Model data processing method and system based on gap detection

Abstract

The invention discloses a model data processing method and system based on gap detection. The method comprises the steps of obtaining model slice data uploaded by a target user; determining a detection gap corresponding to a slice part in the model slice data based on a gap detection model; determining a plurality of suspected damage positions corresponding to the model slice data according to the detection gaps corresponding to the at least two slice parts; and determining at least one damage position and a corresponding damage unreasonable degree from the plurality of suspected damage positions based on a historical model repair record and a prediction algorithm of the target user. Therefore, the damage detection accuracy and detection efficiency in the model slice data can be improved, so that the integrity and quality of the model data of the user are improved, and better use experience is provided for the user.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a model data processing method and system based on gap detection. Background Art

[0002] With the development of FDM technology, more and more users or organizations have begun to pay attention to technical issues such as the management and transmission of printed model data. Among them, how to effectively improve the efficiency of users' analysis and detection of the integrity of model slice data is a technical issue that has attracted much attention. However, most of the existing slice data processing technologies still only determine whether the model has unreasonable damage based on user annotations, and do not use algorithm technology to automatically identify the damage location in the model slice data and analyze its rationality. Therefore, the efficiency and accuracy of model damage detection are low, and the user experience is poor. It can be seen that the existing technology has defects that need to be solved urgently. Summary of the invention

[0003] The technical problem to be solved by the present invention is to provide a model data processing method and system based on gap detection, which can improve the accuracy and detection efficiency of damage detection in model slicing data, so as to improve the integrity and quality of the user's model data, thereby giving the user a better user experience.

[0004] In order to solve the above technical problems, the first aspect of the present invention discloses a model data processing method based on gap detection, the method comprising: Get the model slice data uploaded by the target user; Based on the gap detection model, determining the detection gap corresponding to the slice part in the model slice data; Determining a plurality of suspected damage locations corresponding to the model slice data according to the detection gaps corresponding to at least two of the slice parts; Based on the historical model repair record and prediction algorithm of the target user, at least one damage location and a corresponding unreasonable degree of damage are determined from the multiple suspected damage locations.

[0005] As an optional implementation, in the first aspect of the present invention, the determining, based on the gap detection model, the detection gap corresponding to the slice part in the model slice data includes: Inputting each slice part in the model slice data into the trained gap detection model to obtain the gap existence probability corresponding to each slice part; Screening out the slice parts whose gap existence probability is greater than a probability threshold, and obtaining a plurality of gap slice parts; Each of the notch slice portions is input into a trained image edge notch detection model to obtain a detection notch corresponding to each of the notch slice portions.

[0006] As an alternative embodiment, in the first aspect of the present invention, the notch detection model is an RNN neural network, which is trained by a training data set including a plurality of training slice images and corresponding annotations of whether there are notches; the image edge notch detection model is an RNN neural network, which is trained by a training data set including a plurality of training slice images and corresponding edge notch annotations; the model parameters of the image edge notch detection model are greater than the model parameters of the notch detection model.

[0007] As an alternative embodiment, in the first aspect of the present invention, determining a plurality of suspected damaged positions corresponding to the model slice data according to the detected notches corresponding to at least two of the slice parts includes: For any plurality of adjacent slice parts with the detected notches, determining at least one notch sequence of the detected notches of the plurality of slice parts; the notch sequence includes a plurality of detected notches with close positions arranged from top to bottom according to the positions of the corresponding slice parts in the model slice data; Inputting each notch sequence into a trained notch continuous rationality prediction neural network to obtain a continuous rationality parameter corresponding to each notch sequence; Screening out the notch sequences with the continuous rationality parameters greater than the first parameter threshold, and determining them as damaged notch sequences; Determining the superimposed set of the positions of the detected notches in each damaged notch sequence as a suspected damaged position corresponding to the model slice data.

[0008] As an alternative embodiment, in the first aspect of the present invention, the notch continuous rationality prediction neural network is an LSTM neural network, which is trained by a training data set including a plurality of continuous notch sequences and corresponding annotations of whether they are damaged positions.

[0009] As an alternative embodiment, in the first aspect of the present invention, determining at least one notch sequence of the detected notches of the plurality of slice parts includes: Setting the objective function to include that the number of notches in each notch sequence is minimized; Setting the constraints to include: The distance between the geometric centers of the detected notches corresponding to two adjacent slice parts in each notch sequence is less than the first distance threshold; The distance between the geometric centers of any two detected notches belonging to different notch sequences is greater than the second distance threshold; the second distance threshold is greater than the first distance threshold; The average distance between the geometric center points corresponding to any two of the detected gaps in each gap sequence is less than a third distance threshold; the third distance threshold is greater than the first distance threshold and less than the second distance threshold; Based on the objective function and the constraint conditions, iteratively group and optimize the detected gaps of the multiple slice parts based on the dynamic programming algorithm to obtain an optimal grouping result; at least one gap sequence is included in the grouping result.

