Cutting piece type recognition method and device, sewing equipment and readable storage medium

Abstract

The application provides a kind of identification method of cutting piece type, comprising: obtaining the image information of the cutting piece to be identified. According to the image information of the cutting piece to be identified, the contour of the cutting piece to be identified is identified to determine the first contour of the cutting piece to be identified. According to the first contour, the first key point of the cutting piece to be identified is determined. According to the first contour, the first key point and the preset cutting piece type template, it is determined whether the type of the cutting piece to be identified is front cutting piece or back cutting piece. In this way, the sewing equipment can determine the type of the cutting piece to be identified according to the first contour, the first key point and the preset cutting piece type template after identifying the cutting piece to be identified, thereby providing a basis for subsequent cutting piece sewing related procedures. Because contour determination and key point determination do not require complex steps, the sewing equipment can simply and efficiently complete cutting piece type identification, the identification efficiency of cutting piece type is improved, and the efficiency of garment sewing is also improved.

Description

Technical Field

[0001] This application relates to the field of intelligent manufacturing technology, and more specifically, to a method for identifying fabric piece types, a device for identifying fabric piece types, sewing equipment, and a computer-readable storage medium. Background Technology

[0002] Generally, a garment is made by sewing together several pieces of fabric. Based on the position of the pieces in the finished garment, they can be categorized into front pieces (located in front of the body) and back pieces (located in back of the body). In related technologies, the pieces are typically divided into front and back pieces manually before the sewing equipment is controlled to complete the sewing operation according to the type of piece. However, this method is inefficient due to the low efficiency of manual piece identification, resulting in low overall garment sewing efficiency. Summary of the Invention

[0003] This application provides a method for identifying fabric piece types, a device for identifying fabric piece types, a sewing device, and a computer-readable storage medium.

[0004] This application provides a method for identifying the type of cut piece, including:

[0005] Obtain image information of the cropped piece to be identified;

[0006] The first outline of the cut piece to be identified is determined based on the image information;

[0007] The first key point of the cut piece to be identified is determined based on the first contour;

[0008] Based on the first outline, the first key point, and the preset pattern type template, the type of the pattern to be identified is determined, and the type includes front pattern and back pattern.

[0009] In the method for identifying fabric piece types provided in this application, a sewing device acquires image information of the fabric piece to be identified. Then, based on the image information, contour recognition is performed on the fabric piece to be identified to determine its first contour. Multiple key points of the fabric piece to be identified are determined based on the first contour, thereby obtaining the first key point. With the first contour and the first key point obtained, the sewing device determines whether the fabric piece to be identified is a front or rear fabric piece based on the first contour, the first key point, and a preset fabric piece type template. Thus, the embodiments of this application enable the sewing device to identify the type of the fabric piece after it has been identified—that is, after detecting an unknown type of fabric piece to be identified—based on the first contour, the first key point, and the preset fabric piece type template; or, based on the first contour and the first key point, determine the similarity between the fabric piece to be identified and the preset fabric piece type template, and then identify the type of the fabric piece to be identified, thereby providing a basis for subsequent fabric piece sewing processes. In this embodiment, the sewing equipment can identify the type of the cut piece based on its first key point and first contour. Since contour and key point determination do not require complex steps, the sewing equipment can quickly determine the outline and key points of the cut piece, thus achieving simple and efficient cut piece type identification. This improves the efficiency of cut piece type identification and the overall efficiency of garment sewing. Simultaneously, it reduces the execution cost associated with manual cut piece type identification to some extent. Furthermore, the sewing equipment can perform cut piece type identification using a control chip with lower computing power, thereby reducing the hardware cost of the sewing equipment.

[0010] In some implementations, identifying the type of the cut piece to be identified based on the first outline, the first key point, and a preset cut piece type template includes:

[0011] The second key point is determined based on the second outline of the preset piece type template;

[0012] Based on the first key point and the second key point, perform an affine transformation on the first contour to obtain a transformed image of the first contour.

[0013] The type of the cut piece to be identified is determined based on the transformed image and the second contour.

[0014] Thus, the embodiments of this application are based on affine transformation, so that the first contour of the cut piece to be identified can be the same in size as the second contour of the preset cut piece type template after transformation. Therefore, the sewing equipment can more easily determine the similarities and differences between the transformed first contour and the second contour. That is, the similarities and differences between the transformed image and the second contour can make the type of the cut piece to be identified accurately identified.

[0015] In some implementations, identifying the type of the cut piece to be identified based on the transformed image and the second contour includes:

[0016] Based on the transformed image and the second contour, the similarity between the cut piece to be identified and the preset cut piece type template is determined;

[0017] Based on the degree of similarity, the type of the cut piece to be identified is determined.

[0018] Thus, the embodiments of this application enable the sewing equipment to determine the type of the cut piece to be identified based on the similarity between the transformed image and the second contour, making the identification of the cut piece type simple and efficient, and ensuring the identification efficiency.

[0019] In some implementations, determining the similarity between the cut piece to be identified and the preset cut piece type template based on the transformed image and the second contour includes:

[0020] Based on the first transformed image and the outline of the front piece, the first similarity between the piece to be identified and the front piece template is determined. The first transformed image is obtained by performing an affine transformation on the first outline based on the first key point and the third key point corresponding to the outline of the front piece.

[0021] Based on the second transformed image and the outline of the back piece, the second similarity between the piece to be identified and the template of the back piece is determined. The second transformed image is obtained by performing an affine transformation on the first outline based on the first key point and the fourth key point corresponding to the outline of the back piece.

