Precision part classification method and related equipment

Abstract

The invention provides a precision part classification method and related equipment, and the method comprises the steps: obtaining a three-dimensional part model file in an STEP format, and extracting text data in the three-dimensional part model file; obtaining the word frequency of each word by counting the occurrence frequency of each word in the text data; calculating the inverse document frequency of each word in the text data; for each word, multiplying the word frequency by the inverse document frequency to obtain an importance score of each word in the three-dimensional part model file; all the importance scores are combined to form feature vectors of the three-dimensional part model file, and the feature vectors are used for representing text features of the three-dimensional part model file; inputting the feature vectors into a trained classification model for classification to obtain a classification result; and the classification efficiency and precision can be effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of precision parts processing, and in particular to a precision parts classification method and related equipment. Background Art

[0002] The standard terminal equipment protocol (STEP) format is an international standard and is widely used to store and transmit complex 3D model data. These model data not only contain product geometry information, but also descriptions such as topology, materials, and manufacturing requirements. As the complexity of 3D models increases, how to quickly and effectively classify and analyze 3D models based on these model data has become a key issue. For example, in smart manufacturing, factories may need to automatically identify different types of parts in order to determine the production process.

[0003] Traditional 3D model recognition and classification methods rely on human visual feature classification, usually by analyzing 2D or 3D part drawings to determine the type of part. This visual judgment-based approach is not only time-consuming and labor-intensive, but also susceptible to subjective bias, and as the model complexity increases, the classification accuracy decreases significantly. Most existing automated methods are based on shallow module learning algorithms and cannot fully utilize the potential structure and semantic information in complex data.

[0004] The development of deep learning technology has provided a powerful tool for processing complex data, especially showing extremely high performance when processing text, images, and 3D model data. However, applying deep learning technology to the classification of 3D models in STEP format still faces challenges, mainly because STEP files have large data volumes and complex structures, and traditional methods cannot effectively process this information, resulting in low classification accuracy and efficiency. Summary of the invention

[0005] The invention provides a precision parts classification method and related equipment, the purpose of which is to improve the classification efficiency and accuracy.

[0006] In order to achieve the above object, the present invention provides a precision parts classification method, comprising:

[0007] Step 1, obtaining a 3D part model file in STEP format, and extracting text data in the 3D part model file;

[0008] Step 2, by counting the frequency of each word in the text data, obtain the word frequency of each word;

[0009] Step 3, calculate the inverse document frequency of each word in the text data;

[0010] Step 4, for each word, multiply the word frequency by the inverse document frequency to obtain the importance score of each word in the 3D part model file;

[0011] Step 5, combining all importance scores to form a feature vector of the 3D part model file, where the feature vector is used to characterize the text features of the 3D part model file;

[0012] Step 6: Input the feature vector into the trained classification model for classification to obtain the classification result.

[0013] Specifically, text data includes name, description, and attribute information.

[0014] Specifically, the trained classification model includes:

[0015] An input module, a first fully-connected module, a second fully-connected module, a third fully-connected module, a fourth fully-connected module, a fifth fully-connected module, a sixth fully-connected module, and an output module;

[0016] The input module is used to receive the input feature vector;

[0017] The first fully connected module, the second fully connected module, the third fully connected module, the fourth fully connected module, the fifth fully connected module, and the sixth fully connected module are all used to classify the feature vector;

[0018] The output module is used to output the classification results.

[0019] Further, the first fully-connected module has the same structure as the second fully-connected module, the third fully-connected module, the fourth fully-connected module, the fifth fully-connected module, and the sixth fully-connected module;

[0020] The first fully connected module includes:

[0021] Sequentially connected fully connected layer, BN layer, dropout layer, and activation function layer;

[0022] The input end of the fully connected layer is the input end of the first fully connected module;

[0023] The output end of the activation function layer is the output end of the first fully connected module.

[0024] Further, the output module includes a Softmax activation function;

[0025] The Softmax activation function is used to transform the classified feature vector to obtain the classification result.

