Artificial Intelligence Control Method and System for an Automatic Cloth Cutting and Conveying Device

Abstract

The present application provides an artificial intelligence control method and system for an automatic cutting and conveying device of fabric. By determining the cutting execution block where the target fabric is located in the cutting area; obtaining the fabric loosening signal of the target fabric based on the polygonal shape characteristics and the wrinkle shape characteristics determined by the cutting execution block; further determining multiple intelligent cutting landing points of the target fabric, and determining the first interaction instruction of the target fabric through the surface flattening data during surface tension adjustment of the target fabric and all the intelligent cutting landing points; in response to the operation of the first interaction instruction, determining the deviation cost of the left side and the deviation cost of the right side of the target fabric on the left side of the fabric tensioning shaft in the cutting preparation state, and further obtaining the second interaction instruction of the human-machine interaction under the intelligent cutting of the target fabric; in response to the operation of the second interaction instruction, controlling the cutting execution mechanism to automatically cut the target fabric. It can perform adaptive cutting of the target fabric under abnormal morphological changes by artificial intelligence automation control.

Description

Technical Field

[0001] The present application relates to the technical field of artificial intelligence control of cloth production, and more specifically, to an artificial intelligence control method and system for an automatic cloth cutting and conveying device. Background Art

[0002] With the development of intelligent manufacturing, artificial intelligence technology in the production field has gradually been applied, especially the application of artificial intelligence automatic control systems in cloth cutting has gradually become popular, which has greatly improved the level of automation in this production field and realized the intelligence of the cloth cutting process. Artificial intelligence technology based on computer vision can detect and identify the shape, position and defects of the cloth in real time, and optimize the cutting path through deep learning algorithms, thereby ensuring the accuracy and efficiency of cutting.

[0003] In the prior art, the artificial intelligence automatic control system for cloth cutting first uses a visual sensor to capture the cloth image and preprocess it, analyzes the image based on a deep learning model to identify boundaries, patterns and defects, generates an accurate cutting path, and finally performs the cutting operation through the robot according to the cutting path through accurate automatic control. In addition, it can also continuously monitor and feedback through force sensors and visual sensors in real time, automatically adjust the path or motion parameters, ensure operational stability and accuracy, so as to realize the artificial intelligence control of the cloth cutting process; however, in the cutting control of the existing automatic cloth cutting and conveying device, the target cloth will be subjected to uneven pressure or random extrusion, so that the target cloth will produce irregular and messy wrinkles (i.e., creases and ripples on the surface of the cloth) and edge morphology changes (i.e., the edge of the cloth presents an irregular shape), making it difficult to effectively identify the cutting point of the target cloth during cutting, thereby causing the cutting path to deviate and reduce the production quality of the target cloth. Therefore, how to perform artificial intelligence automatic control of adaptive cutting of the target cloth under abnormal morphological changes, thereby improving the production quality of the target cloth has become a difficult problem faced by the industry. Summary of the invention

[0004] The present application provides an artificial intelligence control method and system for an automatic cloth cutting and conveying device, which can perform artificial intelligence automated control of adaptive cutting of target cloth with abnormal morphological changes, thereby improving the production quality of the target cloth.

[0005] In a first aspect, the present application provides an artificial intelligence control method for an automatic cloth cutting and conveying device, comprising the following steps:

[0006] In response to the operation of the fabric production instruction, an intelligent monitoring image of the target fabric in the cutting area is collected, and a cutting execution block in the cutting area where the target fabric is located is determined based on the intelligent monitoring image;

[0007] Determine the multi - angular morphological features and wrinkling morphological features of the target fabric in the intelligent cutting state according to the cutting execution block, and then determine the fabric loosening signal of the target fabric in the intelligent cutting state through the multi - angular morphological features and the wrinkling morphological features;

[0008] Based on the fabric loosening signal, obtain the surface flattening data of the target fabric when adjusting the surface tension, determine all the characteristic corner points of the target fabric after adjusting the surface tension, and then determine multiple intelligent cutting landing points of the target fabric according to all the characteristic corner points. Determine the first interaction instruction for human - machine interaction of the target fabric in intelligent cutting through the surface flattening data and all the intelligent cutting landing points;

[0009] In response to the operation of the first interaction instruction, set the cutting item in the cutting interaction control panel to the cutting preparation state, determine the deviation cost of the left side and the deviation cost of the right side of the target fabric on the left side of the fabric tensioning axis in the cutting preparation state, and determine the second interaction instruction for human - machine interaction of the target fabric in intelligent cutting according to the deviation cost of the left side and the deviation cost of the right side;

[0010] In response to the operation of the second interaction instruction, control the cutting execution mechanism to automatically cut the target fabric, and display the cutting result on the user control interface of the fabric automatic cutting and conveying device.

[0011] In some embodiments, determining the cutting execution block where the target fabric is located in the cutting area based on the intelligent monitoring image specifically includes:

[0012] Obtain the preset matching sample of the target fabric in the cutting area;

[0013] Perform sliding matching on the intelligent monitoring image with the preset matching sample to obtain multiple sliding matching areas, and determine the matching degree of each sliding matching area;

[0014] Extract the sliding matching area corresponding to the maximum matching degree, and use the sliding matching area as the cutting execution block where the target fabric is located in the cutting area.