[0022] Thus, the embodiments of this application, based on the first similarity degree and the second similarity degree, enable the sewing equipment to efficiently determine the similarities and differences between the cut piece to be identified and the previous cut piece, as well as the similarities and differences between the cut piece to be identified and the subsequent cut piece, thereby accurately determining the type of the cut piece to be identified.

[0023] In some implementations, identifying the type of the cut piece to be identified based on the degree of similarity includes:

[0024] If the first similarity level is higher than the second similarity level, then the type of the cut piece to be identified is determined to be a front cut piece;

[0025] If the second similarity level is higher than the first similarity level, then the type of the cut piece to be identified is determined to be a post-cut piece.

[0026] Thus, the implementation method of this application, based on the relationship between the first similarity and the second similarity, enables the type of the cut piece to be identified to be determined efficiently and accurately.

[0027] In some implementations, determining the similarity between the cut piece to be identified and the preset cut piece type template based on the transformed image and the second contour includes:

[0028] The degree of similarity is determined based on the intersection-union ratio of the transformed image and the second contour.

[0029] Thus, the implementation method of this application is based on intersection-union ratio, which makes it easy and efficient to determine the similarity between the transformed image and the second contour, and also ensures the execution efficiency of the clipping type.

[0030] In some implementations, determining the similarity based on the intersection-union ratio (IUU) of the transformed image and the second contour includes:

[0031] Based on the second contour, the transformed image is subjected to image alignment processing to obtain the aligned transformed image;

[0032] The degree of similarity is determined based on the intersection-union ratio of the second contour and the aligned transformed image.

[0033] Thus, the embodiment of this application is based on image alignment processing, which enables the transformed image to be aligned with the second contour, thereby reducing the negative impact of the positional difference between the transformed image and the second contour on the cross-union ratio calculation, and ensuring the effectiveness of the cross-union ratio calculation result.

[0034] This application provides a device for identifying the type of cut piece, including:

[0035] The information acquisition module is used to acquire image information of the cropped piece to be identified;

[0036] A contour determination module is used to determine the first contour of the cut piece to be identified based on the image information;

[0037] The key point determination module is used to determine the first key point of the cut piece to be identified based on the first contour.

[0038] The type recognition module is used to identify the type of the cut piece to be identified based on the first outline, the first key point and the preset cut piece type template, wherein the type includes front cut piece and back cut piece.

[0039] This application provides a sewing device, including a memory and a processor. The memory stores a computer program, which, when executed by the processor, implements the above-described method for identifying fabric piece types.

[0040] This application provides a computer-readable storage medium storing a computer program that, when executed by one or more processors, implements the above-described method for identifying fabric piece types.

[0041] The fabric piece type identification device, sewing equipment, and computer-readable storage medium provided in this application can identify the type of fabric piece after it is detected (i.e., after detecting an unknown type of fabric piece to be identified), based on the first outline, first key points, and a preset fabric piece type template of the fabric piece to be identified; or, based on the first outline and first key points, determine the similarity between the fabric piece to be identified and the preset fabric piece type template, and then identify the type of the fabric piece to be identified, thereby providing a basis for subsequent fabric piece sewing processes. In this application embodiment, the sewing equipment can complete the fabric piece type identification based on the first key points and first outline of the fabric piece to be identified. Since the outline determination and key point determination do not require complex steps, the sewing equipment can quickly determine the outline and key points of the fabric piece, thereby completing the fabric piece type identification simply and efficiently, improving the efficiency of fabric piece type identification and the efficiency of garment sewing by the sewing equipment. At the same time, it can reduce the execution cost caused by manual fabric piece type identification to a certain extent. In addition, the sewing equipment can complete the fabric piece type identification with a control chip with low computing power, thereby reducing the hardware cost of the sewing equipment.

[0042] Additional aspects and advantages of embodiments of this application will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of embodiments of this application. Attached Figure Description

[0043] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, wherein:

[0044] Figure 1 This is a flowchart illustrating the method for identifying the type of cut piece in certain embodiments of this application;

[0045] Figure 2 This is a schematic diagram of a fabric piece type identification device in some embodiments of this application;

[0046] Figure 3 This is a schematic diagram of the cut piece in some embodiments of this application;

[0047] Figure 4 This is a flowchart illustrating the method for identifying the type of cut piece in certain embodiments of this application;

[0048] Figure 5 This is a flowchart illustrating the method for identifying the type of cut piece in certain embodiments of this application;

[0049] Figure 6 This is a flowchart illustrating the method for identifying the type of cut piece in certain embodiments of this application;

[0050] Figure 7This is a flowchart illustrating the method for identifying the type of cut piece in certain embodiments of this application;

[0051] Figure 8 This is a flowchart illustrating the method for identifying the type of cut piece in certain embodiments of this application;

[0052] Figure 9 This is a flowchart illustrating the method for identifying the type of cut piece in certain embodiments of this application;

[0053] Figure 10 This is a flowchart illustrating a method for identifying fabric piece types in certain embodiments of this application. Detailed Implementation

[0054] The embodiments of this application are described in detail below. Examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the embodiments of this application, and should not be construed as limiting the embodiments of this application.

[0055] In automated garment production, sewing equipment needs to correctly stack different types of cut pieces before sewing. For example, the front cut piece is stacked on top of the back cut piece before they are sewn together to obtain a garment. Therefore, cut piece type identification is an indispensable part of automated garment production.