[0026] Specifically, the trained classification model includes:

[0027] Input module, recurrent neural network module, self-attention module, fully connected module, Dropout module, output module;

[0028] The input module is used for input feature vectors;

[0029] The recurrent neural network module is used to process the feature vector to obtain the hidden state sequence;

[0030] The self-attention module is used to extract key features from the hidden state sequence;

[0031] The fully connected module is used to classify the feature vectors according to the key features;

[0032] The Dropout module is used to prevent the model from overfitting;

[0033] The output module is used to output the classification results.

[0034] Furthermore, the recurrent neural network module is a bidirectional recurrent neural network.

[0035] The present invention also provides a precision parts classification device, comprising:

[0036] An acquisition module is used to acquire a 3D part model file in a STEP format and extract text data from the 3D part model file;

[0037] The statistical module is used to obtain the frequency of each word by counting the frequency of each word in the text data;

[0038] A first calculation module, used to calculate the inverse document frequency of each word in the text data;

[0039] The second calculation module is used to multiply the word frequency by the inverse document frequency for each word to obtain the importance score of each word in the three-dimensional part model file;

[0040] A combination module is used to combine all importance scores to form a feature vector of the three-dimensional part model file, and the feature vector is used to characterize the text features of the three-dimensional part model file;

[0041] The classification module is used to input the feature vector into the trained classification model for classification to obtain the classification result.

[0042] The present invention also provides a terminal device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the precision parts classification method is implemented when the processor executes the computer program.

[0043] The present invention also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the precision parts classification method is implemented.

[0044] The above scheme of the present invention has the following beneficial effects:

[0045] The present invention obtains a three-dimensional part model file in a STEP format and extracts text data in the three-dimensional part model file; obtains the word frequency of each word by counting the frequency of occurrence of each word in the text data; calculates the inverse document frequency of each word in the text data; multiplies the word frequency by the inverse document frequency for each word respectively to obtain the importance score of each word in the three-dimensional part model file; combines all the importance scores to form a feature vector of the three-dimensional part model file, and the feature vector is used to characterize the text features of the three-dimensional part model file; inputs the feature vector into a trained classification model for classification to obtain a classification result; compared with the prior art, the present invention obtains the importance score of each word in the three-dimensional part model file by counting the word frequency of each word in the text data and calculating the inverse document frequency of each word in the text data, so as to vectorize the text data; then uses the powerful feature extraction capability of the trained classification model to extract features and classify the vectorized text data, which can effectively improve the classification efficiency and accuracy.

[0046] Other beneficial effects of the present invention will be described in detail in the subsequent specific implementation section. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 It is a schematic diagram of a flow chart of an embodiment of the present invention;

[0048] Figure 2 A schematic diagram of a structure of a classification model after training in an embodiment of the present invention;

[0049] Figure 3 for Figure 2 Schematic diagram of the structure of the fully connected module in the trained classification model shown;

[0050] Figure 4 Another structural schematic diagram of a classification model after training in an embodiment of the present invention;

[0051] Figure 5 It is a structural schematic diagram of a precision parts classification device in an embodiment of the present invention;

[0052] Figure 6 Schematic diagram of the structure of a terminal device in an embodiment of the present invention. DETAILED DESCRIPTION

[0053] In order to make the technical problems, technical solutions and advantages to be solved by the present invention clearer, the following will be described in detail with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0054] In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance.

[0055] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installation", "connection" and "connection" should be understood in a broad sense, for example, it can be a locking connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0056] In addition, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0057] In view of the existing problems, the present invention provides a precision parts classification method and related equipment.

[0058] like Figure 1 As shown, an embodiment of the present invention provides a precision parts classification method, comprising:

[0059] Step 1, obtaining a 3D part model file in STEP format, and extracting text data in the 3D part model file;

[0060] Step 2, by counting the frequency of each word in the text data, obtain the word frequency of each word;

[0061] Step 3, calculate the inverse document frequency of each word in the text data;

[0062] Step 4, for each word, multiply the word frequency by the inverse document frequency to obtain the importance score of each word in the 3D part model file;

[0063] Step 5, combining all importance scores to form a feature vector of the 3D part model file, where the feature vector is used to characterize the text features of the 3D part model file;

[0064] Step 6, input the feature vector into the trained classification model for classification to obtain classification results, which include plate parts and non-plate parts.

