

[0094] In some embodiments of the present application, since the point cloud has the properties of disorder and irregularity, it needs to be converted into structured data for subsequent processing. The structured conversion of the point cloud can include voxelization processing and multi-view projection processing. The voxelization processing means a process of discretizing a three-dimensional space into a regular small cubic unit (referred to as a voxel) and mapping the point cloud into this three-dimensional grid structure. The multi-view projection processing means projecting three-dimensional points from multiple different perspectives (usually virtual cameras) onto a two-dimensional image plane to generate a set of two-dimensional images (views).

[0095] The voxelization processing of the point cloud obtains a point cloud after voxelization processing as the converted point cloud, including: defining a voxel grid, including defining a spatial range to be occupied by the point cloud and a voxel resolution (a size of each voxel in the three-dimensional space). Each point cloud in the point cloud is traversed, and a voxel grid into which each point cloud falls is determined according to the three-dimensional coordinates of each point cloud. For all point clouds mapped into the same voxel, one or more feature values representing the same voxel are calculated. The calculated feature values are filled into a three-dimensional grid data structure to construct a voxel grid tensor. The voxel grid tensor is taken as the point cloud after voxelization processing, that is, as the converted point cloud.

[0096] The multi-view projection processing of the point cloud obtains a point cloud after multi-view projection processing as the converted point cloud, including: determining how many views need to be generated and from which angles to project, wherein the more views, the more comprehensive the information covered, but the larger the amount of calculation, and too few views may cause key features to be occluded or lost. A virtual camera is defined for each selected view, and virtual camera parameters are set, such as camera position, projection type, image resolution, and the like. For each virtual camera, the point clouds in the point cloud are traversed, and the three-dimensional points are projected onto a two-dimensional image plane based on the virtual camera parameters to generate a view image. The view image is taken as the point cloud after voxelization processing, that is, as the converted point cloud.

[0097] At step S802, a plurality of local features is obtained by the first neural network based on the converted point cloud.

[0098] In some embodiments of the present application, the first neural network can be a convolutional neural network, and the structure of the convolutional neural network at least includes a convolutional layer, a Softmax layer, a pooling layer, and a fully connected layer. The converted point cloud is input into the convolutional neural network, so that a plurality of local features of the converted point cloud, such as the shape, density, contour, and texture of a local region, is extracted by the convolutional neural network.

[0099] At step S803, a global feature is obtained by aggregating the plurality of local features.

[0100] In some embodiments of the present application, after the plurality of local features is determined, the plurality of local features is aggregated, so that all scattered and spatially related local feature information is integrated into a feature vector that can represent the entire object, which is referred to as a global feature.

[0101] At step S804, the first neural network outputs a class analysis probability of the object based on the global feature.

[0102] In some embodiments of the present application, the global feature is processed by one or more fully connected layers and a Softmax layer in the first neural network, and a class analysis probability distribution vector is obtained. For example, for K = 3 (chair, table, and car), the output can be P = [0.85, 0.10, 0.05], which indicates that the first neural network predicts that the class analysis probability of the object is 85% for a chair, 10% for a table, and 5% for a car.

[0103] At step S805, a class analysis of the object is determined based on the class analysis probability.

[0104] In some embodiments of the present application, the maximum class analysis probability is determined from the class analysis probability, and the class analysis corresponding to the maximum class analysis probability is taken as the class analysis of the object scanned by the scanning device. In the above embodiment, the chair corresponding to the class analysis probability of 85% is taken as the class analysis of the object.

[0105] In some embodiments of the present application, desensitization makes one or more sensitive first analyses no longer completely reveal the object, including:

[0106] The one or more sensitive first analyses after desensitization are configured to reveal an aspect of the object, wherein the aspect includes one or more of the material, size, color, and weight of the object.

[0107] It can be understood that one or more sensitive first analyses before desensitization will fully reveal the object, for example, reveal that the object is a specific something / product, for example, by scanning the object field to reveal that the object is a car, a bearing, a vase, etc. The embodiments of the present application can replace the scanned object field with a scanned material field, a scanned size field, a scanned color field, or a scanned weight field so that one or more sensitive first analyses can only reveal one aspect of the object, rather than fully revealing the specific object, which will avoid the leakage of the scanned object caused by the analysis of the scanned data.