[0056] Currently, in the fabric piece identification process, some factories use manual identification of fabric piece types to ensure accurate identification. However, because fabric piece identification is mostly a tedious and repetitive task, and prolonged identification work causes eye strain, the efficiency of manual identification of fabric piece types is low, which in turn limits the sewing efficiency of sewing equipment.

[0057] In other factories, a deep learning-based automated fabric piece type identification method is used. This method predicts the type of each fabric piece based on a pre-trained neural network model, thus automating the identification process. However, due to the uncertainty of neural network models—or rather, because they are black-box models—it is difficult for workers to correct errors in fabric piece type prediction to prevent future mistakes. Furthermore, most neural network models process type prediction tasks based on pixel features of fabric piece images, such as using color features. However, when the neural network model identifies unknown fabric pieces, or detects fabric pieces outside the training set, the style, texture, and print of the fabric pieces are unknown, so the neural network model may output unpredictable results. Therefore, the prediction accuracy of neural network models is difficult to guarantee.

[0058] Furthermore, when the sewing equipment identifies the type of fabric piece based on the neural network model, it is necessary to ensure that the control chip of the sewing equipment can support the operation of the neural network model. In other words, the control chip needs to have high computing power, which leads to a high hardware cost for a single sewing equipment.

[0059] For the issues mentioned above, please refer to [link / reference]. Figure 1 This application provides a method for identifying the type of cut piece, including:

[0060] 01: Obtain the image information of the cropped piece to be identified;

[0061] 02: Determine the first outline of the cut piece to be identified based on the image information;

[0062] 03: Determine the first key point of the cut piece to be identified based on the first outline;

[0063] 04: Based on the first outline, the first key point, and the preset pattern type template, identify the type of the pattern to be identified, which includes front pattern and back pattern.

[0064] Please see Figure 2 This application provides a fabric piece type identification device 200. The fabric piece type identification method of this application can be implemented by the fabric piece type identification device 200. Specifically, the identification device 200 includes an information acquisition module 210, a contour determination module 220, a key point determination module 230, and a type identification module 240. The information acquisition module 210 is used to acquire image information of the fabric piece to be identified. The contour determination module 220 is used to determine a first contour of the fabric piece to be identified based on the image information. The key point determination module 230 is used to determine a first key point of the fabric piece to be identified based on the first contour. The type identification module 240 is used to identify the type of the fabric piece to be identified based on the first contour, the first key point, and a preset fabric piece type template.

[0065] This application also provides a sewing device, which includes a memory and a processor. The method for identifying fabric piece types according to this application can be implemented by the sewing device of this application. Specifically, the memory stores a computer program, and the processor is used to acquire image information of the fabric piece to be identified; determine a first outline of the fabric piece to be identified based on the image information; determine a first key point of the fabric piece to be identified based on the first outline; and identify the type of the fabric piece to be identified based on the first outline, the first key point, and a preset fabric piece type template. The type includes front fabric piece and back fabric piece.

[0066] Specifically, in this application's embodiment, when the sewing recognition detects a piece of unknown type, for example, when a camera detects a piece of fabric at a preset location, or when a piece of fabric is received, the sewing device can capture an image of the piece of fabric to be identified using the camera, i.e., acquire the image information of the piece of fabric to be identified. Then, based on the image information of the piece of fabric to be identified, the sewing device identifies the outline of the piece of fabric to be identified, thereby extracting the outline of the piece of fabric to be identified, i.e., the first outline. With the first outline obtained, a key point extraction operation is performed on the first outline to identify various key points on the first outline, such as identifying the neckline key point and / or cuff key point in the first outline, thereby obtaining the first key point. With the first outline and the first key point information obtained, the sewing device determines the type of the piece of fabric to be identified based on the first outline, the first key point, and a pre-stored preset piece of fabric type template, i.e., determines the piece of fabric to be identified as a front piece of fabric or a back piece of fabric.

[0067] It is understood that the sewing equipment in the embodiments of this application can be understood as a machine or equipment capable of performing the tasks of identifying the type of fabric piece and sewing the fabric piece.

[0068] Furthermore, the specific configuration of the sewing equipment can be configured according to actual conditions. For example, in some embodiments, the sewing equipment may consist of two hardware devices: a host computer and an intelligent sewing machine. The host computer can be understood as the device used to perform fabric piece type identification, i.e., to execute steps 01 to 04 above. The host computer includes, but is not limited to, mobile terminals, personal computers, and tablet computers. The intelligent sewing machine can be used to perform corresponding fabric piece grabbing and sewing based on signals sent by the host computer. In other embodiments, the sewing equipment consists of a single, complete hardware device capable of performing fabric piece type identification, i.e., executing steps 01 to 04 above, based on various modules (such as camera modules) and a preset program.

[0069] It is also understood that, after acquiring the image information of the cut piece to be identified, the sewing device in this application's embodiments first determines the outline of the cut piece, that is, determines the first outline of the cut piece to be identified. For a clearer explanation of the outline in this application's embodiments, please refer to... Figure 3 , Figure 3 This is a schematic diagram of the cut piece in some embodiments of this application. Figure 3 In the image, the closed shape enclosed by the black lines can be seen as the outline of a certain piece of fabric.

[0070] It is understandable that the first contour can characterize image features such as the shape and size of the piece to be identified. Therefore, the embodiments of this application are based on the idea that different types of pieces have different shapes (or contours), and identify (or extract) the contour of the piece to be identified, and then determine the type of the piece to be identified based on the extracted first contour.