[0065] Specifically, step 1 includes:

[0066] Get the 3D part model file in STEP format;

[0067] In the embodiment of the present invention, the 3D part model file in STEP format is an international standard format for describing the geometric information and topological structure of the 3D part model. The file in this format not only contains complex 3D geometric data, but also contains a lot of text data describing the features of the 3D part model, including the name of the part, description of the geometric shape, properties of the material, and functional requirements, etc.

[0068] Extract text data from 3D part model files;

[0069] The text data includes the name of the part, a description of its geometry, and text descriptions related to the material properties.

[0070] Specifically, by counting the frequency of each word in the text data, we can get the word frequency of each word. The calculation expression is as follows:

[0071]

[0072] Among them, w represents the wth word.

[0073] Specifically, calculate the inverse document frequency of each word in the text data:

[0074] Inverse Document Frequency (IDF) is used to measure the importance of a word in all documents. Its calculation expression is:

[0075]

[0076] Wherein, N represents the number of 3D part model files, and DF(w) is the number of documents containing the wth word, that is, the number of 3D part model files in which the word appears.

[0077] Specifically, for each word, the word frequency and the inverse document frequency are multiplied to obtain the importance score of each word in the 3D part model file;

[0078] The importance score of a word is calculated by multiplying the term frequency (TF) by the inverse document frequency (IDF). The calculation expression is as follows:

[0079] W(w)=TF(w)×IDF(w)

[0080] Among them, TF(w) represents the word frequency of the w-th word in a certain 3D part model file, and IDF(w) represents the inverse document frequency of the w-th word in all 3D part model files.

[0081] Specifically, by combining all the importance scores, the text description of each STEP format file is converted into a high-dimensional sparse matrix. The sparse matrix includes the feature vectors of multiple 3D part model files. These feature vectors can well represent the text features of each file. Compared with traditional geometric feature extraction, this method can capture more information in the text description, thereby providing richer feature input for the model.

[0082] Most preferably, Figure 2 As shown, a structure of the trained classification model includes:

[0083] An input module, a first fully-connected module, a second fully-connected module, a third fully-connected module, a fourth fully-connected module, a fifth fully-connected module, a sixth fully-connected module, and an output module;

[0084] The input module is used to receive the input feature vector;

[0085] The first fully connected module, the second fully connected module, the third fully connected module, the fourth fully connected module, the fifth fully connected module, and the sixth fully connected module are all used to classify the feature vector;

[0086] The output module is used to output the classification results.

[0087] Most preferably, Figure 3 As shown, the first fully connected module has the same structure as the second fully connected module, the third fully connected module, the fourth fully connected module, the fifth fully connected module, and the sixth fully connected module, and are all used to improve the expression ability of the model. The first fully connected module includes:

[0088] Sequentially connected fully connected layer, BN layer, dropout layer, and activation function layer;

[0089] The input end of the fully connected layer is the input end of the first fully connected module;

[0090] The output end of the activation function layer is the output end of the first fully connected module.

[0091] Most preferably, the output module includes a Softmax activation function;

[0092] The Softmax activation function is used to transform the classified feature vector to obtain the classification result.

[0093] In the embodiment of the present invention, the BN layer is a batch normalization layer (BN, BatchNormalization), which is used to accelerate the training convergence speed, control gradient explosion, prevent gradient disappearance, and prevent overfitting.

[0094] Most preferably, Figure 4 As shown, another structure of the trained classification model includes:

[0095] Input module, recurrent neural network module, self-attention module, fully connected module, Dropout module, output module;

[0096] The input module is used for input feature vectors;

[0097] The recurrent neural network module is used to process the feature vector to obtain the hidden state sequence;

[0098] The self-attention module is used to extract key features from the hidden state sequence;

[0099] The fully connected module is used to classify the feature vectors according to the key features;

[0100] The Dropout module is used to prevent the model from overfitting;

[0101] The output module is used to output the classification results.

[0102] Most preferably, the recurrent neural network module is a bidirectional recurrent neural network, which provides a more comprehensive understanding of sequence data and higher processing capabilities by comprehensively considering the previous and next information of the sequence.

