Parameterization extraction method, device and equipment for sketch of hand-drawn part and medium

Abstract

The invention provides a parameterized extraction method, device and equipment for a sketch of a hand-drawn part and a medium, which are used for quickly drawing the part by using the sketch of the hand-drawn part. The parameterization extraction method for the sketch of the hand-drawn part comprises the following steps: establishing a standardized parameter representation system of geometric primitives; extracting image features of the sketch of the hand-drawn part through the primitive analysis model, and predicting to obtain a prediction primitive set conforming to the parameter representation system; establishing an optimal matching relationship between the prediction primitive set and the real primitive set by adopting an optimal matching algorithm; calculating a multi-task loss function including primitive type loss, attribute loss and geometric parameter loss based on the matching relationship; and optimizing parameters of the primitive analysis model through the multi-task loss function.

Description

Technical Field

[0001] The present invention relates to the field of automobile design and manufacturing, and in particular to a parameterized extraction method, device, equipment and medium for hand-drawn component sketches. Background Art

[0002] With the continuous increase in car ownership and the rapid development of new energy vehicles, the automotive parts industry is facing unprecedented development opportunities. However, faced with growing market demand and complex design requirements, traditional component design methods are no longer able to meet the requirements for efficient and rapid design. To improve the efficiency of automotive component design and reduce the workload of designers, many companies and enterprises are committed to finding faster and more efficient solutions.

[0003] Hand-drawn sketches of automotive parts are a concise and indirect representation of the actual three-dimensional automotive parts being envisioned. In real-world design environments, mechanical designers typically rely on three-view drawings for initial design of automotive parts. However, this traditional design approach presents several challenges. First, using specific modeling software (such as AutoCAD and SolidWorks) requires extensive learning time and specialized knowledge to master the software, which can be challenging for non-professionals. Second, creating parts requires tedious interactive operations, consuming significant time and effort, and resulting in low efficiency. Therefore, utilizing sketches as a prerequisite for efficiently and conveniently creating the corresponding parts has become a pressing issue. Summary of the Invention

[0004] The present invention provides a parameterized extraction method, device, equipment and medium for hand-drawn component sketches, which are used to realize rapid component drawing using hand-drawn component sketches.

[0005] The technical solution of the present invention is:

[0006] In one aspect, the present application provides a parametric extraction method for hand-drawn component sketches, comprising:

[0007] Establish a standardized parameter representation system for geometric primitives;

[0008] Extracting image features of the hand-drawn component sketch through a primitive parsing model and predicting a set of predicted primitives that conform to the parameter representation system;

[0009] Using an optimal matching algorithm to establish an optimal matching relationship between the predicted primitive set and the real primitive set;

[0010] Based on the matching relationship, a multi-task loss function including primitive type loss, attribute loss and geometric parameter loss is calculated;

[0011] The parameters of the primitive parsing model are optimized by the multi-task loss function.

[0012] Preferably, the step of establishing a standardized parameter representation system for geometric primitives comprises:

[0013] Establish a standardized parameter representation system that includes type identification, attribute parameters and geometric parameters of geometric primitives.

[0014] Preferably, the step of extracting image features of the hand-drawn component sketch by the primitive parsing model and predicting a set of predicted primitives that conform to the parameter representation system includes:

[0015] Divide the input hand-drawn component sketch image into multiple non-overlapping image blocks of fixed size;

[0016] Use the feature extraction network to extract features from each image block to obtain multiple image block sequence feature sequences;

[0017] Use Transformer encoder to fuse all image block sequence features into a global feature sequence;

[0018] Adopting the Transformer decoder and introducing multiple query vectors, the primitive features corresponding to each query vector are extracted from the global feature sequence;

[0019] The primitive features corresponding to each query vector output by the Transformer decoder are fed into multiple multi-layer perceptron fully connected networks, and the primitive type, Boolean attributes and primitive parameters of each primitive feature are respectively determined to generate a set of predicted primitives that conform to the standardized parameter representation system.