[0071] Furthermore, the specific method for determining the first contour based on the image information can be set according to the actual situation. For example, in some embodiments, the first contour can be obtained by processing the image information of the cut piece to be identified using an edge detection algorithm. In other embodiments, the sewing device program can execute a preset foreground segmentation algorithm, so after obtaining the image information of the cut piece to be identified, the sewing device obtains the first contour corresponding to the image information based on the foreground segmentation algorithm.

[0072] After extracting the first contour, the embodiments of this application will perform key point extraction (or key point recognition operation) based on the first contour to extract each key point of the cut piece to be identified, thereby obtaining the first key point. It is understood that since the first key point is obtained based on the first contour, the first key point can characterize the size and contour of the cut piece to be identified.

[0073] Furthermore, the first key point in the embodiments of this application is a content that can be set according to the actual situation. For example, in some embodiments, there are multiple first key points, including but not limited to the cuff key point and the collar key point. In other embodiments, please refer again for details. Figure 3 In other words, the first key point could be Figure 3 The eight points, or rather, Figure 3 The location and key point categories of the eight hollow circles (hem category key points, neckline category key points, etc.).

[0074] Furthermore, the specific method for determining the first key point of the cut piece to be identified can be set according to the actual situation. For example, in some implementations, the first key point is obtained based on a preset key point recognition model. That is, the first outline of the cut piece to be identified is input into the preset key point recognition model to obtain the first key point.

[0075] After obtaining the first outline and the first key points, the sewing machine can determine the specific type of the piece to be identified based on a pre-stored preset piece type template, combined with the first outline and the first key points. It is understood that this embodiment of the application can complete the type identification of the piece to be identified using only the first outline and the first key points of the piece, as well as the preset piece type template. In other words, this embodiment of the application completes type identification using partial information from a complete piece image, without needing to perform operations such as pixel comparison on the complete piece image. The method for identifying the piece type is relatively simple. Furthermore, it is understood that determining the outline and key points of the piece does not require complex algorithms or models. Therefore, when the sewing machine processes the piece type identification task, it can complete the identification based on a control chip with lower computing power, thus controlling the hardware cost of the sewing machine.

[0076] It is understood that the preset pattern type template in the embodiments of this application can be understood as the general style of a certain type of pattern. For example, after image fitting of multiple front patterns, the fitted image is the preset pattern type template.

[0077] It is also understandable that the specific method for identifying the type of the cut piece to be identified based on the first outline, the first key point, and the preset cut piece type template is customizable according to actual circumstances. For example, in some embodiments, the sewing equipment can determine the type of the cut piece to be identified based on the degree of overlap between the first outline and the preset cut piece type template, and the key point placement within the preset cut piece type template. Exemplarily, suppose the preset cut piece type template includes a front cut piece template. When the cut piece to be identified and the front cut piece template are of the same proportion and aligned, it is determined whether the overlapping area between the first outline of the cut piece to be identified and the front cut piece template exceeds a preset area. Then, it is determined whether a preset number or more key points are located within the corresponding graphic of the front cut piece template. If both conditions are met, the cut piece to be identified is considered similar to the front cut piece template, and therefore identified as the front cut piece. If only one condition is met, or neither condition is met, the cut piece to be identified is identified as the rear cut piece.

[0078] Furthermore, it is understandable that after the type of the cut piece is identified, the sewing equipment can perform subsequent sewing processes, such as cut piece picking and cut piece sewing, according to the type of cut piece.

[0079] In summary, the embodiments of this application enable the sewing equipment to identify the type of a cut piece after it has been detected (i.e., after detecting an unknown type of cut piece to be identified), based on the first outline, first key points, and a preset cut piece type template of the cut piece. Alternatively, it can determine the similarity between the cut piece to be identified and the preset cut piece type template based on the first outline and first key points, and then identify the type of the cut piece, thus providing a basis for subsequent sewing processes. In the embodiments of this application, the sewing equipment can complete the identification of the cut piece type based on the first key points and first outline of the cut piece to be identified. Since the outline determination and key point determination do not require complex steps, the sewing equipment can quickly determine the outline and key points of the cut piece, thereby completing the cut piece type identification simply and efficiently. The efficiency of cut piece type identification is improved, and the efficiency of garment sewing by the sewing equipment is also improved. At the same time, it can reduce the execution cost caused by manual identification of cut piece type to a certain extent. In addition, the sewing equipment can complete the cut piece type identification with a control chip with lower computing power, reducing the hardware cost of the sewing equipment.

[0080] Furthermore, since this embodiment does not employ a black-box model (such as a neural network model) to predict the type of the fabric piece, when the type of the fabric piece to be identified is incorrectly identified, the operator can quickly pinpoint the cause of the sewing equipment error, making subsequent maintenance of the sewing equipment more convenient. Simultaneously, because this embodiment performs type identification based on contours and key points, rather than image features such as color features, when the sewing equipment encounters fabric pieces with unknown color, texture, and print, since color, texture, and print do not affect the contours and key points of the fabric piece, the sewing equipment of this embodiment can complete the identification with a high accuracy rate.

[0081] In some implementations, please refer to Figure 4 Step 04 includes:

[0082] 040: Determine the second key point based on the second outline of the preset pattern template;

[0083] 041: Based on the first key point and the second key point, perform an affine transformation on the first contour to obtain the transformed image of the first contour;

[0084] 042: Identify the type of the cut piece to be identified based on the transformed image and the second contour.