[0010] As an optional implementation manner, in the first aspect of the present invention, determining at least one damaged position and the corresponding unreasonable degree of damage from the multiple suspected damaged positions based on the historical model repair records and prediction algorithm of the target user includes: Determine multiple historical model repair positions of the target user according to the historical model repair records of the target user; For each of the suspected damaged positions, calculate the position similarity between the suspected damaged position and the multiple historical model repair positions; Input the position parameter of the suspected damaged position and the image of the corresponding detected gap into the trained damaged reasonableness prediction neural network to obtain the unreasonable degree of damage corresponding to the suspected damaged position; the damaged reasonableness prediction neural network is trained through a training data set including multiple training damaged positions and corresponding gap image sets and damaged unreasonableness annotations; Calculate the weighted sum average of the position similarity and the unreasonable degree of damage to obtain the position parameter corresponding to the suspected damaged position; Screen out the positions with the position parameter greater than the second parameter threshold from the multiple suspected damaged positions to obtain at least one damaged position corresponding to the model slice data.

[0011] As an optional implementation manner, in the first aspect of the present invention, calculating the position similarity between the suspected damaged position and the multiple historical model repair positions includes: Calculate the first vector representation between the suspected damaged position and the model center point position of the model slice data; Calculate the second vector representation between each historical model repair position and the model center point position of its corresponding model data; Calculate the average value of the vector distances between the first vector representation and each second vector representation to obtain the position similarity between the suspected damaged position and the multiple historical model repair positions.

[0012] A second aspect of the embodiments of the present invention discloses a model data processing system based on gap detection, and the system includes: An acquisition module for acquiring model slice data uploaded by a target user; A first determination module for determining a detected notch corresponding to a sliced portion in the model slice data based on a notch detection model; A second determination module for determining a plurality of suspected damaged positions corresponding to the model slice data according to the detected notches corresponding to at least two of the sliced portions; A third determination module for determining at least one damaged position and a corresponding unreasonable degree of damage from the plurality of suspected damaged positions based on the historical model repair records and prediction algorithms of the target user.

[0013] As an optional implementation manner, in the second aspect of the present invention, the specific manner in which the first determination module determines the detected notch corresponding to the sliced portion in the model slice data based on the notch detection model includes: Inputting each sliced portion in the model slice data into a trained notch detection model to obtain a notch existence probability corresponding to each sliced portion; Screening out the sliced portions with the notch existence probability greater than a probability threshold to obtain a plurality of notched sliced portions; Inputting each notched sliced portion into a trained image edge notch detection model to obtain a detected notch corresponding to each notched sliced portion.

[0014] As an optional implementation manner, in the second aspect of the present invention, the notch detection model is an RNN neural network, which is trained by a training data set including a plurality of training sliced images and corresponding labels indicating whether there is a notch; the image edge notch detection model is an RNN neural network, which is trained by a training data set including a plurality of training sliced images and corresponding edge notch labels; the model parameters of the image edge notch detection model are greater than the model parameters of the notch detection model.

[0015] As an optional implementation manner, in the second aspect of the present invention, the specific manner in which the second determination module determines a plurality of suspected damaged positions corresponding to the model slice data according to the detected notches corresponding to at least two of the sliced portions includes: For any plurality of adjacent sliced portions with the detected notches, determining at least one notch sequence of the detected notches of the plurality of sliced portions; the notch sequence includes a plurality of detected notches with close positions arranged from top to bottom according to the positions of the corresponding sliced portions in the model slice data; Inputting each notch sequence into a trained notch continuous rationality prediction neural network to obtain a continuous rationality parameter corresponding to each notch sequence; Screen out the notch sequences with the continuous rationality parameter greater than the first parameter threshold, and determine them as damaged notch sequences; Determine the superimposed set of the positions of the detected notches in each of the damaged notch sequences as a suspected damaged position corresponding to the model slice data.

[0016] As an optional implementation manner, in the second aspect of the present invention, the notch continuous rationality prediction neural network is an LSTM neural network, which is trained by a training data set including a plurality of continuous notch sequences and corresponding labels of whether they are damaged positions.

[0017] As an optional implementation manner, in the second aspect of the present invention, the specific manner in which the second determination module determines at least one notch sequence of the detected notches of the plurality of slice parts includes: Set the objective function to include the minimum number of notches in each notch sequence; Set the constraint conditions to include: The distance between the geometric centers corresponding to the detected notches of two adjacent slice parts in each notch sequence is less than the first distance threshold; The distance between the geometric centers corresponding to any two detected notches belonging to different notch sequences is greater than the second distance threshold; the second distance threshold is greater than the first distance threshold; The average value of the distances between the geometric centers corresponding to any two detected notches in each notch sequence is less than the third distance threshold; the third distance threshold is greater than the first distance threshold and less than the second distance threshold; Based on the objective function and the constraint conditions, perform iterative grouping optimization on the detected notches of the plurality of slice parts based on the dynamic programming algorithm to obtain an optimal grouping result; the grouping result includes at least one notch sequence.