[0020] Preferably, the step of using an optimal matching algorithm to establish an optimal matching relationship between the predicted primitive set and the real primitive set includes:

[0021] Calculate the matching cost between each predicted primitive in the predicted primitive set and each true primitive in the true primitive set to form a cost matrix;

[0022] The cost matrix is processed using the Hungarian algorithm to find the optimal matching solution with the minimum cost;

[0023] Determining, according to the optimal matching solution, a matching relationship between each predicted primitive in the predicted primitive set and a corresponding real primitive in the real primitive set;

[0024] Unmatched prediction primitives are marked as invalid predictions.

[0025] Preferably, based on the matching relationship, the step of calculating a multi-task loss function including primitive type loss, attribute loss and geometric parameter loss includes:

[0026] According to the matching relationship between the predicted primitive set and the real primitive set, the predicted primitives and the real primitives involved in the loss calculation are determined;

[0027] Calculate the primitive type loss and use the cross entropy function to measure the difference between the predicted primitive type and the true primitive type;

[0028] Calculate attribute loss and use the cross entropy function to measure the difference between the predicted Boolean attribute and the true Boolean attribute;

[0029] Calculate the geometric parameter loss and use the standard cross entropy function to measure the difference between the predicted geometric parameters and the true geometric parameters;

[0030] The primitive type loss, attribute loss and geometric parameter loss are weighted and summed to obtain the multi-task loss function.

[0031] Preferably, one query vector corresponds to one primitive object in the hand-drawn component sketch.

[0032] Preferably, the geometric primitive is a two-dimensional curve, which includes a straight line, a circle and / or an arc.

[0033] On the other hand, the present application also provides a parametric extraction device for hand-drawn component sketches, comprising:

[0034] Standard establishment module, used to establish a standardized parameter representation system for geometric primitives;

[0035] A prediction primitive set prediction module is used to extract image features of the hand-drawn component sketch through a primitive parsing model and predict a prediction primitive set that conforms to the parameter representation system;

[0036] A matching relationship establishment module, configured to establish an optimal matching relationship between the predicted primitive set and the real primitive set using an optimal matching algorithm;

[0037] A multi-task loss function calculation module, used to calculate a multi-task loss function including primitive type loss, attribute loss and geometric parameter loss based on the matching relationship;

[0038] A model optimization module is used to optimize the parameters of the primitive parsing model through the multi-task loss function.

[0039] On the other hand, the present application also provides a control device, including a processor, a memory, and a program or instruction stored in the memory and executable on the processor. When the program or instruction is executed by the processor, the steps of the parametric extraction method of the hand-drawn component sketch as described above are implemented.

[0040] On the other hand, the present application also provides a readable storage medium, which stores a program or instruction. When the program or instruction is executed by a processor, the steps of the parametric extraction method of the hand-drawn component sketch as described above are implemented.

[0041] The beneficial effects of the present invention are:

[0042] By parsing hand-drawn sketches into a set of parametric primitives, these primitives can be directly imported into computer-aided design (CAD) software to generate precise component models. Parametric primitives provide precise geometric information, including primitive types, Boolean attributes, and geometric parameters. This information can be directly used for precise modeling in CAD software, reducing human error and improving component model accuracy. Furthermore, this process is fast and accurate, avoiding the inefficiencies and inaccuracies of manual sketch parsing. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] Figure 1 Schematic diagram of the process of parametric extraction of hand-drawn component sketches in an embodiment of the present invention;

[0044] Figure 2 This is a diagram showing the overall architecture of a parametric extraction method for hand-drawn component sketches in an embodiment of the present invention;

[0045] Figure 3 Schematic diagram of the primitive parsing model in this embodiment;

[0046] Figure 4 Schematic diagram of the sketch primitive reasoning process based on the query vector in this embodiment;

[0047] Figure 5 This is a structural block diagram of a parametric extraction device for hand-drawn component sketches in an embodiment of the present invention. DETAILED DESCRIPTION

[0048] To facilitate understanding by those skilled in the art, the present invention is further described below with reference to the accompanying drawings. The detailed description is complete, but it should not be construed as limiting the scope of the present invention. Obvious variations and alternative forms of the following examples are all within the scope of protection of this patent.