[0085] The type identification module in this embodiment is used to determine a second key point based on the second contour of a preset pattern piece type template; perform affine transformation on the first contour based on the first and second key points to obtain a transformed image of the first contour; and identify the type of the pattern piece to be identified based on the transformed image and the second contour.

[0086] The sewing device of this application embodiment is further used to determine a second key point based on the second contour of a preset cut piece type template; perform affine transformation processing on the first contour based on the first key point and the second key point to obtain a transformed image of the first contour; and identify the type of cut piece to be identified based on the transformed image and the second contour.

[0087] That is, the sewing equipment in this application will determine the type of the cut piece to be identified based on the similarity in outline between the cut piece to be identified and the preset cut piece type template.

[0088] Specifically, the sewing equipment first extracts the outline of a preset pattern template to obtain a second outline.

[0089] It is understandable that the specific method for extracting the outline of the preset pattern piece type template can be set according to the actual situation. For example, in some embodiments, the preset pattern piece type template stored in the sewing device is a Drawing Exchange Format (DXF) file. In other words, the preset pattern piece type template is constructed from multiple vector data. Therefore, the sewing device can obtain a second outline by extracting the vector data pre-labeled as "outline" in the preset pattern piece type template.

[0090] After obtaining the second outline, the sewing equipment identifies the key points in the second outline to obtain the second key points of the preset pattern type template.

[0091] Subsequently, to ensure that the first and second contours have the same proportions or dimensions, and to eliminate the size difference (or proportion difference) between the first and second contours, this embodiment of the application uses an affine transformation method. Based on the first key points of the first contour and the second key points of the second contour, a corresponding transformation process is performed on the first contour so that after the transformation, the key points of the first contour can match the key points of the second contour in position. It is understood that the first and second key points can be key points of the same type. Exemplarily, the first key point includes key points similar to... Figure 3 When the eight key points are shown, the second key point also includes something similar to... Figure 3 The eight key points are shown below.

[0092] Furthermore, during affine transformation, the mapping transformation is performed based on the type of keypoints. For example, when both the first contour and the second contour include collar keypoints and hem keypoints, the collar keypoints of the first contour will be mapped to the collar keypoints of the second contour, and the hem keypoints of the first contour will be mapped to the hem keypoints of the second contour, thereby completing the affine transformation.

[0093] It is also understandable that while the proportions of the line segments in the first contour change before and after the affine transformation, the direction of each line segment remains unchanged. That is, a line segment that was a straight line before the transformation remains a straight line after the transformation, and a line segment that was a curve before the transformation remains a curve after the transformation. Therefore, based on the affine transformation, the distortion of the first contour before and after the transformation is likely to be small, ensuring that the shape of the first contour remains unchanged to a certain extent. Furthermore, it is also understandable that the affine transformation operates relatively simply; in other words, the sewing equipment can perform affine transformations with high efficiency, thereby ensuring the efficiency of the sewing equipment in performing complete pattern type recognition.

[0094] After completing the affine transformation of the first contour and obtaining the affine transformation result of the first contour, i.e. the transformed image, the sewing device will identify the type of the cut piece to be identified based on the transformed image and the second contour of the preset cut piece type template.

[0095] It is understandable that if the type of the piece to be identified matches the type of the preset piece type template, for example, if the actual type of the piece to be identified is a front piece and the preset piece type template is a front piece template, then: because the types are the same, the relative positions of each key point in the first contour are similar to the relative positions of each key point in the second contour. After the first contour and the second contour are scaled to the same ratio, the first contour and the second contour are similar, that is, the transformed image and the second contour are similar.

[0096] It is also understandable that if the type of the piece to be identified does not match the type of the preset piece type template, such as the actual type of the piece to be identified being a back piece while the preset piece type template is a front piece template, the first contour will be distorted to a certain extent during the affine transformation process due to the type mismatch, and the transformed image will have a certain difference from the second contour.

[0097] Furthermore, it should be noted that because the cut pieces are made of flexible materials, they may undergo certain deformations, such as stretching, in actual production environments. Therefore, if the type is determined directly based on a complete cut piece image, for example, by calculating the similarity between the complete cut piece image and a preset cut piece type template, the cut piece may exhibit some deformation, leading to differences between the cut piece and the standard template (i.e., the preset cut piece type template), resulting in incorrect cut piece type identification.

[0098] Therefore, in order to avoid the above situation, the implementation method of this application is based on key point mapping, that is, affine mapping based on key points, so that the outline of the deformed piece will be based on affine mapping, which to a certain extent makes up for the error caused by deformation, thereby reducing the difference between the deformed piece and the standard template (i.e., the preset piece type template), and ensuring the accuracy of piece type recognition.

[0099] Thus, the embodiments of this application are based on affine transformation, so that the first contour of the cut piece to be identified can have the same scale (or proportion) as the second contour of the preset cut piece type template after transformation. Therefore, the sewing equipment can more easily determine the similarities and differences between the transformed first contour and the second contour. That is, the similarities and differences between the transformed image and the second contour can make the type of the cut piece to be identified accurately identified.

[0100] In some implementations, please refer to Figure 5 Step 042 includes:

[0101] 0420: Based on the transformed image and the second contour, determine the degree of similarity between the piece to be identified and the preset piece type template;

[0102] 0421: Identify the type of the cut piece to be identified based on the degree of similarity.

[0103] The type recognition module in this embodiment is used to determine the degree of similarity between the cut piece to be recognized and the preset cut piece type template based on the transformed image and the second contour; and to identify the type of the cut piece to be recognized based on the degree of similarity.

