AI-assisted planing machine drawing programming system and programming method thereof

Abstract

The present invention application discloses an AI-assisted planing machine drawing programming system and a programming method thereof, and belongs to the field of control technology. The AI-assisted planing machine drawing programming method includes: extracting drawings from archives and collecting natural language descriptions of programming operations input by users, and then introducing image processing algorithms and natural language processing technologies based on artificial intelligence and deep learning in the back end to analyze the drawing information and natural language descriptions of programming operations, so as to learn and capture the interactive screening features between the user's coding operation semantics and the drawing object image semantics, so as to generate post-programming drawings that meet user needs. The present invention application can use artificial intelligence technology to assist in the programming process of planing machine drawing, thereby improving the intelligence level of planing machine drawing programming, while simplifying the programming process and manual operations, which can not only lower the operating threshold, but also effectively reduce human burden and errors.

Description

Technical Field

[0001] The present invention relates to the field of control technology, and in particular, to an AI-assisted planing machine drawing programming system and a programming method thereof. Background Art

[0002] In modern manufacturing, the rapid development of CNC technology has promoted the automation and intelligentization of mechanical processing. As an important processing equipment, slotting machines are widely used in the cutting and forming of metals, plastics and other materials. The traditional slotting machine control methods are mainly PLC mode and computer board mode. However, the slotting machine controlled by PLC can only perform simple slotting actions, and can only edit one slotting action at a time. The function is relatively single, and it is difficult to realize the simultaneous processing of multiple plates, multiple directions or multiple slots, and it cannot meet the complex logic operation requirements. At the same time, the PLC control method requires certain PLC programming skills, which increases the threshold for use. Although the computer board control method has been improved on the basis of PLC, such as introducing a UI interface to reduce the difficulty of programming, and can realize multiple single-direction slotting operations, its programming methods are still divided into standard mode, semi-automatic and simple programming. In addition, users still need to choose a suitable programming method according to the specific situation, and the operation is not convenient enough.

[0003] In response to the above technical problems, Chinese patent CN112732244A discloses a method for drawing programming and table programming of a slotting machine based on a new generation controller, which allows users to directly perform drawing programming or table programming through a graphical interface, allowing users to freely set the number of processing files, copy, delete, and add, and can be exported to slotting machines of other systems for use. The advantage of this method is that the position, depth, bending direction and angle can be set for each groove, and the state after bending can be intuitively displayed through simulation graphics, which is convenient for rapid verification of the program.

[0004] In the above programming method, although it provides a more intuitive programming experience, it still requires manual drawing programming or table programming, and has not completely gotten rid of the need for manual programming intervention. The level of intelligence is low, especially when dealing with complex planer graphics or scenes that require a lot of repetitive work. It still requires more manual operations, which not only increases the labor burden, but also is prone to human errors. The accuracy and reliability of planer drawing programming are also difficult to guarantee. Therefore, an AI-assisted planer drawing programming system and programming method thereof are desired. Summary of the invention

[0005] This summary is provided to introduce concepts in a brief form that will be described in detail in the detailed description below. This summary is not intended to identify key features or essential features of the claimed technical solution, nor is it intended to limit the scope of the claimed technical solution.

[0006] In the first aspect, the present invention provides an AI-assisted planing machine drawing programming system, comprising:

[0007] Archives management module, used to enter the archives management interface;

[0008] File loading module, used to load the selected file;

[0009] A programming requirement input module, used to receive a natural language description of a programming operation input by a user;

[0010] An AI-assisted programming module, for programming the drawing in the file based on the natural language description of the programming operation to obtain a programmed drawing;

[0011] Wherein, the AI-assisted programming module includes:

[0012] A drawing extraction unit, used for extracting drawings from the archive;

[0013] A drawing and programming operation description semantic coding unit, used for semantically coding the drawing and the programming operation natural language description to obtain drawing object image semantic coding features and programming operation semantic coding features;

[0014] A feature interactive screening unit, used for performing feature interactive screening on the drawing object image semantic coding features and the programming operation semantic coding features based on metadata blocks to obtain programming semantic screening drawing object coding features;

[0015] The post-programming drawing generation module is used to screen the drawing object coding features based on the programming semantics to generate the post-programming drawing.

[0016] Optionally, the drawing and programming operation description semantic coding unit includes: a feature extraction subunit, used to perform feature extraction based on the drawing object on the drawing to obtain initial semantic coding features of the drawing object image; a drawing object image semantic feature enhancement subunit, used to perform drawing object image semantic feature enhancement based on grid energy attention on the initial semantic coding features of the drawing object image to obtain the drawing object image semantic coding features; a semantic coding subunit, used to perform semantic coding on the natural language description of the programming operation to obtain a programming operation semantic coding feature vector as the programming operation semantic coding feature.

[0017] Optionally, the feature extraction subunit is used to: input the drawing into a drawing object feature extractor based on a hole convolutional neural network model to obtain a drawing object image semantic coding feature map as an initial semantic coding feature of the drawing object image.

[0018] Optionally, the drawing object image semantic feature enhancement subunit includes: a grid division secondary subunit, used to grid the drawing object image semantic coding feature map to obtain a set of drawing object image semantic local feature maps; a significant feature enhancement secondary subunit, used to perform significant feature enhancement based on an energy significance description factor on the set of drawing object image semantic local feature maps to obtain a set of enhanced drawing object image semantic local feature maps; a feature aggregation processing secondary subunit, used to perform feature aggregation processing on the set of enhanced drawing object image semantic local feature maps to obtain a drawing object image semantic enhancement coding feature map as the drawing object image semantic coding feature.

[0019] Optionally, the significant feature enhancement secondary sub-unit is used to: calculate the energy significance description factor of each drawing object image semantic local feature map in the set of drawing object image semantic local feature maps to obtain a set of drawing object image semantic local energy significance description factors; input the set of drawing object image semantic local energy significance description factors into a local feature adaptive selector based on a gating function to obtain a set of drawing object image semantic local significance modulation weights; use each drawing object image semantic local significance modulation weight in the set of drawing object image semantic local significance modulation weights as a weight, and weight each drawing object image semantic local feature map in the set of drawing object image semantic local feature maps to obtain the set of enhanced drawing object image semantic local feature maps.

[0020] Optionally, the feature aggregation processing secondary sub-unit is used to: perform feature aggregation on the set of enhanced drawing object image semantic local feature maps in accordance with the grid division method to obtain a drawing object image semantic local significant guided enhanced feature map; input the drawing object image semantic local significant guided enhanced feature map into a feature dispersion module based on a hole convolution layer to obtain the drawing object image semantic enhanced encoding feature map.

[0021] Optionally, the feature interaction screening unit includes: a linear transformation and mask processing sub-unit, used to perform linear transformation and mask processing on the drawing object image semantic coding features and the programming operation semantic coding features respectively to obtain a masked first programming operation semantic coding auxiliary feature mapping pattern representation and a masked second programming operation semantic coding auxiliary feature mapping pattern representation; a feature screening sub-unit, used to perform feature screening on the drawing object image semantic coding features based on the masked first programming operation semantic coding auxiliary feature mapping pattern representation and the masked second programming operation semantic coding auxiliary feature mapping pattern representation to obtain the programming semantically screened drawing object coding features.