[0049] Reference Figure 1 , an embodiment of the present application provides a parametric extraction method for hand-drawn component sketches, comprising:

[0050] S1, establish a standardized parameter representation system for geometric primitives;

[0051] S2, extracting image features of the hand-drawn component sketch through a primitive parsing model and predicting a set of predicted primitives that conform to the parameter representation system;

[0052] S3, using an optimal matching algorithm to establish an optimal matching relationship between the predicted primitive set and the real primitive set;

[0053] S4, based on the matching relationship, calculating a multi-task loss function including primitive type loss, attribute loss and geometric parameter loss;

[0054] S5, optimizing the parameters of the primitive parsing model through the multi-task loss function.

[0055] By parsing hand-drawn sketches into a set of parametric primitives, these primitives can be directly imported into computer-aided design (CAD) software to generate precise component models. Parametric primitives provide precise geometric information, including primitive types, Boolean attributes, and geometric parameters. This information can be directly used for precise modeling in CAD software, reducing human error and improving component model accuracy. Furthermore, this process is fast and accurate, avoiding the inefficiencies and inaccuracies of manual sketch parsing.

[0056] Among them, combined Figure 2 Step S1 specifically includes: defining a parameterized representation specification of a geometric primitive, including:

[0057] Primitive type identification: assign unique type codes to lines, circles, arcs, and points;

[0058] Boolean attribute flag: used to indicate additional characteristics of primitives, encoded as 0 / 1;

[0059] Geometric parameter specifications:

[0060] Line: coordinates of the starting point (x1, y1) and the end point (x2, y2);

[0061] Circle: center (x, y) and radius r;

[0062] Arc: starting point (x1, y1), middle point (xmid, ymid) and end point (x2, y2);

[0063] Point: coordinates (x,y);

[0064] Parameter quantization rule: Use a fixed number of bits to quantize geometric parameters, and fill unused parameters with -1.

[0065] Specifically, in this embodiment, a 6-bit integer is used to quantize the normalized primitive parameters, and a Boolean attribute is encoded at the same time. The Boolean flag is 1 when it is true, and 0 otherwise.

[0066] The quantization process introduces an acceptable error. As shown in Table 1, in the parameter part, the subscripts 1, mid, and 2 represent the starting point, middle point, and end point, respectively, and a value of -1 indicates placeholder padding.

[0067] Table 1 Primitive parameter quantization table

[0068]

[0069] In this embodiment, step S2 includes:

[0070] S21, dividing the input hand-drawn component sketch image into multiple non-overlapping image blocks of fixed size;

[0071] S22, extracting features from each image block using a feature extraction network to obtain multiple image block sequence feature sequences;

[0072] S23, using the Transformer encoder, fuses all image block sequence features into a global feature sequence;

[0073] S24 uses the Transformer decoder and introduces multiple query vectors to extract the primitive features corresponding to each query vector based on the global feature sequence;

[0074] S25, the primitive features corresponding to each query vector output by the Transformer decoder are sent to multiple multi-layer perceptron fully connected networks, and the primitive type, Boolean attributes and primitive parameters of each primitive feature are respectively determined to generate a set of predicted primitives that conform to the standardized parameter representation system.

[0075] For example, the input hand-drawn component sketch is uniformly scaled to a 128×128 pixel grayscale image; non-overlapping square blocks of size 16×16 are extracted from the input 128×128 resolution component sketch image to generate a picture block sequence of 64 flattened blocks; after each picture block is flattened into a 256-dimensional vector, feature extraction is performed on each picture block sequence through a three-layer feature extraction network PatchMLP (the dimension of each layer of the fully connected network is 256) to obtain multiple picture block sequence feature sequences.

[0076] Then, a linear transformation is performed on each image block sequence. This means that each element in each image block sequence is changed through a linear operation to obtain a new image block sequence. A linear transformation can usually be expressed as a matrix-vector multiplication, or more simply as a weighted sum of each element.

[0077] In this example, the Transformer encoder and Transformer decoder components each had twelve layers, and dropout was set to 0.01. The primitive parsing model was trained for 250 epochs; varying the number of epochs did not significantly improve performance. The primitive parsing model was trained using the Adam optimizer, with an initial learning rate of 1e-4 and a batch size of 512 per GPU.