[0015] In some embodiments, determining the multi - angular morphological features and wrinkling morphological features of the target fabric in the intelligent cutting state according to the cutting execution block specifically includes:

[0016] Extract the cutting execution edge of the cutting execution block;

[0017] Perform multi - corner point fitting on the cutting execution edge to obtain the multi - angular morphological features of the target fabric in the intelligent cutting state;

[0018] Perform local feature description on each pixel point in the cutting execution block to obtain the feature descriptors corresponding to each pixel point;

[0019] Determine the texture feature map of the target fabric from the feature descriptors of each pixel point;

[0020] Determine the wrinkle morphology characteristics of the target fabric in the intelligent cutting state according to the texture feature map.

[0021] In some embodiments, performing local feature description on each pixel point in the cutting execution block to obtain the feature descriptors corresponding to each pixel point specifically includes:

[0022] Select a pixel point in the cutting execution block as the selected pixel point, and obtain all adjacent pixel points of the selected pixel point;

[0023] Compare and describe the pixel value of the selected pixel point with the pixel values of each adjacent pixel point respectively to obtain the feature descriptor corresponding to the selected pixel point;

[0024] Continue to determine the feature descriptors corresponding to the remaining pixel points in the cutting execution block.

[0025] In some embodiments, determining the fabric loose signal of the target fabric in the intelligent cutting state through the multi-angle morphology feature and the wrinkle morphology feature specifically includes:

[0026] Obtain a preset multi-angle morphology influence factor and a wrinkle morphology influence factor;

[0027] Determine the fabric looseness of the target fabric according to the multi-angle morphology influence factor, the wrinkle morphology influence factor, the multi-angle morphology feature and the wrinkle morphology feature;

[0028] Generate a fabric loose signal of the target fabric in the intelligent cutting state through the fabric looseness.

[0029] In some embodiments, determining all the characteristic corner points of the target fabric after surface tension adjustment specifically includes:

[0030] Collect the fabric monitoring map of the target fabric after surface tension adjustment;

[0031] Select a pixel point from the fabric monitoring map as the selected pixel point, and determine the horizontal gray-scale change amount and the vertical gray-scale change amount of the selected pixel point;

[0032] Construct a gray-scale autocorrelation matrix of the selected pixel point according to the horizontal gray-scale change amount and the vertical gray-scale change amount;

[0033] Determine the corner point response degree corresponding to the selected pixel point according to the gray-scale autocorrelation matrix;

[0034] Continue to determine the corner response degrees corresponding to the remaining pixel points in the fabric monitoring map;

[0035] Compare the corner response degree corresponding to each pixel point with a preset corner response degree, extract all the corner response degrees greater than the preset corner response degree, and then use the pixel points corresponding to the extracted corner response degrees as the characteristic corner points of the target fabric after surface tension adjustment.

[0036] In some embodiments, an intelligent monitoring image of the target fabric in the cutting area is collected by a camera installed on the fabric automatic cutting and conveying device.

[0037] In a second aspect, the present application provides an artificial intelligence control system for a fabric automatic cutting and conveying device, which includes a cutting control unit, and the cutting control unit includes:

[0038] An acquisition module, which is used to respond to the operation of the fabric production instruction, collect the intelligent monitoring image of the target fabric in the cutting area, and determine the cutting execution block where the target fabric is located in the cutting area based on the intelligent monitoring image;

[0039] A processing module, which is used to determine the multi-corner shape characteristics and fold shape characteristics of the target fabric in the intelligent cutting state according to the cutting execution block, and then determine the fabric loose signal of the target fabric in the intelligent cutting state through the multi-corner shape characteristics and the fold shape characteristics;

[0040] The processing module is further used to obtain the surface flattening data of the target fabric when performing surface tension adjustment based on the fabric loose signal, determine all the characteristic corner points of the target fabric after surface tension adjustment, and then determine multiple intelligent cutting landing points of the target fabric according to all the characteristic corner points. The first interaction instruction for human-machine interaction of the target fabric in intelligent cutting is determined through the surface flattening data and all the intelligent cutting landing points;

[0041] The processing module is further used to respond to the operation of the first interaction instruction, set the cutting item in the cutting interaction control panel to the cutting preparation state, determine the deviation cost of the left side and the deviation cost of the right side of the target fabric at the left side of the fabric tensioning shaft in the cutting preparation state, and determine the second interaction instruction for human-machine interaction of the target fabric in intelligent cutting according to the deviation cost of the left side and the deviation cost of the right side;

[0042] An execution module, which is used to respond to the operation of the second interaction instruction, control the cutting execution mechanism to automatically cut the target fabric, and display the cutting result on the user control interface of the fabric automatic cutting and conveying device.

[0043] In a third aspect, the present application provides a computer device, which includes a memory and a processor. The memory stores code, and the processor is configured to obtain the code and execute the artificial intelligence control method of the automatic cloth cutting and conveying device described above.

[0044] In a fourth aspect, the present application provides a computer-readable storage medium storing a computer program, which when executed by a processor implements the artificial intelligence control method of the automatic cloth cutting and conveying device described above.