[0108] Through the above embodiments, the category analysis of the object can be accurately detected by the convolutional neural network, and the efficiency of calling one or more desensitized point clouds associated with the category analysis is improved.

[0109] Figure 9 is a detection flowchart of the category analysis of the object provided by another embodiment of the present application. As shown in Figure 9 , the first neural network can be a graph neural network, which uses the graph neural network to identify the point cloud to obtain the category analysis of the object, including the following steps:

[0110] Step S901, constructing a graph structure corresponding to the point cloud.

[0111] In some embodiments of the present application, since the point cloud includes an unordered point set, in order to facilitate subsequent processing, the unordered point set can be converted into a graph structure, and the graph structure includes a node set and an edge set, the node set includes the point cloud, and the edge set includes the connection relationship corresponding to the point cloud.

[0112] Step S902, obtaining node features based on the graph structure by the first neural network.

[0113] In some embodiments of the present application, the first neural network is used to learn and represent the features of the nodes in the graph structure, and high-level node features containing local neighborhood structure information are obtained. The multi-layer graph convolution layer (or more generally, the message passing layer) of the first neural network is applied to process the graph structure, and the node features are obtained.

[0114] Step S903, performing feature aggregation on the node features to obtain global features.

[0115] In some embodiments of the present application, the local features of N nodes are aggregated into a single, fixed-length vector, which is used to represent the global features of the entire point cloud. This vector needs to be able to capture the shape and semantic information of the entire object. The feature aggregation method includes but is not limited to one or more of maximum pooling, average pooling, sum pooling, and attention pooling.

[0116] Step S904, the first neural network outputs a class analysis probability of the object based on the global feature.

[0117] In some embodiments of the present application, the global feature is mapped to a probability distribution of the class analysis of the object through one or more fully connected layers and a Softmax function, to obtain the class analysis probability of the object.

[0118] Step S905, determining the class analysis of the object based on the class analysis probability.

[0119] In some embodiments of the present application, the maximum class analysis probability is determined from the class analysis probability, and the class analysis corresponding to the maximum class analysis probability is taken as the class analysis of the object scanned by the scanning device.

[0120] Through the above embodiments, the class analysis of the object can be accurately detected through the graph neural network, and the efficiency of calling one or more desensitized point clouds associated with the class analysis is improved.

[0121] It should be noted that the data processing method described herein can be executed by, for example, a scanning device, and one or more first analyses after desensitization and / or one or more second analyses after desensitization are displayed on the scanning device. The above one or more first analyses after desensitization and one or more second analyses after desensitization can be configured to be displayed together on one display area, for example, one display screen, of the scanning device, or can be displayed on two or more display areas of one display screen of the scanning device.

[0122] The data processing method described herein can also be executed by, for example, an association computing device, and one or more first analyses after desensitization and / or one or more second analyses after desensitization are displayed on the association computing device. The above one or more first analyses after desensitization and one or more second analyses after desensitization can be configured to be displayed together on one display area, for example, one display screen, of the association computing device, or can be displayed on two or more display areas of one display screen of the association computing device.

[0123] Figure 10 is a structural diagram of a computer device provided by an embodiment of the present application. As shown in Figure 10 The computer device 20 can include a display device 210, a communication module 220, a memory 230, a processor 240, an input / output (I / O) interface 250, and a bus 260. The processor 240 is coupled to the display device 210, the communication module 220, the memory 230, and the I / O interface 250 through the bus 260.

[0124] The display device 210 can be a touch screen, which is an inductive touchable liquid crystal display device. Alternatively, the display device 210 can also be a non-touch screen. The display device 210 can be used to display the processing result of the data transmitted by the scanning device 10.

[0125] The communication module 220 can include a wired communication module and / or a wireless communication module. The wired communication module can provide one or more of the following wired communication solutions: Universal Serial Bus (USB), Controller Area Network (CAN) bus, etc. The wireless communication module can provide one or more of the following wireless communication solutions: Wireless Fidelity (Wi-Fi), Bluetooth (BT), mobile communication network, Frequency Modulation (FM), Near Field Communication (NFC), Infrared (IR) technology, etc.

[0126] The memory 230 can include one or more Random Access Memories (RAMs) and one or more Non-Volatile Memories (NVMs). The Random Access Memory can be directly read and written by the processor 240, and can be used to store executable programs (such as machine instructions) of operating systems or other running programs, and also can be used to store user and application data, etc.