[0078] Then, the Transformer encoder is used to further fuse the sequence features of all image blocks into global sequence features.

[0079] The output of the Transformer decoder is represented as decoder output features. Each feature vector output by the decoder corresponds to a query vector and a primitive target, and contains information related to the query vector extracted from the image block feature sequence.

[0080] The output of the Transformer decoder is fed into three multi-layer perceptrons (MLPs). t Prediction primitive type, MLP f Predicting Boolean attributes, MLP p Predict primitive parameters, which are composed of three layers of fully connected networks with an output dimension of 256) to predict the sketch primitive set Each prediction primitive Including primitive types Boolean properties Primitive parameters and N=20 is the upper limit of the number of primitives predicted by the primitive model.

[0081] In the embodiment of the present application, step S3 includes:

[0082] S31, calculating the matching cost between each predicted primitive in the predicted primitive set and each real primitive in the real primitive set to form a cost matrix;

[0083] S32, processing the cost matrix using the Hungarian algorithm to find an optimal matching solution with the minimum cost;

[0084] S33 , determining a matching relationship between each predicted primitive in the predicted primitive set and a corresponding real primitive in the real primitive set according to the optimal matching solution; marking unmatched predicted primitives as invalid predictions.

[0085] In order to find the corresponding relationship between the two sets, the embodiment of the present invention uses the Hungarian matching algorithm to clearly predict the primitive set. The correspondence between the real primitive set P is achieved by searching for the permutation of N elements σ∈N at the lowest cost, as described below:

[0086]

[0087] is the optimal index arrangement found by the optimization algorithm.

[0088] Among them L match is the i-th real primitive P i and the σ(i)th prediction primitive The cost function for successful matching between .

[0089] σ is a permutation that maps each true index i to another predicted index σ(i).

[0090] The indicator function It means that the primitive type meets the conditions The situation when When (i.e., the i-th primitive has a valid, non-null type), the indicator function The value is 1, otherwise it is 0. i Represents the primitive type corresponding to the i-th real primitive.

[0091] Successfully matching the cost function needs to take into account the predicted primitive type and parameters.

[0092] For the prediction of index σ(i), the present invention uses primitive categories Probability, Boolean attributes The probability and primitive parameters of The probability calculation L match , which is expressed by the following formula:

[0093]

[0094] Where l(.,.) represents the standard cross entropy function, w * is the weight of balancing the three items (w t =1,w f =1,w p =5).

[0095] In the embodiment of the present application, step S4 includes:

[0096] S41, determining the predicted primitives and the real primitives involved in the loss calculation according to the matching relationship between the predicted primitive set and the real primitive set;

[0097] S42, calculates the primitive type loss, using the cross entropy function to measure the difference between the predicted primitive type and the true primitive type;

[0098] S43, calculates attribute loss, using the cross entropy function to measure the difference between the predicted Boolean attribute and the true Boolean attribute;

[0099] S44, calculates the geometric parameter loss, using the standard cross entropy function to measure the difference between the predicted geometric parameters and the true geometric parameters;

[0100] S45 , performing weighted summation of the primitive type loss, attribute loss, and geometric parameter loss to obtain the multi-task loss function.

[0101] By using formula (2), we can get the predicted index σ(i) corresponding to the true index i. From this, we can calculate the overall primitive parsing model loss function. i =(t i ,f i ,p i ), the present invention constructs three loss functions L t ,L f and L p , respectively expressed as primitive category loss L t , Boolean attribute loss L f and primitive parameter loss L p , which is expressed by the following formula:

[0102]

[0103] Loss=w t *L t +w f *L f +w p *L p (6)

[0104] In formula (3) The objects are taken into account, starting from i=1 and accumulating to N, ensuring that all objects are calculated. In addition, in order to maintain the class balance, a weight of 10 is assigned to In formula (4) and formula (5), the present invention only relies on considering the objects matched in formula (2), that is, not Therefore, the accumulation starts from i=1 and continues to K. Loss is the weighted sum of the above losses, w * It is the weight coefficient set corresponding to different loss functions, and its value is the same as formula (2).