[0104] The processor in this embodiment is further configured to determine the degree of similarity between the cut piece to be identified and a preset cut piece type template based on the transformed image and the second contour; and to identify the type of the cut piece to be identified based on the degree of similarity.

[0105] That is, the sewing device of this application embodiment will determine the type of the cut piece to be identified based on the similarity between the transformed image and the second contour.

[0106] As an example, let's assume a preset crop type template corresponds to a previous crop, or in other words, the preset crop type template is the previous crop template. Then, we determine the similarity between the transformed image and the second contour of the previous crop template. If the similarity is higher than a preset value, it means the first contour of the crop to be identified is similar to the second contour of the previous crop template, so the type of the crop to be identified is the previous crop. If the similarity is not higher than the preset value, it means the first contour of the crop to be identified is not similar to the second contour of the previous crop template, so the type of the crop to be identified is the subsequent crop.

[0107] It is understandable that the specific method for determining the degree of similarity can be set according to the actual situation. For example, in some implementations, the degree of similarity is the area of ​​intersection of the transformed image and the second contour, or the area of ​​overlap.

[0108] Thus, the embodiments of this application enable the sewing equipment to determine the type of the cut piece to be identified based on the similarity between the transformed image and the second contour, making the identification of the cut piece type simple and efficient, and ensuring the identification efficiency.

[0109] In some implementations, please refer to Figure 6Step 0420 includes:

[0110] 04200: Based on the first transformed image and the outline of the previous piece, determine the first similarity between the piece to be identified and the template of the previous piece. The first transformed image is obtained by performing an affine transformation on the first outline based on the first key point and the third key point corresponding to the outline of the previous piece.

[0111] 04201: Based on the second transformed image and the outline of the back piece, determine the second similarity between the piece to be identified and the template of the back piece. The second transformed image is obtained by performing an affine transformation on the first outline based on the first key point and the fourth key point corresponding to the outline of the back piece.

[0112] The type determination module in this application embodiment is further configured to determine a first similarity between the cut piece to be identified and the front cut piece template based on the first transformed image and the front cut piece outline, wherein the first transformed image is obtained by performing an affine transformation on the first outline based on the first key point and the third key point corresponding to the front cut piece outline; and to determine a second similarity between the cut piece to be identified and the rear cut piece template based on the second transformed image and the rear cut piece outline, wherein the second transformed image is obtained by performing an affine transformation on the first outline based on the first key point and the fourth key point corresponding to the rear cut piece outline.

[0113] The processor in this embodiment is further configured to determine a first similarity between the piece to be identified and the front piece template based on the first transformed image and the front piece outline, wherein the first transformed image is obtained by performing an affine transformation on the first outline based on a first key point and a third key point corresponding to the front piece outline; and to determine a second similarity between the piece to be identified and the rear piece template based on the second transformed image and the rear piece outline, wherein the second transformed image is obtained by performing an affine transformation on the first outline based on a first key point and a fourth key point corresponding to the rear piece outline.

[0114] That is, the preset pattern type template of the present application embodiment may include two types, namely a front pattern template and a back pattern template. Furthermore, when identifying the second outline and the second key point of the preset pattern type template, the front pattern outline and the third key point of the front pattern outline of the front pattern template will be identified, and the back pattern outline and the fourth key point of the back pattern template will be identified. In other words, the second outline of the present application embodiment includes the front pattern outline and the back pattern outline, and the second key point includes the third key point and the fourth key point.

[0115] Meanwhile, the transformed image is also divided into two types: the first transformed image obtained by performing an affine transformation on the first contour based on the first key point and the third key point, and the second transformed image obtained by performing an affine transformation on the first contour based on the first key point and the fourth key point.

[0116] Furthermore, the degree of similarity in the embodiments of this application includes a first degree of similarity and a second degree of similarity. The first degree of similarity refers to the similarity between the first transformed image and the outline of the front piece, and the second degree of similarity refers to the similarity between the second transformed image and the outline of the rear piece.

[0117] Therefore, the embodiments of this application will determine whether the cut piece to be identified is a front cut piece or a back cut piece based on a first similarity level and a second similarity level. It is understood that the specific method for determining the type of the cut piece to be identified based on the first and second similarity levels can be set according to actual circumstances. Exemplarily, in some embodiments, when the quotient (S1 / S2) of the first similarity level (S1) and the second similarity level (S2) is higher than a preset value (e.g., 1), the type of the cut piece to be identified is a front cut piece. If it is lower than the preset value, the type of the cut piece to be identified is determined to be a back cut piece.

[0118] In another implementation, when either the first similarity level or the second similarity level exceeds a preset value, such as when the first similarity level exceeds the preset value, the type of the cut piece to be identified is determined to be a front cut piece. If both the first and second similarity levels exceed the preset values, or neither exceeds the preset values, then an error has occurred in the cut piece identification process (e.g., the cut piece is folded or incomplete, or the sewing equipment's program settings are incorrect), resulting in the cut piece to be identified being similar to both the front and rear cut pieces, or causing the cut piece to be identified to be dissimilar to either the front or rear cut pieces. Optionally, in the event of an identification error, the sewing equipment will issue an alarm message to remind the staff to troubleshoot the error, or the sewing equipment will send preset information to the backend system, allowing the staff to know that the sewing equipment has identified the cut piece incorrectly based on the content displayed on the backend system, thereby detecting whether the cut piece is abnormal, or checking and optimizing the sewing equipment's program.