[0018] As an optional implementation manner, in the second aspect of the present invention, the specific manner in which the third determination module determines at least one damaged position and the corresponding unreasonable degree of damage from the plurality of suspected damaged positions based on the historical model repair records and prediction algorithms of the target user includes: Determine a plurality of historical model repair positions of the target user according to the historical model repair records of the target user; For each of the suspected damaged positions, calculate the position similarity between the suspected damaged position and the plurality of historical model repair positions; Input the position parameter of the suspected damaged position and the corresponding image of the detection notch into the trained damaged rational prediction neural network to obtain the degree of unreasonableness of damage corresponding to the suspected damaged position; the damaged rational prediction neural network is trained by a training data set including a plurality of training damaged positions, corresponding notch image sets, and damaged unreasonableness annotations; Calculate the weighted sum average of the position similarity and the degree of unreasonableness of damage to obtain the position parameter corresponding to the suspected damaged position; Select the positions with the position parameter greater than the second parameter threshold from the plurality of suspected damaged positions to obtain at least one damaged position corresponding to the model slice data.

[0019] As an optional implementation manner, in the second aspect of the present invention, the specific manner in which the third determination module calculates the position similarity between the suspected damaged position and the plurality of historical model repair positions includes: Calculate the first vector representation between the suspected damaged position and the model center point position of the model slice data; Calculate the second vector representation between each historical model repair position and the model center point position of its corresponding model data; Calculate the average value of the vector distances between the first vector representation and each second vector representation to obtain the position similarity between the suspected damaged position and the plurality of historical model repair positions.

[0020] The third aspect of the present invention discloses another model data processing system based on notch detection, and the system includes: A memory storing executable program code; A processor coupled to the memory; The processor calls the executable program code stored in the memory and executes some or all of the steps in the model data processing method based on notch detection disclosed in the first aspect of the present invention.

[0021] The fourth aspect of the present invention discloses a computer storage medium, and the computer storage medium stores computer instructions, which are used to execute some or all of the steps in the model data processing method based on notch detection disclosed in the first aspect of the present invention when called.

[0022] Compared with the prior art, the embodiments of the present invention have the following beneficial effects: The present invention can determine the detected notch corresponding to the sliced part in the model slice data based on the notch detection model, and then determine multiple suspected damaged positions corresponding to the model slice data based on the detected notch, so as to comprehensively and accurately determine the damaged position and the corresponding unreasonable degree of damage, thereby improving the accuracy and detection efficiency of damage detection in the model slice data, improving the integrity and quality of the user's model data, and further giving the user a better use experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the accompanying drawings required for the description of the embodiments. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can be obtained based on these drawings.

[0024] Figure 1 is a schematic flowchart of a model data processing method based on notch detection disclosed in an embodiment of the present invention.

[0025] Figure 2 is a schematic structural diagram of a model data processing system based on notch detection disclosed in an embodiment of the present invention.

[0026] Figure 3 is a schematic structural diagram of another model data processing system based on notch detection disclosed in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0027] In order to enable those skilled in the art to better understand the solutions of the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, rather than all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0028] The terms "first", "second", etc. in the specification and claims of the present invention and the above accompanying drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, device, product or equipment that includes a series of steps or units is not limited to the listed steps or units, but optionally further includes steps or units not listed, or optionally further includes other steps or units inherent to these processes, methods, products or equipment.

[0029] References to "embodiments" in this specification mean that specific features, structures, or characteristics described in connection with the embodiments can be included in at least one embodiment of the invention. The phrase appears in various places in the specification and does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments. Those skilled in the art will explicitly and implicitly understand that the embodiments described herein can be combined with other embodiments.

[0030] The present invention discloses a method and system for processing model data based on notch detection, which can determine the detection notches corresponding to the sliced parts in the model slice data based on a notch detection model, and then determine multiple suspected damaged positions corresponding to the model slice data based on the detection notches, so as to comprehensively and accurately determine the damaged positions and the corresponding unreasonable degrees of damage, thereby improving the accuracy and detection efficiency of damage detection in model slice data, improving the integrity and quality of the user's model data, and thus giving the user a better usage experience. The following will be described in detail respectively.

[0031] Embodiment 1 Please refer to Figure 1 , Figure 1 which is a schematic flowchart of a method for processing model data based on notch detection disclosed in an embodiment of the present invention. Among them, Figure 1 the described method for processing model data based on notch detection can be applied to a data processing system / data processing device / data processing server (wherein, the server includes a local processing server or a cloud processing server). As Figure 1 shown, the method for processing model data based on notch detection can include the following operations: 101. Obtain the model slice data uploaded by the target user.

[0032] 102. Based on the notch detection model, determine the detection notches corresponding to the sliced parts in the model slice data. 103. According to the detection notches corresponding to at least two sliced parts, determine multiple suspected damaged positions corresponding to the model slice data. 104. Based on the historical model repair records and prediction algorithms of the target user, determine at least one damaged position and the corresponding unreasonable degree of damage from the multiple suspected damaged positions.

[0033] It can be seen that the above-mentioned invention embodiments can determine the detection notches corresponding to the sliced parts in the model slice data based on the notch detection model, and then determine multiple suspected damaged positions corresponding to the model slice data based on the detection notches, so as to comprehensively and accurately determine the damaged positions and the corresponding unreasonable degrees of damage, thereby improving the accuracy and detection efficiency of damage detection in model slice data, improving the integrity and quality of the user's model data, and thus giving the user a better usage experience.