[0045] The technical solutions provided by the disclosed embodiments of the present application have the following beneficial effects:

[0046] In the artificial intelligence control method and system of the automatic cloth cutting and conveying device provided by the present application, first, in response to the operation of the cloth production instruction, an intelligent monitoring image of the target cloth in the cutting area is collected, and the cutting execution block where the target cloth is located in the cutting area is determined based on the intelligent monitoring image; secondly, according to the cutting execution block, the multi-corner shape characteristics and the wrinkle shape characteristics of the target cloth in the intelligent cutting state are determined, and then the cloth loosening signal of the target cloth in the intelligent cutting state is determined through the multi-corner shape characteristics and the wrinkle shape characteristics; further, based on the cloth loosening signal, the surface flattening data of the target cloth during surface tension adjustment is obtained, all the characteristic corner points of the target cloth after surface tension adjustment are determined, and then multiple intelligent cutting drop points of the target cloth are determined according to all the characteristic corner points. The first interaction instruction for human-machine interaction of the target cloth in intelligent cutting is determined through the surface flattening data and all the intelligent cutting drop points; then, in response to the operation of the first interaction instruction, the cutting item in the cutting interaction control panel is set to the cutting preparation state, the deviation cost on the left side and the deviation cost on the right side of the target cloth in the cutting preparation state are determined, and the second interaction instruction for human-machine interaction of the target cloth in intelligent cutting is determined according to the deviation cost on the left side and the deviation cost on the right side; finally, in response to the operation of the second interaction instruction, the cutting execution mechanism is controlled to automatically cut the target cloth, and the cutting result is displayed on the user control interface of the automatic cloth cutting and conveying device.

[0047] It can be seen that the present application can achieve artificial intelligence automated control for adaptively cutting a target fabric under morphological changes, thereby improving the production quality of the target fabric. First, the cutting execution block where the target fabric is located in the cutting area is extracted, effectively limiting the recognition range of cutting execution, and thus avoiding the confusion interference brought by the background area. Secondly, based on the limited cutting execution recognition range, the multi-corner morphological features and wrinkled morphological features of the target fabric are extracted, effectively identifying the wrinkled morphology and multi-corner morphology of the target fabric, and thus avoiding the influence brought by the uneven pressure and random extrusion of the target fabric. Further, the surface tension of the target fabric is adjusted according to the fabric loose signal determined by the wrinkled morphology and multi-corner morphology of the target fabric, realizing the adjustment of the loose state of the target fabric, and determining the first state setting instruction for human-computer interaction in the intelligent cutting of the target fabric through all the characteristic corner points and surface flattening data of the target fabric after surface tension adjustment, effectively ensuring that the cutting device is in an appropriate working state before starting cutting, so as to ensure that each cutting can be accurately carried out as expected. Then, based on the deviation cost on the left side and the deviation cost on the right side of the fabric tensioning axis of the target fabric, the second state setting instruction for human-computer interaction in the intelligent cutting of the target fabric is obtained, further determining the cutting state of the cut fabric, and ensuring that the cutting device performs an adaptive cutting operation after meeting the predetermined conditions. Finally, in response to the state setting instruction, the cutting execution mechanism is controlled to automatically cut the target fabric, and the cutting result is displayed on the user control interface of the automatic fabric cutting and conveying device. In summary, the technical solution provided by the present application can perform artificial intelligence automated control for adaptively cutting the target fabric under abnormal morphological changes, thereby improving the production quality of the target fabric. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 is an exemplary flowchart of an artificial intelligence control method for an automatic fabric cutting and conveying device according to some embodiments of the present application;

[0049] Figure 2 is an exemplary flowchart of determining a cutting execution block according to some embodiments of the present application;

[0050] Figure 3 is an exemplary flowchart of determining multi-corner morphological features and wrinkled morphological features according to some embodiments of the present application;

[0051] Figure 4 is a schematic diagram of exemplary hardware and / or software of a cutting control unit according to some embodiments of the present application;

[0052] Figure 5 is a schematic diagram of the structure of a computer device for implementing an artificial intelligence control method for an automatic fabric cutting and conveying device according to some embodiments of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0053] The core of this application is to first respond to the operation of the fabric production instruction, collect the intelligent monitoring images of the target fabric in the cutting area, and determine the cutting execution block where the target fabric is located in the cutting area based on the intelligent monitoring images. Secondly, determine the multi-corner morphological characteristics and wrinkle morphological characteristics of the target fabric in the intelligent cutting state according to the cutting execution block, and then determine the fabric loose signal of the target fabric in the intelligent cutting state through the multi-corner morphological characteristics and the wrinkle morphological characteristics. Further, obtain the surface flattening data of the target fabric when adjusting the surface tension based on the fabric loose signal, determine all the characteristic corner points of the target fabric after adjusting the surface tension, and then determine multiple intelligent cutting landing points of the target fabric according to all the characteristic corner points. Determine the first interaction instruction of human-computer interaction of the target fabric under intelligent cutting through the surface flattening data and all the intelligent cutting landing points. Then, respond to the operation of the first interaction instruction, set the cutting item in the cutting interaction control panel to the cutting preparation state, determine the deviation cost of the left side and the deviation cost of the right side of the target fabric on the left side of the fabric tensioning axis in the cutting preparation state, and determine the second interaction instruction of human-computer interaction of the target fabric under intelligent cutting according to the deviation cost of the left side and the deviation cost of the right side. Finally, respond to the operation of the second interaction instruction, control the cutting execution mechanism to automatically cut the target fabric, and display the cutting result on the user control interface of the fabric automatic cutting and conveying device, so as to perform artificial intelligence automation control for adaptive cutting of the target fabric under abnormal morphological changes, thereby improving the production quality of the target fabric.