[0022] Optionally, the linear transformation and mask processing subunit is used to: perform a linear transformation on the programming operation semantic coding feature vector based on the first weight modifier matrix and the second weight modifier matrix to obtain a first programming operation semantic coding auxiliary feature mapping pattern representation vector and a second programming operation semantic coding auxiliary feature mapping pattern representation vector; based on the gradient amplitude of each position in the first programming operation semantic coding auxiliary feature mapping pattern representation vector, perform a mask processing based on the gradient amplitude on the first programming operation semantic coding auxiliary feature mapping pattern representation vector to obtain a masked first programming operation semantic coding auxiliary feature mapping pattern representation vector as the masked first programming operation semantic coding auxiliary feature mapping pattern representation; based on the gradient amplitude of each position in the second programming operation semantic coding auxiliary feature mapping pattern representation vector, perform a mask processing based on the gradient amplitude on the second programming operation semantic coding auxiliary feature mapping pattern representation vector to obtain the masked second programming operation semantic coding auxiliary feature mapping pattern representation vector as the masked second programming operation semantic coding auxiliary feature mapping pattern representation.

[0023] Optionally, the feature screening subunit is used to: perform preliminary feature screening on the drawing object image semantic enhancement coding feature map based on the masked first programming operation semantic coding auxiliary feature mapping pattern representation vector and hyperbolic tangent function to obtain the drawing object image semantic feature map after preliminary screening; and perform feature screening again on the drawing object image semantic feature map after preliminary screening based on the masked second programming operation semantic coding auxiliary feature mapping pattern representation vector and Sigmoid function to obtain the programming semantic screening drawing object coding feature map as the programming semantic screening drawing object coding feature.

[0024] Optionally, the post-programming drawing generation module is used to: input the programming semantics screening drawing object encoding feature map into a decoder-based drawing code generation module to obtain the post-programming drawing.

[0025] In a second aspect, a planer drawing programming method based on AI assistance adopts a planer drawing programming system based on AI assistance as described in any one of the first aspects above, comprising the following steps:

[0026] Step S201, entering the file management interface;

[0027] Step S202, loading the selected file;

[0028] Step S203, receiving a natural language description of a programming operation input by a user;

[0029] Step S204, programming the drawing in the file based on the natural language description of the programming operation to obtain a programmed drawing;

[0030] Wherein, step S204, programming the drawing in the file based on the natural language description of the programming operation to obtain a programmed drawing, includes:

[0031] Step S2041, extracting the drawing in the file;

[0032] Step S2042, semantically encode the natural language description of the drawing and the programming operation to obtain a drawing object image semantic encoding feature and a programming operation semantic encoding feature;

[0033] Step S2043, performing feature interactive screening on the drawing object image semantic coding features and the programming operation semantic coding features based on metadata blocks to obtain programming semantic screening drawing object coding features;

[0034] Step S2044: Screening drawing object coding features based on the programming semantics to generate the post-programming drawing.

[0035] The beneficial effects of the present invention are as follows:

[0036] The above technical solution is adopted to extract the drawings in the archives and collect the natural language description of the programming operation input by the user, and then introduce the image processing algorithm and natural language processing technology based on artificial intelligence and deep learning in the back end to analyze the drawing information and the natural language description of the programming operation, so as to learn and capture the interactive screening features between the user's coding operation semantics and the drawing object image semantics, so as to generate a post-programming drawing that meets the user's needs. In this way, artificial intelligence technology can be used to assist in the programming process of the planing machine drawing, thereby improving the intelligence level of the planing machine drawing programming, while simplifying the programming process and manual operation, which can not only lower the operating threshold, but also effectively reduce the human burden and errors. Other features and advantages of the present invention application will be described in detail in the subsequent specific implementation method section. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The above and other features, advantages and aspects of the various embodiments of the present invention will become more apparent with reference to the following detailed description in conjunction with the accompanying drawings. Throughout the drawings, the same or similar reference numerals represent the same or similar elements. It should be understood that the drawings are schematic and the originals and elements are not necessarily drawn to scale. In the drawings:

[0038] Figure 1 It is a block diagram of an AI-assisted routing machine drawing and programming system according to an exemplary embodiment.

[0039] Figure 2 It is a flow chart of a routing machine drawing programming method based on AI assistance according to an exemplary embodiment.

[0040] Figure numerals: file management module-101, file loading module-102, programming requirement input module-103, AI-assisted programming module-104, post-programming drawing generation module-105. DETAILED DESCRIPTION

[0041] The embodiments of the present invention will be described in more detail below with reference to the accompanying drawings. Although certain embodiments of the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as being limited to the embodiments described herein. Instead, these embodiments are provided to provide a more thorough and complete understanding of the present invention. It should be understood that the drawings and embodiments of the present invention are only for exemplary purposes and are not intended to limit the scope of protection of the present invention.

[0042] It should be understood that the various steps described in the method implementation of the present invention can be performed in different orders and / or in parallel. In addition, the method implementation may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect.

[0043] The term "including" and its variations used herein are open inclusions, i.e., "including but not limited to". The term "based on" means "based at least in part on". The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments". The relevant definitions of other terms will be given in the following description.

[0044] It should be noted that the concepts such as "first" and "second" mentioned in the present application are only used to distinguish different devices, modules or units, and are not used to limit the order or interdependence of the functions performed by these devices, modules or units.

[0045] It should be noted that the modifications of "one" and "plurality" mentioned in the present application are illustrative rather than restrictive, and those skilled in the art should understand that unless otherwise clearly indicated in the context, it should be understood as "one or more".

[0046] The names of the messages or information exchanged between multiple devices in the implementation manner of the present invention are only used for illustrative purposes, and are not used to limit the scope of these messages or information.

[0047] The specific implementation methods of the present invention are described in detail below with reference to the accompanying drawings.

[0048] With the development of artificial intelligence technology, especially the progress in computer vision and natural language processing, it is possible to use AI technology to assist or replace manual programming of slotting machines. By combining image recognition technology and natural language understanding capabilities, the automatic conversion from graphics to processing instructions can be achieved. This method can not only improve programming efficiency and reduce errors, but also simplify the programming process of complex graphics, so that operators without professional programming knowledge can easily complete programming work.

[0049] First, Figure 1 is a block diagram of an AI-assisted slotting machine drawing programming system according to an exemplary embodiment. Figure 1 As shown, an AI-assisted planing machine drawing programming system 100 includes:

[0050] The file management module 101 is used to enter the file management interface;

[0051] The file loading module 102 is used to load the selected file;

[0052] The programming requirement input module 103 is used to receive a natural language description of a programming operation input by a user;

[0053] The AI-assisted programming module 104 is used to program the drawing in the file based on the natural language description of the programming operation to obtain a programmed drawing;

[0054] The AI-assisted programming module 104 includes:

[0055] A drawing extraction unit 1041, used to extract drawings from the archive;

[0056] A drawing and programming operation description semantic coding unit 1042, used for semantically coding the drawing and the programming operation natural language description to obtain drawing object image semantic coding features and programming operation semantic coding features;

[0057] A feature interactive screening unit 1043 is used to perform feature interactive screening on the drawing object image semantic coding feature and the programming operation semantic coding feature based on metadata blocks to obtain programming semantic screening drawing object coding features;

[0058] The post-programming drawing generation module 105 is used to filter the coding features of the drawing objects based on the programming semantics to generate the post-programming drawing.