[0034] As an optional embodiment, in the above steps, determining the detected notch corresponding to the sliced part in the model slice data based on the notch detection model includes: Inputting each sliced part in the model slice data into the trained notch detection model to obtain the notch existence probability corresponding to each sliced part; Screening out the sliced parts with a notch existence probability greater than the probability threshold to obtain multiple notched sliced parts; Inputting each notched sliced part into the trained image edge notch detection model to obtain the detected notch corresponding to each notched sliced part.

[0035] It can be seen that through the above optional embodiment, it is possible to first screen out the sliced parts with a high notch existence probability based on the notch detection model, and then identify the edge notches based on the image edge notch detection model, so as to facilitate subsequent accurate identification of damage and analysis of the unreasonable degree of model damage, assisting in improving the accuracy and detection efficiency of damage detection in model slice data, improving the integrity and quality of the user's model data, and thus giving the user a better use experience.

[0036] As an optional embodiment, in the above steps, the notch detection model is an RNN neural network, which is trained through a training data set including multiple training slice images and corresponding labels indicating whether there is a notch; the image edge notch detection model is an RNN neural network, which is trained through a training data set including multiple training slice images and corresponding edge notch labels; the model parameters of the image edge notch detection model are greater than those of the notch detection model.

[0037] It can be seen that through the above optional embodiment, the details of the two detection models are defined, enabling the use of a small model to screen out the notched sliced parts first, and then using a large model to detect the edge notches, so as to improve the detection efficiency, facilitate subsequent accurate identification of damage and analysis of the unreasonable degree of model damage, assist in improving the accuracy and detection efficiency of damage detection in model slice data, improve the integrity and quality of the user's model data, and thus give the user a better use experience.

[0038] As an optional embodiment, in the above steps, determining multiple suspected damage positions corresponding to the model slice data according to the detected notches corresponding to at least two sliced parts includes: For any multiple adjacent sliced parts with detected notches, determining at least one notch sequence of the detected notches of the multiple sliced parts; optionally, the notch sequence includes multiple detected notches with close positions arranged from top to bottom according to the positions of the corresponding sliced parts in the model slice data; Input each gap sequence into the trained neural network for predicting the continuity rationality of gaps to obtain the continuity rationality parameter corresponding to each gap sequence; Screen out the gap sequences with continuity rationality parameters greater than the first parameter threshold and determine them as damaged gap sequences; Determine the superimposed set of the positions of the detected gaps in each damaged gap sequence as a suspected damaged position corresponding to the model slice data.

[0039] It can be seen that through the above optional embodiments, it is possible to judge the continuity rationality of the gaps in adjacent slice parts based on the trained neural network for predicting the continuity rationality of gaps to identify the positions where continuous missing occurs to form damage, so as to facilitate subsequent accurate identification of damage and analysis of the unreasonable degree of model damage, and assist in improving the accuracy and detection efficiency of damage detection in model slice data, so as to improve the integrity and quality of the user's model data, and further give the user a better usage experience.

[0040] As an optional embodiment, in the above steps, the neural network for predicting the continuity rationality of gaps is an LSTM neural network, which is trained by a training data set including a plurality of continuous gap sequences and the corresponding annotations of whether they are damaged positions.

[0041] It can be seen that through the above optional embodiments, the details of the neural network for predicting the continuity rationality of gaps are defined, enabling it to predict whether the continuous change rationality of the continuous gap sequence is a damaged change, so as to screen out accurate damaged positions, improve the detection efficiency, facilitate subsequent accurate identification of damage and analysis of the unreasonable degree of model damage, and assist in improving the accuracy and detection efficiency of damage detection in model slice data, so as to improve the integrity and quality of the user's model data, and further give the user a better usage experience.

[0042] As an optional embodiment, in the above steps, determining at least one gap sequence of the detected gaps in the plurality of slice parts includes: Set the objective function to include the minimum number of gaps in each gap sequence; Set the constraints to include: The distance between the geometric centers corresponding to the detected gaps belonging to two adjacent slice parts in each gap sequence is less than the first distance threshold; The distance between the geometric centers corresponding to any two detected gaps belonging to different gap sequences is greater than the second distance threshold; optionally, the second distance threshold is greater than the first distance threshold; The average value of the distances between the geometric centers corresponding to any two detected gaps in each gap sequence is less than the third distance threshold; optionally, the third distance threshold is greater than the first distance threshold and less than the second distance threshold; Based on the objective function and constraints, the detection gaps of the multiple slice parts are iteratively grouped and optimized based on the dynamic programming algorithm to obtain the optimal grouping result; the grouping result includes at least one gap sequence.

[0043] It can be seen that through the above optional embodiments, the grouping of the gap sequences can be specified in detail based on the preset objective function and constraints, so as to divide the gap sequences closer to the more likely same breakage position through the dynamic programming algorithm, which is convenient for subsequent accurate identification of breakage and analysis of the unreasonable degree of model breakage, assisting in improving the accuracy and detection efficiency of breakage detection in model slice data, so as to improve the integrity and quality of the user's model data, and further give the user a better use experience.