[0054] To better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the specification drawings and specific implementation manners. Refer to Figure 1 , this figure is an exemplary flowchart of the artificial intelligence control method of the fabric automatic cutting and conveying device according to some embodiments of the present application. The artificial intelligence control method 100 of the fabric automatic cutting and conveying device mainly includes the following steps:

[0055] In step 101, respond to the operation of the fabric production instruction, collect the intelligent monitoring images of the target fabric in the cutting area, and determine the cutting execution block where the target fabric is located in the cutting area based on the intelligent monitoring images.

[0056] In specific implementation, when a user sends a cloth production instruction to a computer, the computer responds to the operation of the cloth production instruction and collects an intelligent monitoring image of the target cloth in the cutting area through a camera installed on the automatic cloth cutting and conveying device. In this application, the intelligent monitoring image refers to an image used to intelligently monitor the state and position of the target cloth during the cutting process. By collecting the intelligent monitoring image, the accurate position of the cloth in the cutting area can be effectively displayed to ensure correct alignment of the cloth, and the cutting path can be adjusted according to the actual position of the cloth.

[0057] In some embodiments, referring to Figure 2 as shown, this figure is an exemplary flowchart for determining a cutting execution block according to some embodiments of this application. In this embodiment, the following steps can be used to determine the cutting execution block where the target cloth is located in the cutting area based on the intelligent monitoring image:

[0058] First, in step 1011, obtain a preset matching sample of the target cloth in the cutting area;

[0059] Then, in step 1012, perform sliding matching on the intelligent monitoring image with the preset matching sample to obtain multiple sliding matching areas, and determine the matching degree of each sliding matching area;

[0060] Finally, in step 1013, extract the sliding matching area corresponding to the maximum matching degree, and use the sliding matching area as the cutting execution block where the target cloth is located in the cutting area.

[0061] In specific implementation, the preset matching sample of the target cloth in the cutting area can be obtained through the automatic cloth cutting database. In this embodiment, the preset matching sample refers to a set of standardized samples predefined and stored in the automatic cloth cutting system, which is used to compare and match with the actually collected intelligent monitoring image. The preset matching sample contains information such as the position, shape, size, and cutting path of the cloth in the cutting area under ideal conditions. By matching the actual cloth image with the preset sample, the accuracy and consistency of cutting can be effectively ensured, thereby reducing the confusion interference caused by the background area.

[0062] It should be noted that in this embodiment, the automatic cloth cutting database is a database system for storing and managing data related to automatic cloth cutting. The purpose of this database is to support various operations and decisions during the cloth cutting process to ensure the accuracy and efficiency of cutting.

[0063] During specific implementation, the preset matching sample is used to perform sliding matching on the intelligent monitoring image to obtain multiple sliding matching regions, that is: starting from the upper left corner of the intelligent monitoring image, the preset matching sample traverses and matches from left to right and from top to bottom. Each time a match is made, a sliding matching region at the corresponding matching position in the intelligent monitoring image is obtained, and thus multiple sliding matching regions are obtained.

[0064] During specific implementation, the matching degree of each sliding matching region can be calculated by the cross-correlation matching method, which will not be elaborated here. In addition, in other embodiments, other matching algorithms can also be used to calculate the matching degree of each sliding matching region, which is not limited here.

[0065] It should be noted that in this embodiment, sliding matching refers to the matching process where the preset matching sample starts from the upper left corner of the target image and traverses from left to right and from top to bottom, and the matching degree is calculated each time a match is made. That is, in this application, it is the process of the preset matching sample matching the intelligent monitoring image, which will not be elaborated here.

[0066] It should also be noted that in this embodiment, the sliding matching region refers to the region at the corresponding matching position in the intelligent monitoring image each time the preset matching sample makes a match. In this embodiment, the matching degree refers to the similarity between the preset matching sample and the image region at the matching position, that is, the greater the matching degree, the more similar they are, and vice versa.

[0067] In addition, it should be noted that in this application, the cutting execution block represents the image region where the target fabric performs the cutting operation, that is, the maximum range of this image region includes the position information and the range of the target fabric. By determining the cutting execution block, the position and state of the target fabric can be effectively identified during the cutting operation, so as to dynamically adjust the cutting execution range and the state of the target fabric, ensuring the accuracy and flexibility of the cutting operation.

[0068] In step 102, based on the cutting execution block, the polygonal shape feature and the wrinkle shape feature of the target fabric in the intelligent cutting state are determined, and then the fabric looseness signal of the target fabric in the intelligent cutting state is determined through the polygonal shape feature and the wrinkle shape feature.

[0069] In some embodiments, refer to Figure 3 As shown, this figure is an exemplary flowchart for determining the polygonal shape feature and the wrinkle shape feature according to some embodiments of the present application. In this embodiment, the polygonal shape feature and the wrinkle shape feature of the target fabric in the intelligent cutting state can be determined based on the cutting execution block by the following steps:

[0070] First, in step 1021, the cutting execution edge of the cutting execution block is extracted;

[0071] Secondly, in step 1022, perform multi-corner fitting on the cutting execution edge to obtain the multi-corner morphological features of the target fabric in the intelligent cutting state;

[0072] Furthermore, in step 1023, perform local feature description on each pixel point in the cutting execution block to obtain the feature descriptors corresponding to each pixel point;

[0073] Then, in step 1024, determine the texture feature map of the target fabric from the feature descriptors of each pixel point;

[0074] Finally, in step 1025, determine the wrinkle morphological features of the target fabric in the intelligent cutting state according to the texture feature map.