[0103] The training process of the embodiment of the present invention is completed in the NVIDIA GPU environment. The classification model is trained using the cross entropy loss function and the RMSprop optimizer. The learning rate of the optimizer is 0.0001 and the weight decay is 1e -6 The model performance is evaluated by the Stratified K-Fold cross-validation method. The model is trained on the GPU at an accelerated speed. The optimal configuration is obtained by continuously adjusting the model's hyperparameters such as the number of layers, number of neurons, and learning rate to obtain the trained classification model.

[0104] The specific application process is:

[0105] A set of 3D part model files in STEP format to be classified is obtained, and text data of each 3D part model file to be classified in the 3D part model file set is extracted; then the word frequency of each word in all the text data is counted, and then the inverse document frequency of each word in the text data is calculated; for each word, the word frequency and the inverse document frequency are multiplied to obtain the importance score of each word in the 3D part model file; all importance scores are combined to form a feature vector of the 3D part model file; the feature vector is input into a trained classification model for classification to obtain a binary classification result or a multi-classification result, the binary classification result is a plate class and a non-plate class, and the multi-classification result is a plate class, a shaft class, a bracket class, and a box class.

[0106] The embodiment of the present invention obtains a three-dimensional part model file in a STEP format and extracts text data in the three-dimensional part model file; obtains the word frequency of each word by counting the frequency of occurrence of each word in the text data; calculates the inverse document frequency of each word in the text data; multiplies the word frequency by the inverse document frequency for each word respectively to obtain the importance score of each word in the three-dimensional part model file; combines all the importance scores to form a feature vector of the three-dimensional part model file, and the feature vector is used to characterize the text features of the three-dimensional part model file; inputs the feature vector into a trained classification model for classification to obtain a classification result; compared with the prior art, the embodiment of the present invention obtains the importance score of each word in the three-dimensional part model file by counting the word frequency of each word in the text data and calculating the inverse document frequency of each word in the text data, so as to vectorize the text data; then uses the powerful feature extraction capability of the trained classification model to extract features and classify the vectorized text data, which can effectively improve the classification efficiency and accuracy.

[0107] Corresponding to the precision parts classification method described in the above embodiment, Figure 2 As shown, the embodiment of the present invention further provides a precision parts classification device 100, and the precision parts classification device 100 includes:

[0108] The acquisition module 101 is used to acquire a three-dimensional part model file in a STEP format and extract text data from the three-dimensional part model file;

[0109] A statistical module 102 is used to obtain the word frequency of each word by counting the frequency of occurrence of each word in the text data;

[0110] A first calculation module 103, used to calculate the inverse document frequency of each word in the text data;

[0111] The second calculation module 104 is used to multiply the word frequency and the inverse document frequency for each word to obtain the importance score of each word in the three-dimensional part model file;

[0112] A combination module 105 is used to combine all importance scores to form a feature vector of the three-dimensional part model file, and the feature vector is used to characterize the text features of the three-dimensional part model file;

[0113] The classification module 106 is used to input the feature vector into the trained classification model for classification to obtain a classification result.

[0114] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present application. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0115] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0116] The embodiment of the present invention also provides a terminal device, such as Figure 6 As shown, the terminal device D10 of this embodiment includes: at least one processor D100 ( Figure 6 Only one processor is shown in the figure), a memory D101, and a computer program D102 stored in the memory D101 and executable on the at least one processor D100, wherein the processor D100 implements the above-mentioned precision parts classification method when executing the computer program D102.

[0117] The terminal device D10 may be a computing device such as a desktop computer, a notebook, a PDA, a server, a server cluster, a cloud server, etc. The terminal device may include, but is not limited to, a processor D100 and a memory D101. Those skilled in the art will appreciate that Figure 6This is only an example of the terminal device D10 and does not constitute a limitation on the terminal device D10. The terminal device D10 may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, it may also include input and output devices, network access devices, etc.