[0119] In some implementations, when the sewing equipment cannot identify the type of the cut piece, or when the first similarity level is the same as the second similarity level, causing the sewing equipment to be unable to determine the type of the cut piece to be identified, the sewing equipment will identify the current cut piece to be identified as a waste cut piece. Then, the corresponding gripping device can be controlled to grip the current cut piece to be identified and place it at a preset waste collection point, and the next cut piece to be identified will be identified. In this way, the sewing equipment can perform corresponding operations for both identifiable and unidentifiable cut pieces, ensuring stable operation of the sewing equipment.

[0120] Furthermore, if the sewing equipment repeatedly identifies different pieces of fabric as waste, it can send preset information to the backend system. This allows staff to know that the sewing equipment is continuously misidentifying fabric pieces, thus detecting whether the fabric pieces are abnormal or testing and optimizing the sewing equipment's program.

[0121] Thus, the embodiments of this application, based on the first similarity degree and the second similarity degree, enable the sewing equipment to efficiently determine the similarities and differences between the cut piece to be identified and the previous cut piece, as well as the similarities and differences between the cut piece to be identified and the subsequent cut piece, thereby accurately determining the type of the cut piece to be identified.

[0122] In some implementations, please refer to Figure 7 Step 0421 includes:

[0123] 04210: If the first similarity level is higher than the second similarity level, then the type of the cut piece to be identified is determined to be a front cut piece;

[0124] 04211: If the second similarity level is higher than the first similarity level, then the type of the cut piece to be identified is determined to be a post-cut piece.

[0125] The type identification module in this application embodiment is further configured to determine the type of the cut piece to be identified as a front cut piece if the first similarity is higher than the second similarity; and to determine the type of the cut piece to be identified as a back cut piece if the second similarity is higher than the first similarity.

[0126] The processor in this application embodiment is further configured to determine the type of the cut piece to be identified as a front cut piece if the first similarity is higher than the second similarity; and to determine the type of the cut piece to be identified as a back cut piece if the second similarity is higher than the first similarity.

[0127] That is, the implementation method of this application will determine the template that is more similar to the cut piece to be identified among the two cut piece templates based on the relationship between the first similarity degree and the second similarity degree.

[0128] Specifically, if the first similarity level is greater than the second similarity level, it means that the cut piece to be identified is more similar to the previous cut piece, so the cut piece to be identified will be identified as the previous cut piece. If the second similarity level is greater than the first similarity level, it means that the cut piece to be identified is more similar to the subsequent cut piece, so the cut piece to be identified will be identified as the subsequent cut piece.

[0129] Thus, the implementation method of this application, based on the relationship between the first similarity and the second similarity, enables the type of the cut piece to be identified to be determined efficiently and accurately.

[0130] In some implementations, please refer to Figure 9 Step 0420 includes:

[0131] 04202: Determine the degree of similarity based on the intersection-union ratio of the transformed image and the second contour.

[0132] The implementation type determination module of this application is also used to determine the degree of similarity based on the intersection-union ratio of the transformed image and the second contour.

[0133] The processor in this embodiment is further configured to determine the degree of similarity based on the intersection-union ratio of the transformed image and the second contour.

[0134] In other words, the embodiments of this application determine the similarity between the transformed image and the second contour based on the intersection and merging of the transformed image and the second contour, i.e., the Intersection Over Union (IOU). It can be understood that the IOU can be interpreted as the ratio of the overlapping area to the merged area of ​​two images; a larger ratio indicates that the two images may have a high degree of similarity in contour. Conversely, a smaller ratio indicates that the two images have significant differences.

[0135] Thus, the implementation method of this application is based on intersection-union ratio, which makes it easy and efficient to determine the similarity between the transformed image and the second contour, and also ensures the execution efficiency of the clipping type.

[0136] In some implementations, please refer to Figure 10 Step 04202 includes:

[0137] 042020: Based on the second contour, perform image alignment processing on the transformed image to obtain the aligned transformed image;

[0138] 042021: Determine the degree of similarity based on the intersection-union ratio of the second contour and the aligned transformed image.

[0139] The type determination module in this application embodiment is further configured to perform image alignment processing on the transformed image according to the second contour to obtain an aligned transformed image; and determine the degree of similarity according to the intersection-union ratio of the second contour and the aligned transformed image.

[0140] The processor in this embodiment is further configured to perform image alignment processing on the transformed image according to the second contour to obtain an aligned transformed image; and determine the degree of similarity according to the intersection-union ratio of the second contour and the aligned transformed image.

[0141] That is, since there may be a certain difference between the position of the transformed image and the position of the second contour, and the difference in position will affect the calculation result of the cross-union ratio, in order to reduce the interference of cross-union ratio calculation caused by different positions, the sewing device of this application embodiment will also perform image alignment processing on the transformed image based on the second contour, so that the processed transformed image can be aligned with the second contour, thereby reducing the interference of cross-union ratio calculation to a certain extent.

[0142] In some implementations, the sewing device performs image alignment processing on the transformed image based on the center point of the second contour, such that the center point of the aligned transformed image coincides with the center point of the second contour.

[0143] Thus, the embodiment of this application is based on image alignment processing, which enables the transformed image to be aligned with the second contour, thereby reducing the negative impact of the positional difference between the transformed image and the second contour on the cross-union ratio calculation, and ensuring the effectiveness of the cross-union ratio calculation result.