[0075] Specifically, the Canny algorithm can be used to extract the cutting execution edge of the cutting execution block. In addition, in other embodiments, other edge extraction methods can also be used to extract the cutting execution edge of the cutting execution block, such as threshold segmentation method and morphological processing method, which are not limited here. It should be noted that in this embodiment, the cutting execution edge represents the edge contour of the target fabric in the cutting execution block. By determining the cutting execution edge, the edge state of the current target fabric can be effectively identified.

[0076] Specifically, the Douglas-Peucker polygon simplification algorithm can be used to perform multi-corner fitting on the cutting execution edge to obtain the multi-corner morphological features of the target fabric in the intelligent cutting state. The Douglas-Peucker polygon simplification algorithm is an algorithm used to recursively reduce the number of edge vertices while maintaining the shape features of the original contour, which will not be elaborated here. In addition, in other embodiments, other multi-corner fitting methods can also be used for fitting, such as the Visvalingam-Whyatt polygon simplification algorithm, which is not limited here.

[0077] It should be noted that in this embodiment, multi-corner fitting refers to the process of simplifying a complex contour into a polygon composed of a small number of vertices by retaining key points and deleting unnecessary intermediate points. In addition, in this application, the multi-corner morphological features refer to the multi-corner features of the target fabric in the cutting execution block, which are used to describe the complex contour of the target fabric in the cutting execution block. For example, the number of vertices on the edge of the target fabric in the cutting execution block, and the number of vertices is usually directly related to the shape complexity of the target fabric. In addition, the multi-corner morphological features can also be other features, such as the perimeter or area of the edge of the target fabric in the cutting execution block, which are not limited here. By determining the multi-corner morphological features, the boundary and shape of the fabric can be accurately identified, ensuring that the cutting tool moves along the correct path to avoid cutting errors and improve the cutting accuracy and quality.

[0078] Among them, in some embodiments, local feature description is performed on each pixel point in the cutting execution block, and the specific steps for obtaining the feature descriptor corresponding to each pixel point can be as follows:

[0079] Select a pixel point in the cutting execution block as the selected pixel point, and obtain all adjacent pixel points of the selected pixel point;

[0080] Compare and describe the pixel value of the selected pixel point with the pixel values of each adjacent pixel point respectively to obtain the feature descriptor corresponding to the selected pixel point;

[0081] Continue to determine the feature descriptors corresponding to the remaining pixel points in the cutting execution block.

[0082] Specifically, all adjacent pixel points of the selected pixel point can be obtained by the eight-neighborhood method, that is, taking the selected pixel point as the center, and taking all pixel points adjacent to the selected pixel point as adjacent pixel points, which will not be elaborated here. In addition, in other embodiments, other methods can also be used to obtain all adjacent pixel points of the selected pixel point, which is not limited here.

[0083] Specifically, when comparing and describing the pixel value of the selected pixel point with the pixel values of each adjacent pixel point respectively to obtain the feature descriptor corresponding to the selected pixel point, that is: compare the pixel value of the selected pixel point with the pixel values of each adjacent pixel point respectively. When the pixel value of an adjacent pixel point is greater than the pixel value of the selected pixel point, set the pixel value of this adjacent pixel point to 1, otherwise set the pixel value of this adjacent pixel point to 0. Combine each pixel value obtained after comparison to get a binary number, and perform decimal conversion on the binary number to obtain the feature descriptor corresponding to the selected pixel point.

[0084] It should be noted that in this embodiment, local feature description represents the process of describing the local features of each pixel point in the cutting execution block. In this embodiment, local feature description is performed on each pixel point in the cutting execution block, that is: select a pixel point in the cutting execution block as the selected pixel point, and obtain all adjacent pixel points of the selected pixel point; compare and describe the pixel value of the selected pixel point with the pixel values of each adjacent pixel point respectively to obtain the feature descriptor corresponding to the selected pixel point; continue to determine the feature descriptors corresponding to the remaining pixel points in the cutting execution block, that is, to realize local feature description of each pixel point in the cutting execution block.

[0085] It should also be noted that in this embodiment, the feature descriptor represents a parameter for describing the local neighborhood information of feature points in the cutting execution block. The feature descriptor encodes the texture, color, or shape information around the key points in the image, enabling these feature points to be matched and compared between different images. During the cutting process of the fabric, wrinkles may occur on the fabric. When wrinkles occur, the wrinkled areas usually present a situation that is not similar to the surrounding features. Therefore, by determining the feature descriptor, the wrinkled form of the target fabric can be effectively identified.

[0086] When specifically implemented, the texture feature map of the target fabric is determined by the feature descriptor of each pixel point, that is: the feature descriptor of each pixel point is correspondingly substituted for the original pixel value to obtain the texture feature map of the target fabric.

[0087] It should be noted that in this embodiment, the texture feature map represents an image used to describe the local texture features in the image.