[0127] The Random Access Memory can include Static Random-Access Memory (SRAM), Dynamic Random Access Memory (DRAM), Synchronous Dynamic Random Access Memory (SDRAM), Double Data Rate Synchronous Dynamic Random Access Memory (DDR SDRAM), etc.

[0128] The Non-Volatile Memory can also store executable programs and store user and application data, etc., which can be loaded in advance into the Random Access Memory for direct reading and writing by the processor 240. The Non-Volatile Memory can include disk storage devices, Flash Memory.

[0129] The memory 230 is configured to store one or more computer programs. The one or more computer programs are configured to be executed by the processor 240. The one or more computer programs include a plurality of instructions, which, when executed by the processor 240, can implement the data desensitization operation performed on the computer device 20.

[0130] In other embodiments, the computer device 20 further includes an external memory interface configured to connect an external memory, so as to extend the storage capability of the computer device 20.

[0131] The processor 240 can include one or more processing units, for example: the processor 240 can include an application processor (AP), a modem processor, a graphics processing unit (GPU), an image signal processor (ISP), a controller, a video codec, a digital signal processor (DSP), a baseband processor, and / or a neural-network processing unit (NPU), etc. Different processing units can be independent devices or integrated into one or more processors.

[0132] The processor 240 provides computing and control capabilities, for example, the processor 240 is configured to execute the computer program stored in the memory 230, so as to implement the processing of the data sent by the scanning device 10.

[0133] The I / O interface 250 is configured to provide a channel for user input or output, for example, the I / O interface 250 can be used to connect various input and output devices, such as a mouse, a keyboard, a touch device, a display screen, etc., so that the user can input information or make the information visualized.

[0134] The bus 260 is at least configured to provide a channel for communication between the communication module 220, the memory 230, the processor 240, and the I / O interface 250 in the computer device 20.

[0135] It can be understood that the structure illustrated in the embodiments of the present application does not constitute a specific limitation on the computer device 20. In other embodiments of the present application, the computer device 20 can include more or fewer components than the illustration, or combine certain components, or split certain components, or different arrangement of components. The illustrated components can be implemented in hardware, software, or a combination of software and hardware.

[0136] The embodiment of the present application further provides a computer readable storage medium, and the computer readable storage medium stores a computer program. The computer program includes program instructions. The method realized by executing the program instructions can refer to the method in each of the above embodiments of the present application.

[0137] The computer readable storage medium can be an internal memory of the electronic device, such as a hard disk or a memory of the electronic device. The computer readable storage medium can also be an external storage device of the electronic device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, and the like.

[0138] In some embodiments, the computer readable storage medium can include a program storage area and a data storage area. The program storage area can store an operating system, an application required by at least one function, and the like. The data storage area can store data created according to the use of the electronic device, and the like.

[0139] In the above embodiments, the description of each embodiment has its own focus. The part not described or recorded in detail in an embodiment can refer to the related description of other embodiments.

[0140] Those skilled in the art can understand that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be realized by electronic hardware or a combination of computer software and electronic hardware. Whether the functions are realized in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to realize the described functions for each specific application, but such implementation should not be considered beyond the scope of the present application.

[0141] In the embodiments provided by the present application, it should be understood that the disclosed apparatus / terminal device and method can be implemented by other ways. For example, the above-mentioned apparatus / terminal device embodiments are only schematic, and the division of the modules or units is only a logical function division, and there can be another division way in actual implementation. For example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections between different units, or the logical couplings or communication connections between different units, can be indirect couplings or communication connections through some interfaces, devices or units, and can be electrical, mechanical or other forms.

[0142] The units described as separate components may or may not be physically separate, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may also be distributed to multiple network units. Part or all of the units can be selected to achieve the purpose of the embodiment scheme according to actual needs.

[0143] The above embodiments are only used to illustrate the technical solutions of the present application, but not limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that the technical solutions recorded in the foregoing embodiments can still be modified, or some technical features can be replaced by equivalents; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application.