[0044] As an optional embodiment, in the above steps, based on the historical model repair records of the target user and the prediction algorithm, at least one breakage position and the corresponding unreasonable degree of breakage are determined from multiple suspected breakage positions, including: According to the historical model repair records of the target user, multiple historical model repair positions of the target user are determined; For each suspected breakage position, calculate the position similarity between the suspected breakage position and the multiple historical model repair positions; Input the position parameter of the suspected breakage position and the image of the corresponding detection gap into the trained breakage reasonableness prediction neural network to obtain the unreasonable degree of breakage corresponding to the suspected breakage position; optionally, the breakage reasonableness prediction neural network is trained through a training data set including multiple training breakage positions and the corresponding set of gap images and breakage unreasonableness annotations; Calculate the weighted sum average of the position similarity and the unreasonable degree of breakage to obtain the position parameter corresponding to the suspected breakage position; Select the positions with position parameters greater than the second parameter threshold from multiple suspected breakage positions to obtain at least one breakage position corresponding to the model slice data.

[0045] It can be seen that through the above optional embodiments, the degree of unreasonable breakage of each suspected breakage position can be comprehensively determined based on the position similarity between the user's historical model repair records and each suspected breakage position and the prediction reasonableness degree of the breakage reasonableness prediction neural network, so as to accurately identify breakage and analyze the unreasonable degree of model breakage, and achieve improvement of the accuracy and detection efficiency of breakage detection in model slice data, so as to improve the integrity and quality of the user's model data, and further give the user a better use experience.

[0046] As an optional embodiment, in the above steps, calculating the position similarity between the suspected breakage position and the multiple historical model repair positions includes: Calculate a first vector representation between the suspected damaged position and the model center point position of the model slice data; Calculate a second vector representation between each historical model repair position and the model center point position of its corresponding model data; Calculate the average value of the vector distances between the first vector representation and each second vector representation to obtain the position similarity between the suspected damaged position and multiple historical model repair positions.

[0047] It can be seen that through the above optional embodiments, it is possible to accurately determine the position similarity between the suspected damaged position and the multiple historical model repair positions based on the vector distance between the vector representation of the damaged position and the model center point and the vector representation of the historical repair positions, so as to accurately identify the damage and analyze the unreasonable degree of model damage, realize improving the accuracy and detection efficiency of damage detection in model slice data, improve the integrity and quality of the user's model data, and thus give the user a better use experience.

[0048] Embodiment 2 Please refer to Figure 2 , Figure 2 which is a schematic structural diagram of a model data processing system based on notch detection disclosed in an embodiment of the present invention. Among them, Figure 2 The described model data processing system based on notch detection can be applied to a data processing system / data processing device / data processing server (wherein, the server includes a local processing server or a cloud processing server). As Figure 2 shown, the model data processing system based on notch detection may include: An acquisition module 201, configured to acquire model slice data uploaded by a target user.

[0049] A first determination module 202, configured to determine a detection notch corresponding to a sliced portion in the model slice data based on a notch detection model. A second determination module 203, configured to determine multiple suspected damaged positions corresponding to the model slice data according to the detection notches corresponding to at least two sliced portions. A third determination module 204, configured to determine at least one damaged position and the corresponding unreasonable degree of damage from multiple suspected damaged positions based on the historical model repair records of the target user and a prediction algorithm.

[0050] It can be seen that the above-described invention embodiments can determine the detection notch corresponding to the sliced part in the model slice data based on the notch detection model, and then determine multiple suspected damaged positions corresponding to the model slice data based on the detection notch, so as to comprehensively and accurately determine the damaged position and the corresponding unreasonable degree of damage, thereby improving the accuracy and detection efficiency of damage detection in the model slice data, improving the integrity and quality of the user's model data, and further giving the user a better usage experience.

[0051] As an optional embodiment, the specific manner in which the first determination module determines the detection notch corresponding to the sliced part in the model slice data based on the notch detection model includes: Input each sliced part in the model slice data into the trained notch detection model to obtain the notch existence probability corresponding to each sliced part; Select the sliced parts with a notch existence probability greater than the probability threshold to obtain multiple notched sliced parts; Input each notched sliced part into the trained image edge notch detection model to obtain the detection notch corresponding to each notched sliced part.

[0052] It can be seen that through the above optional embodiment, it is possible to first screen out the sliced parts with a high notch existence probability based on the notch detection model, and then identify the edge notch based on the image edge notch detection model, so as to facilitate the subsequent accurate identification of damage and analysis of the unreasonable degree of model damage, assisting in improving the accuracy and detection efficiency of damage detection in the model slice data, improving the integrity and quality of the user's model data, and further giving the user a better usage experience.

[0053] As an optional embodiment, the notch detection model is an RNN neural network, which is trained through a training data set including multiple training slice images and corresponding annotations of whether there is a notch; the image edge notch detection model is an RNN neural network, which is trained through a training data set including multiple training slice images and corresponding edge notch annotations; the model parameters of the image edge notch detection model are greater than the model parameters of the notch detection model.

[0054] It can be seen that through the above optional embodiment, the details of the two detection models are defined, enabling the use of a small model to screen the notched sliced parts first, and then using a large model to detect the edge notch, so as to improve the detection efficiency, facilitate the subsequent accurate identification of damage and analysis of the unreasonable degree of model damage, assisting in improving the accuracy and detection efficiency of damage detection in the model slice data, improving the integrity and quality of the user's model data, and further giving the user a better usage experience.