[0088] When specifically implemented, the wrinkled form feature of the target fabric in the intelligent cutting state is determined according to the texture feature map, that is: the total number of pixel points in the texture feature map whose pixel values are greater than the preset pixel value is obtained, the total number of pixel points in the texture feature map is obtained, and the wrinkled form feature of the target fabric in the intelligent cutting state is determined by the ratio of the total number of pixel points in the texture feature map whose pixel values are greater than the preset pixel value to the total number of pixel points in the texture feature map. In addition, in other embodiments, other calculation methods may also be used to calculate the wrinkled form feature, which is not limited here.

[0089] It should be noted that in this application, the wrinkled form feature represents the wrinkled features formed on the surface of the target fabric. During the cutting process of the target fabric, when there are wrinkles on the target fabric, the effectiveness and accuracy of cutting will decrease, and when the target fabric is photographed, there will be a large amount of texture features in the wrinkled part. Therefore, by obtaining the texture features of the target fabric and using the proportion of the texture features as the wrinkled form feature of the target fabric in the intelligent cutting state, the effectiveness and accuracy of cutting can be effectively improved, thereby avoiding mis-cutting caused by fabric wrinkles, and ensuring the accuracy and consistency of cutting.

[0090] In some embodiments, the following steps may be specifically adopted to determine the fabric looseness signal of the target fabric in the intelligent cutting state through the multi-angle form feature and the wrinkled form feature, that is:

[0091] Obtain the preset multi-angle form influence factor and wrinkled form influence factor;

[0092] Determine the fabric looseness of the target fabric according to the multi-angle form influence factor, the wrinkled form influence factor, the multi-angle form feature, and the wrinkled form feature;

[0093] Generate a fabric looseness signal of the target fabric in the intelligent cutting state based on the fabric looseness.

[0094] When specifically implemented, the preset multi - angular shape influence factor and the wrinkle shape influence factor can be obtained through the fabric automatic cutting database. The preset multi - angular shape influence factor is the weight of the multi - angular shape affecting the cutting quality of the target fabric, which can be specifically set according to its influencing degree. The preset wrinkle shape influence factor is the weight of the wrinkle shape affecting the cutting quality of the target fabric, which can be specifically set according to its influencing degree. In this embodiment, since the influence of the wrinkle shape on the cutting quality of the target fabric is greater than that of the multi - angular shape on the cutting quality of the target fabric, therefore, the multi - angular shape influence factor is set to 0.46 and the wrinkle shape influence factor is set to 0.64. In addition, in other embodiments, it can also be set according to actual needs and is not limited here.

[0095] When specifically implemented, determine the fabric looseness of the target fabric according to the multi - angular shape influence factor, the wrinkle shape influence factor, the multi - angular shape feature, and the wrinkle shape feature, that is: First, perform a product calculation on the multi - angular shape influence factor and the multi - angular shape feature to obtain a first product calculation result. Second, perform a product calculation on the wrinkle shape influence factor and the wrinkle shape feature to obtain a second product calculation result. Finally, sum the first product calculation result and the second product calculation result to obtain the fabric looseness of the target fabric. In addition, in other embodiments, other calculation methods can also be used to calculate the fabric looseness of the target fabric and are not limited here.

[0096] It should be noted that in this embodiment, the fabric looseness represents the degree of looseness presented on the fabric surface, that is, the greater the fabric looseness, the greater the degree of looseness presented on the fabric surface, and the smaller the fabric looseness, the smaller the degree of looseness presented on the fabric surface. During the cutting of the target fabric, the loose target fabric is prone to movement, which will cause the cutting edge to be uneven, resulting in a rough or irregular edge. Therefore, by determining the fabric looseness, the cutting parameters and fixing measures can be effectively adjusted to ensure the cutting accuracy and quality.

[0097] When specifically implemented, generate a fabric looseness signal of the target fabric in the intelligent cutting state based on the fabric looseness, that is: when the artificial intelligence system of the fabric automatic cutting and conveying device receives the fabric looseness, the artificial intelligence system automatically generates a fabric looseness signal of the target fabric in the intelligent cutting state.

[0098] It should be noted that in this embodiment, the fabric looseness signal is information characterizing the degree of fabric looseness. The fabric looseness signal can play an important role in the intelligent cutting system. By analyzing these signals, the cutting process can be optimized and the cutting quality can be improved.

[0099] It should also be noted that in this application, the artificial intelligence system refers to an intelligent system for the automatic cutting and conveying device of fabrics. It realizes the automation, intelligence, and precision in the fabric cutting process through artificial intelligence (AI) technology, which will not be elaborated here.

[0100] In step 103, based on the fabric looseness signal, obtain the surface flattening data of the target fabric when adjusting the surface tension, determine all the characteristic corner points of the target fabric after adjusting the surface tension, and then determine multiple intelligent cutting landing points of the target fabric according to all the characteristic corner points. Determine the first interaction instruction of human-computer interaction in the intelligent cutting of the target fabric through the surface flattening data and all the intelligent cutting landing points.

[0101] In some embodiments, the specific steps for obtaining the surface flattening data of the target fabric when adjusting the surface tension based on the fabric looseness signal are as follows:

[0102] Send the fabric looseness signal to the artificial intelligence system of the fabric automatic cutting and conveying device. When the artificial intelligence system receives the fabric looseness signal, adjust the surface tension of the target fabric according to a preset adjustment amount, and use an ultrasonic sensor to scan the surface of the target fabric to obtain the surface flattening data of the target fabric when adjusting the surface tension.