[0105] like Figure 4As shown in the figure, this paper shows how to obtain the implicit parameterized primitives in the hand-drawn sketch of the car gear through the query vector. In the figure, the gray curve represents the primitives that the query vector focuses on, and the black curve represents the real sketch. Each query vector focuses on a specific area in the hand-drawn sketch to observe the primitives that exist in the area. After the model is trained, each query vector has its own corresponding primitive parsing preference, which can obtain primitives in different positions, including primitive types and primitive parameters. Figure 3 As shown, taking the hand-drawn sketch as input and using the trained query vector 1, the corresponding parameterized primitive 1 in the hand-drawn sketch is obtained (as shown by the gray primitive in the figure).

[0106] The above method has the following effects:

[0107] (1) The task of parsing hand-drawn sketches of automotive parts is transformed into a set prediction problem, effectively simplifying the primitive parsing network framework in sketch parsing. Instead of designing separate modules for each primitive type, a unified network framework is used to predict the entire primitive set. This greatly simplifies the network structure, reduces the complex interactions between modules, and improves the manageability and scalability of the system.

[0108] (2) Decoupling the predicted primitive type from the corresponding primitive parameters, and independently predicting the primitive type and parameters, can more accurately parse the primitives in the hand-drawn sketch.

[0109] (3) The present invention has excellent performance and has reached the best level; at the same time, it has also achieved optimization in running speed and has also reached the best state.

[0110] Reference Figure 5 , an embodiment of the present invention further provides a parametric extraction device for hand-drawn component sketches, comprising:

[0111] A standard establishing module 201 is used to establish a standardized parameter representation system for geometric primitives;

[0112] A prediction primitive set prediction module 202 is configured to extract image features of the hand-drawn component sketch using a primitive parsing model and predict a prediction primitive set that conforms to the parameter representation system;

[0113] A matching relationship establishing module 203 is used to establish an optimal matching relationship between the predicted primitive set and the real primitive set using an optimal matching algorithm;

[0114] A multi-task loss function calculation module 204 is used to calculate a multi-task loss function including primitive type loss, attribute loss and geometric parameter loss based on the matching relationship;

[0115] The model optimization module 205 is used to optimize the parameters of the primitive parsing model through the multi-task loss function.

[0116] The present invention also provides a control device, including a processor, a memory, and a program or instruction stored in the memory and executable on the processor. When the program or instruction is executed by the processor, the steps of the parametric extraction method of the hand-drawn component sketch as described above are implemented.

[0117] The present invention also provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, the steps of the parametric extraction method of the hand-drawn component sketch as described above are implemented.

[0118] It should be noted that the various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same and similar parts between the various embodiments can be referenced to each other.

[0119] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they become aware of the basic creative concepts. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present invention.

[0120] It should also be noted that, in this document, the terms "center", "up", "down", "left", "right", "vertical", "horizontal", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the accompanying drawings, and are for the purpose of facilitating the description of the present invention and simplifying the description, rather than indicating or implying that the devices or components referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention. In addition, relational terms such as "first" and "second" are used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply that there is any actual relationship or order between these entities or operations, nor should they be understood as indicating or implying relative importance. Moreover, the terms "comprises", "includes" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device comprising a series of elements does not include those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, an element defined by the sentence "comprising a ..." does not exclude the presence of other identical elements in the process, method, article or terminal device comprising the element.

[0121] The technical solutions provided by the present invention have been described in detail above. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is intended only to facilitate understanding of the present invention, and the contents of this specification should not be construed as limiting the present invention. Furthermore, those skilled in the art will appreciate that various modifications may be made to the specific implementation methods and scope of application according to the present invention. It is not necessary and impossible to exhaustively enumerate all implementation methods herein, and any obvious variations or modifications derived therefrom remain within the scope of protection of the present invention.

Claims

1. A parametric extraction method for hand-drawn component sketches, characterized in that: include: Establish a standardized parameter representation system for geometric primitives; Extracting image features of the hand-drawn component sketch through a primitive parsing model and predicting a set of predicted primitives that conform to the parameter representation system; Using an optimal matching algorithm to establish an optimal matching relationship between the predicted primitive set and the real primitive set; Based on the matching relationship, a multi-task loss function including primitive type loss, attribute loss and geometric parameter loss is calculated; The parameters of the primitive parsing model are optimized by the multi-task loss function.