[0055] As an optional embodiment, the specific manner in which the second determination module determines multiple suspected damage positions corresponding to the model slice data based on the detection notches corresponding to at least two slice portions includes: For any multiple adjacent slice portions with detection notches, determine at least one notch sequence of the detection notches of the multiple slice portions; optionally, the notch sequence includes multiple detection notches with close positions arranged from top to bottom according to the positions of the corresponding slice portions in the model slice data; Input each notch sequence into the trained notch continuous rationality prediction neural network to obtain the continuous rationality parameter corresponding to each notch sequence; Screen out the notch sequences with continuous rationality parameters greater than the first parameter threshold and determine them as damaged notch sequences; Determine the superimposed set of the positions of the detection notches in each damaged notch sequence as a suspected damage position corresponding to the model slice data.

[0056] It can be seen that through the above optional embodiment, it is possible to judge the continuous rationality of the notches in adjacent slice portions based on the trained notch continuous rationality neural network to identify the continuous missing positions to form damaged positions, so as to accurately identify the damage and analyze the unreasonable degree of model damage subsequently, assist in improving the accuracy and detection efficiency of damage detection in model slice data, improve the integrity and quality of the user's model data, and thus give the user a better use experience.

[0057] As an optional embodiment, the notch continuous rationality prediction neural network is an LSTM neural network, which is trained through a training data set including multiple continuous notch sequences and the corresponding annotations of whether they are damaged positions.

[0058] It can be seen that through the above optional embodiment, the details of the notch continuous rationality prediction neural network are defined, enabling it to predict whether the continuous change rationality of the continuous notch sequence is a damaged change, so as to screen out accurate damaged positions, improve the detection efficiency, accurately identify the damage and analyze the unreasonable degree of model damage subsequently, assist in improving the accuracy and detection efficiency of damage detection in model slice data, improve the integrity and quality of the user's model data, and thus give the user a better use experience.

[0059] As an optional embodiment, the specific manner in which the second determination module determines at least one notch sequence of the detection notches of the multiple slice portions includes: Set the objective function to include that the number of notches in each notch sequence is minimized; Set the constraint conditions to include: The distance between the geometric centers corresponding to the detected gaps of two adjacent slice parts belonging to each gap sequence is less than the first distance threshold; The distance between the geometric centers corresponding to any two detected gaps belonging to different gap sequences is greater than the second distance threshold; Optionally, the second distance threshold is greater than the first distance threshold; The average value of the distances between the geometric centers corresponding to any two detected gaps in each gap sequence is less than the third distance threshold; Optionally, the third distance threshold is greater than the first distance threshold and less than the second distance threshold; According to the objective function and the constraint conditions, based on the dynamic programming algorithm, the detected gaps of the multiple slice parts are iteratively grouped and optimized to obtain the optimal grouping result; The grouping result includes at least one gap sequence.

[0060] It can be seen that through the above optional embodiments, the grouping of the gap sequences can be detailedly defined based on the preset objective function and constraint conditions, so as to divide the gap sequences closer to the more likely same damaged position through the dynamic programming algorithm, which is convenient for subsequent accurate identification of damage and analysis of the unreasonable degree of model damage, assisting in improving the accuracy and detection efficiency of damage detection in model slice data, so as to improve the integrity and quality of the user's model data, and further give the user a better use experience.

[0061] As an optional embodiment, the specific manner in which the third determination module determines at least one damaged position and the corresponding unreasonable degree of damage from multiple suspected damaged positions based on the historical model repair records and prediction algorithm of the target user includes: According to the historical model repair records of the target user, determine multiple historical model repair positions of the target user; For each suspected damaged position, calculate the position similarity between the suspected damaged position and the multiple historical model repair positions; Input the position parameter of the suspected damaged position and the image of the corresponding detected gap into the trained damaged reasonableness prediction neural network to obtain the unreasonable degree of damage corresponding to the suspected damaged position; Optionally, the damaged reasonableness prediction neural network is trained through a training data set including multiple training damaged positions and corresponding gap image sets and damaged unreasonableness annotations; Calculate the weighted sum average of the position similarity and the unreasonable degree of damage to obtain the position parameter corresponding to the suspected damaged position; Select the positions with position parameters greater than the second parameter threshold from multiple suspected damaged positions to obtain at least one damaged position corresponding to the model slice data.

[0062] It can be seen that through the above optional embodiments, it is possible to comprehensively determine the degree to which each suspected damage position belongs to unreasonable damage based on the position similarity between the user's historical model repair records and each suspected damage position and the prediction rationality degree of the damage rational prediction neural network, so as to accurately identify the damage and analyze the unreasonable degree of model damage, realize improving the accuracy and detection efficiency of damage detection in model slice data, improve the integrity and quality of the user's model data, and thus give the user a better usage experience.

[0063] As an optional embodiment, the specific manner in which the third determination module calculates the position similarity between the suspected damage position and multiple historical model repair positions includes: Calculating a first vector representation between the suspected damage position and the model center point position of the model slice data; Calculating a second vector representation between each historical model repair position and the model center point position of its corresponding model data; Calculating the average value of the vector distances between the first vector representation and each second vector representation to obtain the position similarity between the suspected damage position and multiple historical model repair positions.