[0059] In particular, in the above-mentioned AI-assisted slotting machine drawing programming system, in order to meet the needs of users to perform programming and generate programming drawings that meet the needs of users, it is necessary to accurately parse the natural language description of the programming operation input by the user, and read the drawing information in the archive selected by the user, so as to perform corresponding programming operations on the drawing to generate programming drawings that meet the user's requirements. Specifically, the technical concept of the present invention application is to extract the drawings in the archive, collect the natural language description of the programming operation input by the user, and then introduce the image processing algorithm and natural language processing technology based on artificial intelligence and deep learning in the back end to analyze the drawing information and the natural language description of the programming operation, so as to learn and capture the interactive screening features between the user's coding operation semantics and the image semantics of the drawing object, so as to generate programming drawings that meet the needs of users. In this way, artificial intelligence technology can be used to assist the programming process of slotting machine drawing, thereby improving the intelligence level of slotting machine drawing programming, while simplifying the programming process and manual operation, which can not only reduce the operation threshold, but also effectively reduce the human burden and errors.

[0060] Specifically, in the AI-assisted programming module, first, the drawing in the file is extracted. It should be understood that in order to program the drawing according to the user's needs, it is necessary to read the drawing information in the file selected by the user, which includes but is not limited to key elements such as geometric features, size, and position of the drawing object.

[0061] In one embodiment of the present invention application, the drawing and programming operation description semantic coding unit includes: a feature extraction subunit, used to perform feature extraction based on the drawing object on the drawing to obtain initial semantic coding features of the drawing object image; a drawing object image semantic feature enhancement subunit, used to perform drawing object image semantic feature enhancement based on grid energy attention on the initial semantic coding features of the drawing object image to obtain the drawing object image semantic coding features; a semantic coding subunit, used to perform semantic coding on the natural language description of the programming operation to obtain a programming operation semantic coding feature vector as the programming operation semantic coding feature.

[0062] Next, the drawing is input into a drawing object feature extractor based on a dilated convolutional neural network model for feature mining to extract implicit semantic feature information about the drawing object in the drawing, thereby obtaining a drawing object image semantic coding feature map.

[0063] In one embodiment of the present invention, the feature extraction subunit is used to input the drawing into a drawing object feature extractor based on a hole convolutional neural network model to obtain a drawing object image semantic coding feature map as the initial semantic coding feature of the drawing object image.

[0064] It should be understood that in the actual process of image semantic recognition and understanding of drawing objects, since the drawing will not only contain semantic features related to the drawing object, but also some background interference and noise influence, in order to enhance the expressiveness and saliency of the feature map, so as to better capture the key information in the image, in the technical solution of the present invention application, the initial semantic coding features of the drawing object image are further enhanced based on the grid energy attention to obtain the semantic coding features of the drawing object image. The grid energy attention semantic feature enhancement processing method can use the grid to standardize the attention window and imitate the attention structure based on energy saliency, so as to improve the expressiveness and model performance of the semantic features of the drawing object in the drawing object image semantic coding feature map.

[0065] In one embodiment of the present invention application, the drawing object image semantic feature enhancement subunit includes: a grid division secondary subunit, which is used to grid the drawing object image semantic coding feature map to obtain a set of drawing object image semantic local feature maps; a significant feature enhancement secondary subunit, which is used to perform significant feature enhancement based on an energy significance description factor on the set of drawing object image semantic local feature maps to obtain a set of enhanced drawing object image semantic local feature maps; a feature aggregation processing secondary subunit, which is used to perform feature aggregation processing on the set of enhanced drawing object image semantic local feature maps to obtain a drawing object image semantic enhancement coding feature map as the drawing object image semantic coding feature.

[0066] Specifically, the steps of semantic feature enhancement processing with grid energy attention include: first, gridding the semantic coding feature map of the drawing object image to obtain a set of semantic local feature maps of the drawing object image, that is, decomposing the semantic coding feature map of the drawing object image into multiple local feature maps through grid division, each of the local feature maps represents a small segment in the original semantic coding feature map of the drawing object image, and this division enables the model to focus on specific areas in the feature map, which helps to identify and process local features of the semantics of the drawing object image, and provide more refined control for subsequent feature enhancement and analysis.

[0067] Next, the energy significance description factor of each drawing object image semantic local feature map is calculated using the statistical features of each drawing object image semantic local feature map in the set of drawing object image semantic local feature maps, and the energy significance description factor reflects the significance or importance of the drawing object image semantic local feature map in the global feature map. It should be understood that the calculation of the energy significance description factor helps to identify the key areas in the original drawing object image semantic local feature map. That is, to identify which parts are more important or more worthy of attention, and these areas may contain information that has an important impact on the semantic understanding of the drawing object image and the drawing programming task. These significance description factors are actually a quantification of the importance of the local feature map in the global context, so that the model can allocate attention resources more intelligently.

[0068] Then, the set of descriptors of the semantic local energy significance of the drawing object image is input into the local feature adaptive selector based on the gating function for processing, so as to use the gating function to dynamically adjust the weight according to the energy significance of each semantic local feature map of the drawing object image. This means that the model can self-regulate during the processing, strengthen those more important features related to subsequent tasks, and suppress those less relevant features, thereby enhancing the model's responsiveness to key information in the feature map, thereby improving the robustness and effectiveness of feature representation.

[0069] Then, each drawing object image semantic local significant modulation weight in the set of drawing object image semantic local significant modulation weights is used as a weight to weight each drawing object image semantic local feature map in the set of drawing object image semantic local feature maps, so as to further highlight and strengthen those drawing object image local semantic features with strong significance, while reducing the influence of noise or irrelevant features. It should be understood that the essence of the weighted enhancement step is attention application, which enables the model to more effectively utilize the key semantic information in the feature map, improves the discrimination and expression ability of the feature representation, and helps to improve the performance and generalization ability of the model in the programming task of slotting machine drawing.

[0070] Next, the set of enhanced drawing object image semantic local feature maps is feature aggregated according to the grid division method to form a drawing object image semantic local significant guided enhanced feature map. In this way, it is ensured that the enhanced drawing object image semantic local feature map can be integrated back into a complete representation form, retaining global information while enhancing local details.

[0071] Finally, considering that the attention granularity of this processing method is the grid granularity, there will be feature transition gaps at the grid boundary. In order to avoid the abnormal feature expression transition caused by grid division, a dilated convolution is used to disperse the features, that is, the semantic local significant guided enhancement feature map of the drawing object image is input into the feature dispersion module based on the dilated convolution layer to obtain the semantic enhancement encoding feature map of the drawing object image. The dilated convolution can capture more contextual information without increasing the receptive field, thereby smoothing the feature expression differences between different grids and ensuring the consistency and coherence of the entire feature map.