[0103] It should be noted that in this embodiment, a surface tension threshold is set for the surface tension, and the adjustment degree during the adjustment process cannot exceed the surface tension threshold. In this embodiment, the surface flattening data includes multiple surface flattening values, and the surface flattening value represents a measure of the flatness degree of the surface of the target fabric. In addition, in other embodiments, other scanning sensors can also be used to scan the surface of the target fabric to obtain the surface flattening data of the target fabric, which is not limited here.

[0104] In some embodiments, the specific steps for determining all the characteristic corner points of the target fabric after adjusting the surface tension are as follows:

[0105] Collect the fabric monitoring image of the target fabric after adjusting the surface tension;

[0106] Select a pixel point from the fabric monitoring image as the selected pixel point, and determine the horizontal gray level change amount and the vertical gray level change amount of the selected pixel point;

[0107] Construct a gray - level autocorrelation matrix of the selected pixel point according to the horizontal gray - level change amount and the vertical gray - level change amount;

[0108] Determine the corner response degree corresponding to the selected pixel point according to the gray - level autocorrelation matrix;

[0109] Continue to determine the corner response degrees corresponding to the remaining pixel points in the fabric monitoring image;

[0110] Compare the corner response degree corresponding to each pixel point with a preset corner response degree, extract all corner response degrees greater than the preset corner response degree, and then take the pixel points corresponding to the extracted corner response degrees as the characteristic corner points of the target fabric after surface tension adjustment.

[0111] Specifically, when implemented, a camera installed on the fabric automatic cutting and conveying device is used to collect a fabric monitoring image of the target fabric after surface tension adjustment, and the fabric monitoring image represents an image of the target fabric taken after surface tension adjustment.

[0112] Specifically, when implemented, the horizontal gray - level change amount and the vertical gray - level change amount of the selected pixel point can be calculated by using the Sobel operator in image - processing technology, which will not be elaborated here. In addition, in other embodiments, the Prewitt operator can also be used to calculate the horizontal gray - level change amount and the vertical gray - level change amount of the selected pixel point, which is not limited here.

[0113] It should be noted that in this embodiment, the horizontal gray - level change amount represents the gray - level change amount of the current pixel point in the horizontal direction, and the vertical gray - level change amount represents the gray - level change amount of the current pixel point in the vertical direction.

[0114] Specifically, when implemented, construct a gray - level autocorrelation matrix of the selected pixel point according to the horizontal gray - level change amount and the vertical gray - level change amount, that is: take the square value of the horizontal gray - level change amount and the square value of the vertical gray - level change amount as the first element and the second element on the main diagonal of the gray - level autocorrelation matrix respectively, and take the product value of the horizontal gray - level change amount and the vertical gray - level change amount as the element on the secondary diagonal of the gray - level autocorrelation matrix, and then obtain the gray - level autocorrelation matrix of the selected pixel point. In addition, in other implementations, other construction methods can also be used to construct the gray - level autocorrelation matrix, which is not limited here.

[0115] It should be noted that in this embodiment, the gray - level autocorrelation matrix represents the correlation matrix between the gray - level values of the current pixel point in the horizontal direction and the vertical direction. By constructing the gray - level autocorrelation matrix, the spatial gray - level distribution characteristics of the current pixel point can be effectively identified.

[0116] In specific implementation, the corner response function in image processing can be used to calculate the gray - level autocorrelation matrix to obtain the corner response degree corresponding to the selected pixel point. The corner response function is a mathematical function in image processing used to detect and locate feature corners in an image. In addition, in other embodiments, other calculation methods can also be used to calculate the corner response degree corresponding to the selected pixel point, which is not limited here. It should be noted that in this embodiment, the corner response degree represents a metric value used to evaluate whether the current pixel point is a feature corner, that is, the possibility of a feature corner is determined by calculating the eigenvalues of the autocorrelation matrix. By determining the corner response degree, important feature points in the image can be effectively identified and extracted.

[0117] It should be noted that the preset corner response degree in this embodiment is a pre - set judgment corner response degree, which can be specifically set according to actual needs and will not be elaborated here. In addition, in this embodiment, all corner response degrees less than or equal to the preset corner response degree are not processed.

[0118] It also needs to be noted that in this embodiment, feature corners represent unique points in the cloth monitoring image. These points have obvious gray - level changes or texture changes in different regions and are prominent features within a local range. During the intelligent cutting process of the target cloth, feature corners, as significant local feature points in the cloth monitoring image, can accurately identify the position, posture, and boundary of the cloth. The cutting device can precisely control the movement path of the cutter or scissors according to the position and features of the feature corners, thereby avoiding rough or irregular problems at the cloth edge.

[0119] In some embodiments, the following steps can be specifically adopted to determine multiple intelligent cutting landing points of the target cloth based on all feature corners, that is:

[0120] Perform cutting matching on all feature corners to obtain the cutting matching deviation degree of each feature corner;

[0121] Compare the cutting matching deviation degree of each feature corner with the preset matching deviation degree, extract all cutting matching deviation degrees greater than the preset matching deviation degree, and use the feature corners corresponding to the extracted cutting matching deviation degrees as the intelligent cutting landing points when the target cloth is intelligently cut.