[0118] The processor D100 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0119] In some embodiments, the memory D101 may be an internal storage unit of the terminal device D10, such as a hard disk or memory of the terminal device D10. In other embodiments, the memory D101 may also be an external storage device of the terminal device D10, such as a plug-in hard disk, a smart memory card (SMC, SmartMedia Card), a secure digital (SD, Secure Digital) card, a flash card (Flash Card), etc. equipped on the terminal device D10. Further, the memory D101 may also include both an internal storage unit of the terminal device D10 and an external storage device. The memory D101 is used to store an operating system, an application program, a boot loader (BootLoader), data and other programs, such as the program code of the computer program, etc. The memory D101 may also be used to temporarily store data that has been output or is to be output.

[0120] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present application. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0121] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0122] An embodiment of the present invention further provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, the precision parts classification method is implemented.

[0123] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the process in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may at least include: any entity or device that can carry the computer program code to the construction device / terminal device, recording medium, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. For example, a USB flash drive, a mobile hard disk, a disk or an optical disk.

[0124] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A precision parts classification method, characterized in that: include: Step 1, obtaining a three-dimensional part model file in STEP format, and extracting text data in the three-dimensional part model file; Step 2, obtaining the word frequency of each word by counting the frequency of occurrence of each word in the text data; Step 3, calculating the inverse document frequency of each word in the text data; Step 4, for each word, multiplying the word frequency by the inverse document frequency to obtain an importance score of each word in the three-dimensional part model file; Step 5, combining all importance scores to form a feature vector of the 3D part model file, wherein the feature vector is used to characterize text features of the 3D part model file; Step 6: Input the feature vector into the trained classification model for classification to obtain a classification result.

2. The precision parts classification method according to claim 1, characterized in that: The text data includes name, description and attribute information.

3. The precision parts classification method according to claim 2, characterized in that: The trained classification model includes: An input module, a first fully-connected module, a second fully-connected module, a third fully-connected module, a fourth fully-connected module, a fifth fully-connected module, a sixth fully-connected module, and an output module; The input module is used to receive an input feature vector; The first fully connected module, the second fully connected module, the third fully connected module, the fourth fully connected module, the fifth fully connected module, and the sixth fully connected module are all used to classify feature vectors; The output module is used to output the classification results.

4. The precision parts classification method according to claim 3, characterized in that: The first fully-connected module has the same structure as the second fully-connected module, the third fully-connected module, the fourth fully-connected module, the fifth fully-connected module, and the sixth fully-connected module; The first fully connected module includes: Sequentially connected fully connected layer, BN layer, dropout layer, and activation function layer; The input end of the fully connected layer is the input end of the first fully connected module; The output end of the activation function layer is the output end of the first fully connected module.

5. The precision parts classification method according to claim 4, characterized in that: The output module includes a Softmax activation function; The Softmax activation function is used to transform the classified feature vector to obtain the classification result.

6. The precision parts classification method according to claim 2, characterized in that: The trained classification model includes: Input module, recurrent neural network module, self-attention module, fully connected module, Dropout module, output module; The input module is used to input feature vectors; The recurrent neural network module is used to process the feature vector to obtain a hidden state sequence; The self-attention module is used to extract key features from the hidden state sequence; The fully connected module is used to classify feature vectors according to key features; The Dropout module is used to prevent the model from overfitting; The output module is used to output the classification results.

7. The precision parts classification method according to claim 6, characterized in that: The recurrent neural network module is a bidirectional recurrent neural network.

8. A precision parts classification device, characterized in that: include: An acquisition module, used for acquiring a three-dimensional part model file in a STEP format and extracting text data from the three-dimensional part model file; A statistical module, used for obtaining the word frequency of each word by counting the frequency of occurrence of each word in the text data; A first calculation module, used for calculating the inverse document frequency of each word in the text data; A second calculation module is used for multiplying the word frequency by the inverse document frequency for each word to obtain an importance score of each word in the three-dimensional part model file; A combination module, used for combining all importance scores to form a feature vector of the three-dimensional part model file, wherein the feature vector is used for characterizing text features of the three-dimensional part model file; The classification module is used to input the feature vector into the trained classification model for classification to obtain a classification result.

9. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the precision parts classification method according to any one of claims 1 to 7 is implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the precision parts classification method according to any one of claims 1 to 7 is implemented.

Images