Claims

1. A data processing method, wherein, The method comprises: in response to a request for generating a data analysis report by a scanning device, obtaining source data generated by the scanning device, wherein the source data at least includes a point cloud generated when the scanning device scans an object; in response to generating one or more first analyses of the scan based on the point cloud, performing de-sensitization on one or more sensitive first analyses of the object revealed in the one or more first analyses, wherein the de-sensitization makes the one or more sensitive first analyses no longer completely reveal the object; the one or more sensitive first analyses include a category analysis of the object, and the de-sensitization on the category analysis of the object comprises: replacing the category analysis with a de-sensitized category associated with the category analysis; the one or more sensitive first analyses include a visual effect analysis of the object, and the de-sensitization on the visual effect analysis of the object comprises: replacing at least part of the point cloud representing one or more visual effects of the category analysis in the point cloud with one or more de-sensitized point clouds associated with the category analysis, wherein the one or more de-sensitized point clouds correspond to changes in the actual presentation of the one or more visual effects; obtaining the point cloud after replacing the at least part of the point cloud; analyzing global features of the replaced point cloud using a regression model to obtain a quality assessment result; and generating the data analysis report based on the quality assessment result.

2. The method of claim 1, wherein, The category analysis of the object is obtained by using a first neural network to recognize the point cloud, comprising: performing a structured conversion on the point cloud to obtain a converted point cloud; based on the converted point cloud, obtaining a plurality of local features through the first neural network; aggregating the plurality of local features to obtain a global feature; based on the global feature, the first neural network outputs a category analysis probability of the object; determining the category analysis of the object based on the category analysis probability.

3. The method of claim 2, wherein, The structured conversion of the point cloud to obtain the converted point cloud comprises: performing voxelization processing on the point cloud to obtain a voxelized point cloud; or performing multi-view projection processing on the point cloud to obtain a multi-view projected point cloud.

4. The method of claim 1, wherein, The category analysis of the object is obtained by using a first neural network to recognize the point cloud, comprising: constructing a graph structure corresponding to the point cloud; based on the graph structure, obtaining node features through the first neural network; performing feature aggregation on the node features to obtain a global feature; based on the global feature, the first neural network outputs a category analysis probability of the object; determining the category analysis of the object based on the category analysis probability.

5. The method of claim 1, wherein, Further comprising: determining the at least part of the point cloud representing one or more visual effects of the category analysis from the point cloud based on the category analysis; generating the one or more de-sensitized point clouds using a second neural network based on the at least part of the point cloud and the category analysis.

6. The method of claim 1, wherein, The source data further comprises an operation log generated when the scanning device scans the object, and the method further comprises: In response to generating operation-related one or more second analyses based on the operation log, desensitizing one or more sensitive second analyses in the one or more second analyses that reveal the operator related to the generated operation, wherein the desensitization causes the one or more sensitive second analyses to no longer reveal the operator.

7. The method of any one of claims 1 to 6, wherein, Applied to an associated computing device in communication connection with the scanning device, and displaying the one or more desensitized first analyses on the associated computing device.

8. The method of any one of claims 1 to 6, wherein, Applied to the scanning device, and displaying the one or more desensitized first analyses on the scanning device.

9. The method of claim 1, wherein, Further comprising: In response to the scanning device's data transmission operation to an associated computing device, obtaining source data generated by the scanning device; Or, In response to the scanning device's operation of requesting to generate a log analysis report, obtaining source data generated by the scanning device.

10. The method of claim 1, wherein, After replacing at least part of the point cloud representing one or more visual effects of the category analysis in the point cloud with one or more desensitized point clouds associated with the category analysis, further comprising: In response to the scanning device's request for data transmission operation to an associated computing device, obtaining the point cloud in which the at least part of the point cloud has been replaced; Sending the replaced point cloud to the associated computing device; Receiving the associated computing device's quality evaluation result of the replaced point cloud.

11. The method of claim 1, wherein, The desensitization causes the one or more sensitive first analyses to no longer completely reveal the object, comprising: The one or more sensitive first analyses after desensitization are configured to reveal one aspect of the object, wherein the one aspect includes one or more of the object's material, size, color, and weight.

12. A scanning device, wherein, The scanning device comprises a processor and a memory, the memory stores a computer program, and the processor implements the data processing method as claimed in any one of claims 1 to 11 when executing the computer program.

13. A computing device, wherein, The computing device comprises a processor and a memory, the memory stores a computer program, and the processor implements the data processing method as claimed in any one of claims 1 to 11 when executing the computer program.

14. A computer readable storage medium, wherein, The computer readable storage medium stores at least one instruction, which is executed by the processor to implement the data processing method as claimed in any one of claims 1 to 11.

Images