[0064] It can be seen that through the above optional embodiments, it is possible to accurately determine the position similarity between the suspected damage position and the multiple historical model repair positions based on the vector distances between the vector representations of the damage position and the model center point and the vector representations of the historical repair positions, so as to accurately identify the damage and analyze the unreasonable degree of model damage, realize improving the accuracy and detection efficiency of damage detection in model slice data, improve the integrity and quality of the user's model data, and thus give the user a better usage experience.

[0065] Embodiment III Please refer to Figure 3 , Figure 3 which is another model data processing system based on notch detection disclosed in the embodiments of the present invention. Figure 3 The described model data processing system based on notch detection is applied to a data processing system / data processing device / data processing server (wherein, the server includes a local processing server or a cloud processing server). As Figure 3 shown, the model data processing system based on notch detection may include: A memory 301 storing executable program code; A processor 302 coupled to the memory 301; Wherein, the processor 302 calls the executable program code stored in the memory 301 to execute the steps of the model data processing method described in Embodiment I.

[0066] Embodiment IV An embodiment of the present invention discloses a computer-readable storage medium that stores a computer program for electronic data exchange. Among them, the computer program enables a computer to execute the steps of the model data processing method based on notch detection described in Embodiment 1.

[0067] Embodiment 5 An embodiment of the present invention discloses a computer program product. The computer program product includes a non-transitory computer-readable storage medium storing a computer program, and the computer program is operable to enable a computer to execute the steps of the model data processing method based on notch detection described in Embodiment 1.

[0068] The above describes specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve the desired result. Additionally, the processes depicted in the drawings do not necessarily have to be performed in the specific order or continuous order shown to achieve the desired result. In certain embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0069] The systems, devices, modules, or units illustrated in the above embodiments can be specifically implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, the computer can be, for example, a personal computer, a laptop computer, a cellular phone, a camera phone, a smart phone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or any combination of these devices.

[0070] For the convenience of description, when describing the above devices, they are described separately as various units according to their functions. Of course, when implementing this specification, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0071] Those skilled in the art should understand that the embodiments of this specification can be provided as a method, a system, or a computer program product. Therefore, the embodiments of this specification can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the embodiments of this specification can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0072] This specification is described with reference to the flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the specification. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and combinations of flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to the processors of general purpose computers, special purpose computers, embedded processors, or other programmable data processing devices to produce a machine, such that the instructions executed by the processors of the computer or other programmable data processing devices produce means for implementing the functions specified in the Figure 1 one or more flows and / or blocks Figure 1 one or more blocks

[0073] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a particular manner, such that the instructions stored in the computer-readable memory produce a manufacture including instruction means for implementing the functions specified in the Figure 1 one or more flows and / or blocks Figure 1 one or more blocks

[0074] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operational steps are performed on the computer or other programmable device to produce a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in the Figure 1 one or more flows and / or blocks Figure 1 one or more blocks

[0075] In a typical configuration, a computing device includes one or more processors (CPUs), an input / output interface, a network interface, and memory.

[0076] The memory may include non-permanent memory in the form of computer-readable media, random access memory (RAM), and / or non-volatile memory such as read-only memory (ROM) or flash memory (flash RAM). The memory is an example of computer-readable media.

[0077] A computer-readable medium includes permanent and non-permanent, removable and non-removable media that can implement information storage by any method or technology. The information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette tapes, magnetic tape disk storage or other magnetic storage devices, or any other non-transitory medium that can be used to store information accessible by a computing device. As defined herein, a computer-readable medium does not include transitory computer-readable media, such as modulated data signals and carrier waves.

[0078] It should also be noted that the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, commodity or device comprising a series of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, commodity or device. Without further limitation, an element defined by the statement "comprising an..." does not exclude the presence of additional identical elements in the process, method, commodity or device comprising the element.

[0079] This specification can be described in the general context of computer-executable instructions executed by a computer, such as program modules. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. This specification can also be practiced in a distributed computing environment where tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.

[0080] Each embodiment in this specification is described in a progressive manner. For the same or similar parts among the embodiments, reference can be made to each other. Each embodiment focuses on the differences from other embodiments. In particular, for system embodiments, since they are basically similar to method embodiments, the description is relatively simple, and reference can be made to the corresponding parts of the method embodiments for the relevant content.

[0081] Finally, it should be noted that: What is disclosed by a model data processing method and system based on notch detection disclosed in the embodiments of the present invention is only the preferred embodiments of the present invention, which are only used to illustrate the technical solutions of the present invention, rather than to limit it; Although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: They can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements on some of the technical features; 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 invention.

Claims

1. A method for processing model data based on notch detection, characterized in that, The method includes: Obtaining model slice data uploaded by a target user; Based on a notch detection model, determining a detection notch corresponding to a sliced part in the model slice data; Determining multiple suspected damaged positions corresponding to the model slice data according to the detection notches corresponding to at least two of the sliced parts; Based on the historical model repair records and prediction algorithm of the target user, determining at least one damaged position and the corresponding unreasonable degree of damage from the multiple suspected damaged positions.