[0072] In one embodiment of the present invention application, the salient feature enhancement secondary subunit is used to: calculate the energy significance description factor of each drawing object image semantic local feature map in the set of drawing object image semantic local feature maps to obtain a set of drawing object image semantic local energy significance description factors; input the set of drawing object image semantic local energy significance description factors into a local feature adaptive selector based on a gating function to obtain a set of drawing object image semantic local significance modulation weights; use each drawing object image semantic local significance modulation weight in the set of drawing object image semantic local significance modulation weights as a weight, and weight each drawing object image semantic local feature map in the set of drawing object image semantic local feature maps to obtain the set of enhanced drawing object image semantic local feature maps.

[0073] Furthermore, in one embodiment of the present invention application, the feature aggregation processing secondary sub-unit is used to: perform feature aggregation on the set of enhanced drawing object image semantic local feature maps in accordance with the grid division method to obtain a drawing object image semantic local significant guided enhanced feature map; input the drawing object image semantic local significant guided enhanced feature map into a feature dispersion module based on a hole convolution layer to obtain the drawing object image semantic enhanced encoded feature map.

[0074] In the present invention, calculating the energy significance description factor of each drawing object image semantic local feature map in the set of drawing object image semantic local feature maps to obtain a set of drawing object image semantic local energy significance description factors includes:

[0075] Extracting the maximum value and the minimum value of each of the semantic local feature maps of the drawing object image respectively to obtain the maximum value of the semantic local feature of the drawing object image and the minimum value of the semantic local feature of the drawing object image;

[0076] Calculating the difference between the maximum value of the semantic local feature of the drawing object image and the minimum value of the semantic local feature of the drawing object image to obtain a difference value of the semantic local feature of the drawing object image;

[0077] Respectively calculating the global mean and variance of the semantic local feature map of the drawing object image to obtain the semantic local feature mean and the semantic local feature variance of the drawing object image;

[0078] The value obtained by multiplying the variance of the semantic local feature of the drawing object image by a constant of two and the mean value of the semantic local feature of the drawing object image are added to the hyperparameter to obtain a statistical modulation value of the semantic local feature of the drawing object image;

[0079] The difference value of the semantic local feature of the drawing object image is divided by the statistical modulation value of the semantic local feature of the drawing object image to obtain a semantic local energy significance description factor of the drawing object image corresponding to the semantic local feature map of the drawing object image.

[0080] The step of inputting the set of semantic local energy significance description factors of the drawing object image into a local feature adaptive selector based on a gating function to obtain a set of semantic local significance modulation weights of the drawing object image comprises:

[0081] Calculating an exponential function value with the natural constant e as the base and the negative number of each drawing object image semantic local energy significance description factor in the set of drawing object image semantic local energy significance description factors as the exponent to obtain a set of drawing object image semantic local energy significance description exponential factors;

[0082] Calculating the reciprocal of the sum of each drawing object image semantic local energy significant description index factor in the set of drawing object image semantic local energy significant description index factors and a constant one to obtain a set of normalized drawing object image semantic local energy significant description factors;

[0083] Inputting each normalized drawing object image semantic local energy saliency description factor in the set of normalized drawing object image semantic local energy saliency description factors into a gating function for masking processing to obtain a sequence of the drawing object image semantic local saliency modulation weights;

[0084] Among them, each normalized drawing object image semantic local energy significance description factor in the set of normalized drawing object image semantic local energy significance description factors is input into a gating function for masking processing to obtain a sequence of the drawing object image semantic local significance modulation weights, including: according to whether the normalized drawing object image semantic local energy significance description factor in the set of normalized drawing object image semantic local energy significance description factors is greater than a predetermined threshold, taking the original value of the normalized drawing object image semantic local energy significance description factor greater than the predetermined threshold, otherwise the other normalized drawing object image semantic local energy significance description factor values ​​are 0, so as to obtain the sequence of the drawing object image semantic local significance modulation weights.

[0085] Specifically, the initial semantic coding features of the drawing object image are enhanced based on grid energy attention using the following significant gated enhancement formula to obtain the semantic coding features of the drawing object image;

[0086] Wherein, the significant gated enhancement formula is:

[0087]

[0088]

[0089]

[0090]

[0091]

[0092]

[0093] in, A semantic encoding feature map for the drawing object image, is the meshing operation, are respectively the semantic local feature maps of the drawing object images in the set of the semantic local feature maps of the drawing object images, is the first in the set of semantic local feature maps of the drawing object image A semantic local feature map of the drawing object image, and Extract the maximum and minimum values ​​in the feature map respectively. and They are The mean and variance of is a hyperparameter, Indicates The semantic local energy saliency description factor of the drawing object image corresponding to the semantic local feature map of the drawing object image, For masking, It is The semantic local salient modulation weights of the drawing object image corresponding to the semantic local feature maps of the drawing object image, is a predetermined threshold, To perform feature aggregation operations on each weighted feature map, is the semantically local significant guided enhanced feature map of the drawing object image, is the dilated convolution operation, It is the semantic enhancement encoding feature map of the drawing object image.

[0094] Furthermore, semantic coding is performed on the natural language description of the programming operation to extract semantic coding features in the natural language description of the programming operation, so as to understand the user's drawing programming requirements, thereby obtaining a programming operation semantic coding feature vector.

[0095] It should be understood that since the drawing object image semantic enhancement coding feature map and the programming operation semantic coding feature vector respectively contain the semantic enhancement coding features and coding operation description semantic coding features related to the drawing object image, in order to select those parts that best reflect the programming intention or are most associated with the programming operation from the drawing object image semantic coding features, so as to more accurately generate the post-programming drawing that meets the user's needs, in the technical solution of the present invention application, the drawing object image semantic coding features and the programming operation semantic coding features are further subjected to feature interaction screening based on metadata blocks to obtain programming semantic screening drawing object coding features. The feature interaction screening process based on metadata blocks is a new feature cross-modal filtering process that uses meta-modal feature vectors to perform feature interaction and screening on the main modal feature map. Its main function is to extract information from one modality (meta-modality) to guide the feature screening of another modality (main modality). In this scenario, the "drawing object image semantic coding features" can be regarded as the main modal feature map, and the "programming operation semantic coding features" can be regarded as the meta-modal feature vector. The metadata block-based feature interactive screening of the drawing object image semantic enhancement coding feature map and the programming operation semantic coding feature vector is performed in order to select those parts of the drawing object image semantic coding features that best reflect the programming intention or are most associated with the programming operation, so as to obtain a more refined and targeted feature, namely the "programming semantically filtered drawing object coding feature".

[0096] In one embodiment of the present invention application, the feature interaction screening unit includes: a linear transformation and mask processing sub-unit, which is used to perform linear transformation and mask processing on the drawing object image semantic coding features and the programming operation semantic coding features respectively to obtain a masked first programming operation semantic coding auxiliary feature mapping pattern representation and a masked second programming operation semantic coding auxiliary feature mapping pattern representation; a feature screening sub-unit, which is used to perform feature screening on the drawing object image semantic coding features based on the masked first programming operation semantic coding auxiliary feature mapping pattern representation and the masked second programming operation semantic coding auxiliary feature mapping pattern representation to obtain the programming semantically screened drawing object coding features.