[0144] In some implementations, please refer to Figure 10 The method for identifying the type of cut piece includes: capturing an image of the cut piece to be identified by the sewing equipment camera to obtain image information (refer to step 01 above). Extracting the outline of the cut piece to be identified based on the image information to obtain the first outline, and then performing key point extraction based on the first outline to obtain the first key points of the cut piece to be identified (refer to steps 02-03 above).

[0145] The sewing equipment can also acquire the front cut piece template and the back cut piece template, and extract the outline of the front cut piece template to obtain the front cut piece outline, and extract the outline of the back cut piece template to obtain the back cut piece outline. Then, the key points are extracted based on the front cut piece outline to obtain the third key point, and the key points are extracted based on the back cut piece outline to obtain the fourth key point (refer to step 040 above).

[0146] Subsequently, the sewing device performs an affine transformation of the first contour based on the first key point and the third key point to obtain a first transformed image, and performs an affine transformation of the first contour based on the first key point and the fourth key point to obtain a second transformed image (refer to step 041 above).

[0147] Subsequently, the sewing equipment uses the front cut piece outline to perform image alignment processing on the first transformed image, thereby obtaining the aligned first transformed image. At the same time, according to the rear cut piece outline, the second transformed image is performed to perform image alignment processing to obtain the aligned second transformed image (refer to steps 042020-042021 above).

[0148] Subsequently, based on the aligned first transformed image and the front piece outline, the intersection-union ratio (IUGR) is calculated to obtain the first similarity. At the same time, based on the aligned second transformed image and the rear piece outline, the IUGR is calculated to obtain the second similarity (refer to steps 04200-04201 above).

[0149] Finally, determine whether the first similarity level (i.e., the crossover ratio corresponding to the front piece) is higher than the second similarity level (i.e., the crossover ratio corresponding to the back piece). If so, the type of the piece to be identified is the front piece; otherwise, it is the back piece (refer to steps 04210-04211 above).

[0150] It needs to be understood that, Figure 10 The method for identifying the type of cut piece shown is only one of the feasible embodiments of this application.

[0151] This application provides a computer-readable storage medium containing a computer program. When the computer program is executed by one or more processors, it causes the one or more processors to perform the pattern piece type identification method of this application.

[0152] In the description of this specification, the references to terms such as "some embodiments," "in one example," "exemplarily," etc., indicate that a specific feature, structure, material, or characteristic described in connection with an embodiment or example is included in at least one embodiment or example of this application. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0153] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process, and the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the function involved, as will be understood by those skilled in the art to which embodiments of this application pertain.

[0154] Although embodiments of this application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting this application. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of this application.

Claims

1. A method for identifying the type of cut piece, characterized in that, include: Obtain image information of the cropped piece to be identified; The first outline of the cut piece to be identified is determined based on the image information; The first key point of the cut piece to be identified is determined based on the first contour; Determine the second key point based on the second outline of the preset pattern template; Based on the first key point and the second key point, perform an affine transformation on the first contour to obtain a transformed image of the first contour. Based on the transformed image and the second contour, the type of the cut piece to be identified is determined, and the type includes front cut piece and back cut piece.

2. The method according to claim 1, characterized in that, The step of identifying the type of the cut piece to be identified based on the transformed image and the second contour includes: Based on the transformed image and the second contour, the similarity between the cut piece to be identified and the preset cut piece type template is determined; Based on the degree of similarity, the type of the cut piece to be identified is determined.

3. The method according to claim 2, characterized in that, The step of determining the similarity between the cut piece to be identified and the preset cut piece type template based on the transformed image and the second contour includes: Based on the first transformed image and the outline of the front piece, the first similarity between the piece to be identified and the front piece template is determined. The first transformed image is obtained by performing an affine transformation on the first outline based on the first key point and the third key point corresponding to the outline of the front piece. Based on the second transformed image and the outline of the back piece, the second similarity between the piece to be identified and the template of the back piece is determined. The second transformed image is obtained by performing an affine transformation on the first outline based on the first key point and the fourth key point corresponding to the outline of the back piece.

4. The method according to claim 3, characterized in that, The step of identifying the type of the cut piece to be identified based on the similarity includes: If the first similarity level is higher than the second similarity level, then the type of the cut piece to be identified is determined to be a front cut piece; If the second similarity level is higher than the first similarity level, then the type of the cut piece to be identified is determined to be a post-cut piece.

5. The method according to claim 2, characterized in that, The step of determining the similarity between the cut piece to be identified and the preset cut piece type template based on the transformed image and the second contour includes: The degree of similarity is determined based on the intersection-union ratio of the transformed image and the second contour.

6. The method according to claim 5, characterized in that, Determining the similarity degree based on the intersection-union ratio of the transformed image and the second contour includes: Based on the second contour, the transformed image is subjected to image alignment processing to obtain the aligned transformed image; The degree of similarity is determined based on the intersection-union ratio of the second contour and the aligned transformed image.

7. A device for identifying the type of cut piece, characterized in that, include: The information acquisition module is used to acquire image information of the cropped piece to be identified; A contour determination module is used to determine the first contour of the cut piece to be identified based on the image information; The key point determination module is used to determine the first key point of the cut piece to be identified based on the first contour. The type recognition module is used to identify the type of the cut piece to be identified based on the degree of overlap between the first outline and the preset cut piece type template, and the key point falling into the preset cut piece type template. The type includes front cut piece and back cut piece.

8. A sewing machine, characterized in that, The method includes a memory and a processor, wherein the memory stores a computer program, which, when executed by the processor, implements the method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by one or more processors, implements the method of any one of claims 1-6.

Images