2. The model data processing method based on notch detection according to claim 1, wherein The determining a detection notch corresponding to a sliced part in the model slice data based on the notch detection model includes: Inputting each sliced part in the model slice data into a trained notch detection model to obtain a notch existence probability corresponding to each sliced part; Screening out the sliced parts with the notch existence probability greater than a probability threshold to obtain multiple notched sliced parts; Inputting each notched sliced part into a trained image edge notch detection model to obtain a detection notch corresponding to each notched sliced part.

3. The method for processing model data based on notch detection according to claim 2, wherein The notch detection model is an RNN neural network, which is trained by a training data set including multiple training slice images and corresponding annotations of whether there is a notch; the image edge notch detection model is an RNN neural network, which is trained by a training data set including multiple training slice images and corresponding edge notch annotations; the model parameters of the image edge notch detection model are greater than those of the notch detection model.

4. The method for processing model data based on notch detection according to claim 2, wherein, The determining multiple suspected damaged positions corresponding to the model slice data according to the detection notches corresponding to at least two of the sliced parts includes: For any multiple adjacent sliced parts with the detection notch, determining at least one notch sequence of the detection notches of the multiple sliced parts; the notch sequence includes multiple detection notches with close positions arranged from top to bottom according to the positions of the corresponding sliced parts in the model slice data; Inputting each notch sequence into a trained notch continuous rationality prediction neural network to obtain a continuous rationality parameter corresponding to each notch sequence; Screening out the notch sequences with the continuous rationality parameter greater than a first parameter threshold and determining them as damaged notch sequences; Determining the superimposed set of the positions of the detection notches in each damaged notch sequence as a suspected damaged position corresponding to the model slice data.

5. The method for processing model data based on notch detection according to claim 4, wherein The notch continuous rationality prediction neural network is an LSTM neural network, which is trained by a training data set including multiple continuous notch sequences and corresponding annotations of whether they are damaged positions.

6. The method for processing model data based on notch detection according to claim 4, wherein The determining at least one notch sequence of the detection notches of the multiple sliced parts includes: Setting an objective function to include that the number of notches in each notch sequence is minimized; Setting constraint conditions to include: The distance between the geometric centers of the detection notches corresponding to two adjacent sliced parts in each notch sequence is less than a first distance threshold; The distance between the geometric centers of any two of the detected gaps belonging to different gap sequences is greater than a second distance threshold; the second distance threshold is greater than the first distance threshold; The average value of the distances between the geometric centers of any two of the detected gaps in each gap sequence is less than a third distance threshold; the third distance threshold is greater than the first distance threshold and less than the second distance threshold; Based on the objective function and the constraints, iteratively group and optimize the detected gaps of the multiple slice parts based on the dynamic programming algorithm to obtain an optimal grouping result; the grouping result includes at least one gap sequence.

7. The method for processing model data based on notch detection according to claim 1, wherein Determining at least one damaged position and the corresponding unreasonable degree of damage from the multiple suspected damaged positions based on the historical model repair records and prediction algorithm of the target user includes: Determine multiple historical model repair positions of the target user according to the historical model repair records of the target user; For each of the suspected damaged positions, calculate the position similarity between the suspected damaged position and the multiple historical model repair positions; Input the position parameters of the suspected damaged position and the image of the corresponding detected gap into the trained damaged reasonableness prediction neural network to obtain the unreasonable degree of damage corresponding to the suspected damaged position; the damaged reasonableness prediction neural network is trained by a training data set including multiple training damaged positions and corresponding gap image sets and damaged unreasonableness annotations; Calculate the weighted sum average of the position similarity and the unreasonable degree of damage to obtain the position parameters corresponding to the suspected damaged position; Screen out the positions with the position parameters greater than the second parameter threshold from the multiple suspected damaged positions to obtain at least one damaged position corresponding to the model slice data.

8. The method for processing model data based on notch detection according to claim 7, wherein Calculating the position similarity between the suspected damaged position and the multiple historical model repair positions includes: Calculate the first vector representation between the suspected damaged position and the model center point position of the model slice data; Calculate the second vector representation between each historical model repair position and the model center point position of its corresponding model data; Calculate the average value of the vector distances between the first vector representation and each second vector representation to obtain the position similarity between the suspected damaged position and the multiple historical model repair positions.

9. A model data processing system based on notch detection, characterized in that, The system includes: An acquisition module for acquiring model slice data uploaded by a target user; A first determination module for determining the detected gaps corresponding to the slice parts in the model slice data based on a gap detection model; A second determination module for determining multiple suspected damaged positions corresponding to the model slice data according to the detected gaps corresponding to at least two of the slice parts; A third determination module for determining at least one damaged position and the corresponding unreasonable degree of damage from the multiple suspected damaged positions based on the historical model repair records and prediction algorithm of the target user.

10. A model data processing system based on notch detection, characterized in that, The system includes: A memory storing executable program code; A processor coupled to the memory; The processor calls the executable program code stored in the memory and executes the model data processing method based on notch detection according to any one of claims 1-8.

Images