[0097] Specifically, the step of feature interaction screening based on metadata blocks includes: first, the programming operation semantic coding features (metamodal feature vectors) are linearly transformed to generate two auxiliary feature mapping mode representation vectors. This step enables the features of the programming operation to be expressed in a new form, which is convenient for comparison and interaction with the drawing object image semantic coding features (main modal feature map). Next, the two programming operation semantic coding auxiliary feature mapping mode representation vectors are subjected to gradient amplitude-based masking. This step emphasizes the key parts of the programming operation semantic coding features to facilitate the subsequent feature screening process. Then, the drawing object image semantic enhancement coding feature map is preliminarily screened using the masked first programming operation semantic coding auxiliary feature mapping mode representation vector combined with the hyperbolic tangent function. This step helps to select those image features that are closely related to the programming operation. Finally, the results of the preliminary screening are further refined using the second programming operation semantic coding auxiliary feature mapping mode representation vector combined with the Sigmoid function. The Sigmoid function can help determine which features are most important in the programming semantic context, thereby generating the final programming semantic screening drawing object coding feature map.

[0098] In one embodiment of the present invention, the linear transformation and mask processing subunit is used to: perform a linear transformation on the programming operation semantic coding feature vector based on the first weight modifier matrix and the second weight modifier matrix to obtain a first programming operation semantic coding auxiliary feature mapping pattern representation vector and a second programming operation semantic coding auxiliary feature mapping pattern representation vector; based on the gradient amplitude of each position in the first programming operation semantic coding auxiliary feature mapping pattern representation vector, perform a mask processing based on the gradient amplitude on the first programming operation semantic coding auxiliary feature mapping pattern representation vector to obtain a masked first programming operation semantic coding auxiliary feature mapping pattern representation vector as the masked first programming operation semantic coding auxiliary feature mapping pattern representation; based on the gradient amplitude of each position in the second programming operation semantic coding auxiliary feature mapping pattern representation vector, perform a mask processing based on the gradient amplitude on the second programming operation semantic coding auxiliary feature mapping pattern representation vector to obtain the masked second programming operation semantic coding auxiliary feature mapping pattern representation vector as the masked second programming operation semantic coding auxiliary feature mapping pattern representation.

[0099] Furthermore, in one embodiment of the present invention, the feature screening subunit is used to: perform preliminary feature screening on the drawing object image semantic enhancement coding feature map based on the masked first programming operation semantic coding auxiliary feature mapping pattern representation vector and the hyperbolic tangent function to obtain the drawing object image semantic feature map after preliminary screening; perform feature screening again on the drawing object image semantic feature map after preliminary screening based on the masked second programming operation semantic coding auxiliary feature mapping pattern representation vector and the Sigmoid function to obtain the programming semantic screening drawing object coding feature map as the programming semantic screening drawing object coding feature.

[0100] In the present invention, the programming operation semantic coding feature vector is subjected to a linear transformation based on a first weight modifier matrix and a second weight modifier matrix to obtain a first programming operation semantic coding auxiliary feature mapping mode representation vector and a second programming operation semantic coding auxiliary feature mapping mode representation vector, including:

[0101] Calculate the programming operation semantic encoding feature vector and the transpose of the first weight modifier matrix, multiply the result, and then add the result to the first bias modifier vector by position to obtain the first programming operation semantic encoding auxiliary feature mapping mode representation vector;

[0102] The programming operation semantic encoding feature vector is calculated and multiplied by the transpose of the second weight modifier matrix, and then added to the second bias modifier vector by position to obtain the second programming operation semantic encoding auxiliary feature mapping mode representation vector.

[0103] Wherein, based on the gradient amplitude of each position in the first programming operation semantic encoding auxiliary feature mapping pattern representation vector, performing a masking process based on the gradient amplitude on the first programming operation semantic encoding auxiliary feature mapping pattern representation vector to obtain a masked first programming operation semantic encoding auxiliary feature mapping pattern representation vector as the masked first programming operation semantic encoding auxiliary feature mapping pattern representation, including:

[0104] Calculating the square of the difference between the previous position feature value and the next position feature value of each position feature value in the first programming operation semantic encoding auxiliary feature mapping mode representation vector to obtain a first programming operation semantic gradient amplitude representation vector;

[0105] Calculating the absolute value of the square root of each first programming operation semantic gradient amplitude in the first programming operation semantic gradient amplitude representation vector, and inputting the obtained feature vector into the mask function for processing to obtain the masked first programming operation semantic encoding auxiliary feature mapping pattern representation vector;

[0106] Among them, in the mask function, in response to the first programming operation semantic gradient amplitude in the first programming operation semantic gradient amplitude representation vector being greater than a predetermined threshold, the first programming operation semantic gradient amplitude greater than the predetermined threshold is set to 1, otherwise it is set to 0, so as to obtain the masked first programming operation semantic encoding auxiliary feature mapping pattern representation vector.

[0107] Furthermore, based on the masked first programming operation semantic encoding auxiliary feature mapping pattern representation vector and the hyperbolic tangent function, preliminary feature screening is performed on the drawing object image semantic enhancement encoding feature map to obtain the drawing object image semantic feature map after preliminary screening, including: using the masked first programming operation semantic encoding auxiliary feature mapping pattern representation vector as the channel weight vector, weightedly enhancing the drawing object image semantic enhancement encoding feature map, and then inputting the weighted enhanced feature map into the hyperbolic tangent function for activation to obtain the drawing object image semantic feature map after preliminary screening.

[0108] Furthermore, based on the masked second programming operation semantic encoding auxiliary feature mapping pattern representation vector and the Sigmoid function, the image semantic feature map of the drawing object after the preliminary screening is subjected to feature screening again to obtain the programming semantic screening drawing object encoding feature map as the programming semantic screening drawing object encoding feature, including: using the masked second programming operation semantic encoding auxiliary feature mapping pattern representation vector as the channel weight vector, performing weighted enhancement on the image semantic feature map of the drawing object after the preliminary screening, and inputting the weighted enhanced feature map into the Sigmoid function for activation to obtain the programming semantic screening drawing object encoding feature map.