2. The parametric extraction method of hand-drawn component sketches according to claim 1, characterized in that: The steps to establish a standardized parameter representation system for geometric primitives include: Establish a standardized parameter representation system that includes type identification, attribute parameters and geometric parameters of geometric primitives.

3. The parametric extraction method of hand-drawn component sketches according to claim 1, characterized in that: The steps of extracting image features of the hand-drawn component sketch by the primitive parsing model and predicting a set of predicted primitives that conform to the parameter representation system include: Divide the input hand-drawn component sketch image into multiple non-overlapping image blocks of fixed size; Use the feature extraction network to extract features from each image block to obtain multiple image block sequence feature sequences; Use Transformer encoder to fuse all image block sequence features into a global feature sequence; Adopting the Transformer decoder and introducing multiple query vectors, the primitive features corresponding to each query vector are extracted from the global feature sequence; The primitive features corresponding to each query vector output by the Transformer decoder are fed into multiple multi-layer perceptron fully connected networks, and the primitive type, Boolean attributes and primitive parameters of each primitive feature are respectively determined to generate a set of predicted primitives that conform to the standardized parameter representation system.

4. The parametric extraction method of hand-drawn component sketches according to claim 1, characterized in that: The steps of using the optimal matching algorithm to establish the optimal matching relationship between the predicted primitive set and the real primitive set include: Calculate the matching cost between each predicted primitive in the predicted primitive set and each true primitive in the true primitive set to form a cost matrix; The cost matrix is processed using the Hungarian algorithm to find the optimal matching solution with the minimum cost; Determining, according to the optimal matching solution, a matching relationship between each predicted primitive in the predicted primitive set and a corresponding real primitive in the real primitive set; Unmatched prediction primitives are marked as invalid predictions.

5. The parametric extraction method of hand-drawn component sketches according to claim 1, characterized in that: Based on the matching relationship, the steps of calculating a multi-task loss function including primitive type loss, attribute loss, and geometric parameter loss include: According to the matching relationship between the predicted primitive set and the real primitive set, the predicted primitives and the real primitives involved in the loss calculation are determined; Calculate the primitive type loss and use the cross entropy function to measure the difference between the predicted primitive type and the true primitive type; Calculate attribute loss and use the cross entropy function to measure the difference between the predicted Boolean attribute and the true Boolean attribute; Calculate the geometric parameter loss and use the standard cross entropy function to measure the difference between the predicted geometric parameters and the true geometric parameters; The primitive type loss, attribute loss and geometric parameter loss are weighted and summed to obtain the multi-task loss function.

6. The parametric extraction method of hand-drawn component sketches according to claim 3, characterized in that: A query vector corresponds to a primitive target in the hand-drawn component sketch.

7. The parametric extraction method of hand-drawn component sketches according to claim 1, characterized in that: The geometric primitives are two-dimensional curves, including straight lines, circles and / or arcs.

8. A parametric extraction device for hand-drawn component sketches, characterized in that: include: Standard establishment module, used to establish a standardized parameter representation system for geometric primitives; A prediction primitive set prediction module is used to extract image features of the hand-drawn component sketch through a primitive parsing model and predict a prediction primitive set that conforms to the parameter representation system; A matching relationship establishment module, configured to establish an optimal matching relationship between the predicted primitive set and the real primitive set using an optimal matching algorithm; A multi-task loss function calculation module, used to calculate a multi-task loss function including primitive type loss, attribute loss and geometric parameter loss based on the matching relationship; A model optimization module is used to optimize the parameters of the primitive parsing model through the multi-task loss function.

9. A control device, characterized in that: It includes a processor, a memory, and a program or instruction stored in the memory and executable on the processor, wherein when the program or instruction is executed by the processor, the steps of the parametric extraction method of the hand-drawn component sketch as described in any one of claims 1 to 5 are implemented.

10. A readable storage medium, characterized in that: The readable storage medium stores a program or instruction, and when the program or instruction is executed by the processor, the steps of the parametric extraction method of the hand-drawn component sketch as described in any one of claims 1 to 5 are implemented.

Images