[0109] Specifically, the drawing object image semantic coding feature and the programming operation semantic coding feature are subjected to feature interactive screening based on metadata blocks using the following feature interactive screening formula to obtain a programming semantic screening drawing object coding feature graph as the programming semantic screening drawing object coding feature;

[0110] The feature interaction screening formula includes:

[0111]

[0112]

[0113]

[0114]

[0115]

[0116]

[0117]

[0118] in, is the feature vector encoding the programming operation semantics, and are the first weight modifier matrix and the first bias modifier vector, respectively, and are the second weight modifier matrix and the second bias modifier vector, respectively, and are the first linear transformation and the second linear transformation respectively, and are respectively a first programming operation semantics encoding auxiliary feature mapping mode representation vector and a second programming operation semantics encoding auxiliary feature mapping mode representation vector, and The first programming operation semantics encoding auxiliary feature mapping mode representation vector is respectively Position and The eigenvalues ​​at the positions, and The first and second programming operation semantics encoding auxiliary feature mapping mode representation vectors are respectively Position and The eigenvalues ​​at the positions, For mask operation, and The first masked first programming operation semantics encoding auxiliary feature mapping mode representation vector and the second masked second programming operation semantics encoding auxiliary feature mapping mode representation vector are respectively The eigenvalues ​​at the positions, is the preset threshold, A semantic enhancement encoding feature map for the drawing object image, encoding an auxiliary feature map pattern representation vector for the masked first programming operation semantics, encoding an auxiliary feature map pattern representation vector for the masked second programming operation semantics, is the position-wise weighted multiplication process along the channel dimension, is the hyperbolic tangent function, After preliminary screening, the semantic feature map of the drawing object image is obtained. for function, A drawing object encoding feature map is screened for the programming semantics.

[0119] Then, the programming semantics-screened drawing object coding feature map is input into the decoder-based drawing coding generation module to obtain the programmed drawing. That is to say, the drawing object coding features that have been screened by programming semantics are decoded and regressed to generate a programmed drawing that meets user needs. In this way, artificial intelligence technology can be used to assist in the programming process of the planer drawing, thereby improving the intelligence level of the planer drawing programming, while simplifying the programming process and manual operation, which can not only lower the operating threshold, but also effectively reduce the human burden and errors.

[0120] In one embodiment of the present invention, the post-programming drawing generation module is used to: input the programming semantics screening drawing object encoding feature map into a decoder-based drawing code generation module to obtain the post-programming drawing.

[0121] In a preferred example, the step of passing the programming semantics screening drawing object encoding feature map through a decoder-based drawing encoding generation module to obtain a programmed drawing includes:

[0122] Calculating the sum of the absolute values ​​of each feature value of the programming semantic screening drawing object encoding feature map to obtain a first programming semantic screening drawing object encoding and modulation value, and calculating the square root of the sum of the squares of each feature value of the programming semantic screening drawing object encoding feature map to obtain a second programming semantic screening drawing object encoding and modulation value;

[0123] After performing dot-wise subtraction on the programming semantics screening drawing object encoding feature map and the second programming semantics screening drawing object encoding and modulation value, perform dot-wise multiplication on the number of feature values ​​of the programming semantics screening drawing object encoding feature map and the reciprocal of the first programming semantics screening drawing object encoding and modulation value, and take the reciprocal of each feature value to obtain the first programming semantics screening drawing object encoding phase conversion feature map;

[0124] After subtracting the programming semantics screening drawing object encoding feature map from the first programming semantics screening drawing object code and modulation value, dot multiplication is performed with the square root of the number of eigenvalues ​​of the programming semantics screening drawing object encoding feature map and the reciprocal of the second programming semantics screening drawing object code and modulation value, and the reciprocal of each eigenvalue is taken to obtain the second programming semantics screening drawing object encoding phase transition feature map; and subtracting the first programming semantics screening drawing object encoding phase transition feature map from the dot product feature map of the second programming semantics screening drawing object encoding phase transition feature map and the weighted hyperparameter to obtain an optimized programming semantics screening drawing object encoding feature map; and passing the optimized programming semantics screening drawing object encoding feature map through a decoder-based drawing code generation module to obtain a programmed drawing.

[0125] Here, the programming semantics screens the drawing object encoding feature map, denoted as The optimization expression is:

[0126]

[0127]

[0128]

[0129]

[0130]

[0131]

[0132]

[0133] in, is each feature value of the feature map encoded by the programming semantics screening drawing object, is the first programming semantic filter drawing object encoding and modulation value, is the second programming semantics filter drawing object encoding and modulation value, is the programming semantics screening drawing object encoding feature map, is the number of eigenvalues ​​of the feature map encoded by the programming semantic screening drawing object, is the reciprocal of each eigenvalue, is the first programming semantic filter drawing object encoding phase transition feature map, The second programming semantics filter drawing object encodes the phase transition feature map, is a weighted hyperparameter, It is an optimized programming semantics filter drawing object encoding feature map, is the length of the feature map encoded by the programming semantics filter drawing object, is the number of channels of the feature map encoded by the programming semantics filter drawing object, is the height of the feature map encoded by the programming semantics filter drawing object, It is the point multiplication by position. It is subtracted by position point. Represents the set of real numbers.

[0134] In particular, in the technical solution of the present invention, considering that the drawing object image semantic enhancement coding feature map and the programming operation semantic coding feature vector respectively represent the image semantics of the drawing in the archive based on the grid energy significance guided enhancement features and the text semantic features of the natural language description of the programming operation, when they are passed through the feature interaction screening encoder based on the metadata block, the feature interaction screening results based on the cross-modal feature distribution will also have cross-modal fusion context association interaction coding differences, so it is expected to improve the detail semantic aggregation expression effect of the programming semantic screening drawing object coding feature map based on multi-modal semantic association interaction screening representation differences.

[0135] Therefore, in the preferred example, the difference of the feature value of the programming semantic screening drawing object coding feature map relative to the overall feature set of the programming semantic screening drawing object coding feature map and the modulation representation is used as the semantic change intensity information, and the phase-like conversion corresponding to the position-based intensity modulation is performed through the different modulation representation forms, so that the aggregation enhancement of the semantic change phase perception can improve the axial aggregation receptive field along the feature aggregation direction by performing the spatial translation operation based on alternating stacking under the set scale balance of the programming semantic screening drawing object coding feature map, thereby improving the perception effect of the aggregation semantics of the programming semantic screening drawing object coding feature map for the detail semantic change, so as to improve the expression effect of the programming semantic screening drawing object coding feature map, and improve the accuracy of the programming drawing obtained by the decoder-based drawing coding generation module. In this way, a more accurate programming drawing that meets user needs can be generated. In this way, artificial intelligence technology can be used to assist the programming process of the planing machine drawing, thereby improving the intelligence level of the planing machine drawing programming, while simplifying the programming process and manual operation, which can not only reduce the operation threshold, but also effectively reduce the human burden and errors.

[0136] In summary, the above scheme is adopted to extract the drawings in the archives and collect the natural language description of the programming operation input by the user, and then introduce the image processing algorithm and natural language processing technology based on artificial intelligence and deep learning in the back end to analyze the drawing information and the natural language description of the programming operation, so as to learn and capture the interactive screening features between the user's coding operation semantics and the image semantics of the drawing object, so as to generate a post-programming drawing that meets the user's needs. In this way, artificial intelligence technology can be used to assist the programming process of the planing machine drawing, thereby improving the intelligent level of the planing machine drawing programming, while simplifying the programming process and manual operation, which can not only lower the operation threshold, but also effectively reduce the human burden and errors.

[0137] Second, Figure 2 The present invention is a flowchart of an AI-assisted slotting machine drawing programming method according to an exemplary embodiment. The AI-assisted slotting machine drawing programming method adopts any one of the AI-assisted slotting machine drawing programming systems described in the first aspect, and includes the following steps:

[0138] Step S201, entering the file management interface;

[0139] Step S202, loading the selected file;

[0140] Step S203, receiving a natural language description of a programming operation input by a user;

[0141] Step S204, programming the drawing in the file based on the natural language description of the programming operation to obtain a programmed drawing;

[0142] Wherein, step S204, programming the drawing in the file based on the natural language description of the programming operation to obtain a programmed drawing, includes:

[0143] Step S2041, extracting the drawing in the file;

[0144] Step S2042, semantically encode the natural language description of the drawing and the programming operation to obtain a drawing object image semantic encoding feature and a programming operation semantic encoding feature;

[0145] Step S2043, performing feature interactive screening on the drawing object image semantic coding features and the programming operation semantic coding features based on metadata blocks to obtain programming semantic screening drawing object coding features;

[0146] Step S2044: Screening drawing object coding features based on the programming semantics to generate the post-programming drawing.

[0147] In the present invention application, the AI-assisted slotting machine drawing programming method protects any technical scheme and all technical effects of the AI-assisted slotting machine drawing programming system in the first aspect, and the present invention application will not be repeated here.

[0148] In the AI-assisted planing machine drawing programming method applied in the present invention, by extracting the drawings in the archives and collecting the natural language description of the programming operation input by the user, the image processing algorithm and natural language processing technology based on artificial intelligence and deep learning are introduced in the back end to analyze the drawing information and the natural language description of the programming operation, so as to learn and capture the interactive screening features between the user's coding operation semantics and the drawing object image semantics, so as to generate a post-programming drawing that meets the user's needs. In this way, artificial intelligence technology can be used to assist the programming process of the planing machine drawing, thereby improving the intelligent level of the planing machine drawing programming, while simplifying the programming process and manual operation, which can not only lower the operation threshold, but also effectively reduce the human burden and errors.

[0149] The terminal devices in the embodiments of the present invention may include but are not limited to mobile terminals such as mobile phones, laptop computers, digital broadcast receivers, PDAs (personal digital assistants), PADs (tablet computers), PMPs (portable multimedia players), etc., and fixed terminals such as digital TVs, desktop computers, etc. The above-mentioned electronic devices are only examples and should not bring any limitations to the functions and scope of use of the embodiments of the present invention.

[0150] An electronic device may include a processing device (e.g., a central processing unit, a graphics processing unit, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) or a program loaded from a storage device into a random access memory (RAM). In the RAM, various programs and data required for the operation of the electronic device are also stored. The processing device, ROM, and RAM are connected to each other via a bus. An input / output (I / O) interface is also connected to the bus.

[0151] Typically, the following devices can be connected to the I / O interface: input devices including, for example, a touch screen, a touchpad, a keyboard, a mouse, a camera, a microphone, an accelerometer, a gyroscope, etc.; output devices including, for example, a liquid crystal display (LCD), a speaker, a vibrator, etc.; storage devices including, for example, a magnetic tape, a hard disk, etc.; and communication devices. The communication device can allow the electronic device to communicate with other devices wirelessly or by wire to exchange data. Although electronic devices with various devices are described in the embodiments of the present application, it should be understood that it is not required to implement or have all the devices shown. More or fewer devices may be implemented or provided alternatively.

[0152] In particular, according to an embodiment of the present invention, the process described above with reference to the flowchart can be implemented as a computer software program. For example, an embodiment of the present invention includes a computer program product, which includes a computer program carried on a non-transitory computer-readable medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from a network via a communication device, or installed from a storage device, or installed from a ROM. When the computer program is executed by a processing device, the above-mentioned functions defined in the method of the embodiment of the present invention are executed.

[0153] It should be noted that the computer-readable medium mentioned in the present application may be a computer-readable signal medium or a computer-readable storage medium or any combination of the above two. The computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples of computer-readable storage media may include, but are not limited to: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in combination with an instruction execution system, device or device. In the present application, a computer-readable signal medium may include a data signal propagated in a baseband or as part of a carrier wave, which carries a computer-readable program code. This propagated data signal may take a variety of forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. The computer readable signal medium may also be any computer readable medium other than a computer readable storage medium, which may send, propagate or transmit a program for use by or in conjunction with an instruction execution system, apparatus or device. The program code contained on the computer readable medium may be transmitted using any suitable medium, including but not limited to: wires, optical cables, RF (radio frequency), etc., or any suitable combination of the above.

[0154] In some embodiments, the client and the server may communicate using any currently known or future developed network protocol such as HTTP (HyperText Transfer Protocol), and may be interconnected with any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include local area networks (LANs), wide area networks (WANs), internets (e.g., the Internet), and peer-to-peer networks, as well as any currently known or future developed networks. The above-mentioned computer-readable medium may be included in the above-mentioned electronic device; or it may exist separately without being assembled into the electronic device.

[0155] Computer program code for performing the operations of the present application may be written in one or more programming languages ​​or a combination thereof, including but not limited to object-oriented programming languages, such as Java, Smalltalk, C++, and conventional procedural programming languages, such as "C" or similar programming languages. The program code may be executed entirely on the user's computer, partially on the user's computer, as a separate software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0156] The flow chart and block diagram in the accompanying drawings illustrate the possible architecture, function and operation of the system, method and computer program product according to various embodiments of the present invention. In this regard, each box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or flow chart, and the combination of the boxes in the block diagram and / or flow chart can be implemented with a dedicated hardware-based system that performs a specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0157] The modules involved in the embodiments of the present invention can be implemented by software or hardware. The name of the module does not limit the module itself in some cases. For example, the test parameter acquisition module can also be described as a "module for acquiring device test parameters corresponding to the target device".

[0158] The functions described above herein may be performed at least in part by one or more hardware logic components. For example, without limitation, exemplary types of hardware logic components that may be used include: field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), complex programmable logic devices (CPLDs), and the like.

[0159] In the context of the present application, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0160] The above description is only a preferred embodiment of the present invention and an explanation of the technical principles used. Those skilled in the art should understand that the scope of disclosure involved in the present invention is not limited to the technical solution formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the above disclosed concept. For example, the above features are replaced with the technical features with similar functions disclosed in the present invention (but not limited to) by each other.

[0161] In addition, although each operation is described in a specific order, this should not be construed as requiring these operations to be performed in the specific order shown or to be performed in a sequential order. Under certain circumstances, multitasking and parallel processing may be advantageous. Similarly, although some specific implementation details are included in the above discussion, these should not be construed as limiting the scope of the present application. Some features described in the context of a separate embodiment can also be implemented in a single embodiment in combination. On the contrary, the various features described in the context of a single embodiment can also be implemented in multiple embodiments individually or in any suitable sub-combination mode.

[0162] Although the subject matter has been described in language specific to structural features and / or method logic actions, it should be understood that the subject matter defined is not necessarily limited to the specific features or actions described above. On the contrary, the specific features and actions described above are merely exemplary forms of implementation. With respect to the apparatus in the above-mentioned embodiments, the specific manner in which each module performs the operation has been described in detail in the embodiments related to the method, and will not be described in detail here.

Claims

1. An AI-assisted planing machine drawing programming system, characterized in that: include: Archives management module, used to enter the archives management interface; File loading module, used to load the selected file; A programming requirement input module, used to receive a natural language description of a programming operation input by a user; An AI-assisted programming module, for programming the drawing in the file based on the natural language description of the programming operation to obtain a programmed drawing; Wherein, the AI-assisted programming module includes: A drawing extraction unit, used for extracting drawings from the archive; A drawing and programming operation description semantic coding unit, used for semantically coding the drawing and the programming operation natural language description to obtain drawing object image semantic coding features and programming operation semantic coding features; A feature interactive screening unit, used for performing feature interactive screening on the drawing object image semantic coding features and the programming operation semantic coding features based on metadata blocks to obtain programming semantic screening drawing object coding features; A post-programming drawing generation module, used for screening drawing object coding features based on the programming semantics to generate the post-programming drawing; The drawing and programming operations describe a semantic coding unit, including: A feature extraction subunit, configured to extract features of the drawing based on the drawing object to obtain initial semantic coding features of the drawing object image; A drawing object image semantic feature enhancement subunit, used for performing grid energy-focused drawing object image semantic feature enhancement on the initial semantic coding feature of the drawing object image to obtain the drawing object image semantic coding feature; A semantic encoding subunit, configured to semantically encode the natural language description of the programming operation to obtain a programming operation semantic encoding feature vector as the programming operation semantic encoding feature; The feature extraction subunit is used to input the drawing into a drawing object feature extractor based on a hole convolutional neural network model to obtain a drawing object image semantic coding feature map as the initial semantic coding feature of the drawing object image; The drawing object image semantic feature enhancement subunit includes: A grid division secondary sub-unit, used for performing grid division on the drawing object image semantic coding feature map to obtain a set of drawing object image semantic local feature maps; A salient feature enhancement secondary subunit, configured to perform salient feature enhancement on the set of semantic local feature maps of the drawing object image based on an energy saliency description factor to obtain a set of enhanced semantic local feature maps of the drawing object image; A feature aggregation processing secondary subunit is used to perform feature aggregation processing on the set of enhanced drawing object image semantic local feature maps to obtain a drawing object image semantic enhancement coding feature map as the drawing object image semantic coding feature; The significant feature enhancement secondary subunit is used to: Calculating the energy significance description factor of each drawing object image semantic local feature map in the set of drawing object image semantic local feature maps to obtain a set of drawing object image semantic local energy significance description factors; Inputting the set of semantic local energy significance descriptors of the drawing object image into a local feature adaptive selector based on a gating function to obtain a set of semantic local significance modulation weights of the drawing object image; Taking each drawing object image semantic local saliency modulation weight in the set of drawing object image semantic local saliency modulation weights as a weight, each drawing object image semantic local feature map in the set of drawing object image semantic local feature maps is weighted respectively to obtain the set of enhanced drawing object image semantic local feature maps.

2. The AI-assisted planing machine drawing programming system according to claim 1 is characterized in that: The feature aggregation processing secondary subunit is used to: Performing feature aggregation on the set of the enhanced drawing object image semantic local feature maps according to the grid division method to obtain the drawing object image semantic local significant guided enhanced feature map; The semantically local significant guided enhanced feature map of the drawing object image is input into a feature dispersion module based on a hole convolution layer to obtain the semantically enhanced encoded feature map of the drawing object image.

3. The AI-assisted planing machine drawing programming system according to claim 2 is characterized in that: The feature interactive screening unit comprises: A linear transformation and mask processing subunit, used for performing linear transformation and mask processing on the image semantic coding features of the drawing object and the programming operation semantic coding features respectively to obtain a masked first programming operation semantic coding auxiliary feature mapping mode representation and a masked second programming operation semantic coding auxiliary feature mapping mode representation; The feature screening subunit is used to perform feature screening on the image semantic coding features of the drawing object based on the masked first programming operation semantic coding auxiliary feature mapping mode representation and the masked second programming operation semantic coding auxiliary feature mapping mode representation to obtain the programming semantic screening drawing object coding features.

4. The AI-assisted planing machine drawing programming system according to claim 3 is characterized in that: The linear transformation and mask processing subunit is used for: Performing a linear transformation based on a first weight modifier matrix and a second weight modifier matrix on the programming operation semantic encoding feature vector to obtain a first programming operation semantic encoding auxiliary feature mapping mode representation vector and a second programming operation semantic encoding auxiliary feature mapping mode representation vector; Based on the gradient magnitude of each position in the first programming operation semantically encoded auxiliary feature mapping pattern representation vector, performing a masking process based on the gradient magnitude on the first programming operation semantically encoded auxiliary feature mapping pattern representation vector to obtain a masked first programming operation semantically encoded auxiliary feature mapping pattern representation vector as the masked first programming operation semantically encoded auxiliary feature mapping pattern representation; Based on the gradient amplitude of each position in the second programming operation semantically encoded auxiliary feature mapping pattern representation vector, the second programming operation semantically encoded auxiliary feature mapping pattern representation vector is subjected to gradient amplitude-based mask processing to obtain the masked second programming operation semantically encoded auxiliary feature mapping pattern representation vector as the masked second programming operation semantically encoded auxiliary feature mapping pattern representation.

5. The AI-assisted planing machine drawing programming system according to claim 4 is characterized in that: The feature screening subunit is used to: Performing preliminary feature screening on the drawing object image semantic enhancement coding feature map based on the masked first programming operation semantic coding auxiliary feature mapping pattern representation vector and the hyperbolic tangent function to obtain a drawing object image semantic feature map after preliminary screening; Based on the masked second programming operation semantic coding auxiliary feature mapping pattern representation vector and the Sigmoid function, the initially screened drawing object image semantic feature map is subjected to feature screening again to obtain the programming semantic screening drawing object coding feature map as the programming semantic screening drawing object coding feature; The post-programming drawing generation module is used to: input the programming semantics screening drawing object encoding feature map into the decoder-based drawing code generation module to obtain the post-programming drawing.

6. An AI-assisted planing machine drawing programming method, using an AI-assisted planing machine drawing programming system according to any one of claims 1 to 5, characterized in that: The following steps are involved: Step S201, entering the file management interface; Step S202, loading the selected file; Step S203: receiving a natural language description of a programming operation input by a user; Step S204, programming the drawing in the file based on the natural language description of the programming operation to obtain a programmed drawing; Wherein, step S204, programming the drawing in the file based on the natural language description of the programming operation to obtain a programmed drawing, includes: Step S2041, extracting the drawing in the file; Step S2042, semantically encode the natural language description of the drawing and the programming operation to obtain a drawing object image semantic encoding feature and a programming operation semantic encoding feature; Step S2043, performing feature interactive screening on the drawing object image semantic coding features and the programming operation semantic coding features based on metadata blocks to obtain programming semantic screening drawing object coding features; Step S2044: Screening drawing object coding features based on the programming semantics to generate the post-programming drawing.

Images