[0122] It should be noted that in this application, one feature corner corresponds to one pixel point, and one pixel point corresponds to one gray - level value. Therefore, one feature corner also corresponds to one gray - level value.

[0123] Among them, in some embodiments, the following steps can be specifically adopted to perform cutting matching on all feature corners to obtain the cutting matching deviation degree of each feature corner, that is:

[0124] Obtain the gray values corresponding to each feature corner point, perform encoding conversion on the gray values corresponding to each feature corner point, and obtain the gray encodings corresponding to each feature corner point;

[0125] Match and associate the gray encodings corresponding to each feature corner point with a preset gray encoding, and then obtain the cutting matching skewness of each feature corner point.

[0126] Specifically, binary encoding conversion can be performed on the gray values corresponding to each feature corner point to obtain the gray encodings corresponding to each feature corner point. The gray encoding represents the encoding obtained after binary conversion of the gray value. In addition, in other embodiments, other encoding conversions can also be used for conversion, which is not limited here.

[0127] Specifically, the gray encodings corresponding to each feature corner point can be matched and associated with the preset gray encoding through the Hamming distance in the prior art, and then the cutting matching skewness of each feature corner point can be obtained, that is, the gray encoding is compared with the preset gray encoding bit by bit, and the number of different bits is counted to obtain the cutting matching skewness. In addition, in other embodiments, other matching and association methods can also be used for matching and association, which is not limited here.

[0128] It should be noted that the cutting matching skewness in this embodiment represents the deviation degree of cutting matching. In addition, the cutting matching in this embodiment represents the process of performing cutting landing point matching. Among them, cutting matching is performed on all feature corner points, that is: obtain the gray values corresponding to each feature corner point, perform encoding conversion on the gray values corresponding to each feature corner point, and obtain the gray encodings corresponding to each feature corner point; match and associate the gray encodings corresponding to each feature corner point with the preset gray encoding, and then obtain the cutting matching skewness of each feature corner point, that is, the cutting matching of all feature corner points is realized.

[0129] It should also be noted that the preset gray encoding in this embodiment is a standard gray encoding set in advance, which can be specifically set according to actual needs and is not limited here. In addition, all cutting matching skewnesses less than or equal to the preset matching skewness are not processed.

[0130] In addition, it should be noted that the intelligent cutting landing point in this application represents the actual cutting point when cutting the target fabric. The intelligent cutting landing point defines the points that the cutting process needs to pass through. Therefore, by determining the intelligent cutting landing point, the accuracy and efficiency of the cutting operation can be effectively ensured.

[0131] Among them, in some embodiments, the first interaction instruction of the human-computer interaction of the target fabric under intelligent cutting is specifically determined by the surface flattening data and all intelligent cutting landing points, and the following steps can be adopted, that is:

[0132] Determine the surface flattening trend of the target fabric under intelligent cutting according to the surface flattening data;

[0133] Determine the longest cutting radius of the target fabric under intelligent cutting according to all intelligent cutting landing points;

[0134] Determine the cutting path loss of the target fabric cutting through the surface flattening trend and the longest cutting radius, and transmit the cutting path loss to the artificial intelligence control system of the fabric automatic cutting and conveying device to automatically generate the first interaction instruction for human-computer interaction of the target fabric under intelligent cutting.

[0135] When specifically implemented, determine the surface flattening trend of the target fabric under intelligent cutting according to the surface flattening data. The surface flattening trend can be determined by calculating the standard deviation of the surface flattening data. In addition, in other embodiments, other calculation methods can also be used to calculate the surface flattening trend, which is not limited here.

[0136] It should be noted that in this embodiment, the surface flattening trend represents the flattening change trend of the surface of the target fabric after surface tension adjustment. By determining the surface flattening trend, it can be effectively judged whether the target fabric meets the cutting conditions. When the cutting conditions are met, the cutting operation is performed; otherwise, the surface of the target fabric needs to be adjusted for surface tension.

[0137] When specifically implemented, determine the longest cutting radius of the target fabric under intelligent cutting according to all intelligent cutting landing points, that is: select an intelligent cutting landing point, calculate the distances between this intelligent cutting landing point and the remaining intelligent cutting landing points respectively, and take the maximum calculated distance as the longest cutting radius of the target fabric during intelligent cutting.

[0138] It should be noted that the distances between this intelligent cutting landing point and the remaining intelligent cutting landing points can be calculated through the Euclidean distance. In addition, in other embodiments, other distance calculation methods can also be used for distance calculation, which is not limited here. Among them, the longest cutting radius in this embodiment represents the maximum cutting diameter of the target fabric cutting.

[0139] When specifically implemented, determine the cutting path loss of the target fabric cutting through the surface flattening trend and the longest cutting radius, that is: perform weighted summation on the surface flattening trend and the longest cutting radius to obtain the cutting path loss of the target fabric cutting. It should be noted that during the weighted summation process of the surface flattening trend and the longest cutting radius in this embodiment, the weights of the surface flattening trend and the longest cutting radius can be set according to actual needs. For example, they can be set respectively according to the influence degree of the surface flattening trend and the longest cutting radius on the cutting loss. In this embodiment, the weight of the surface flattening trend is set to 0.67, and the weight of the longest cutting radius is set to 0.33.

[0140] It should be noted that in this application, the first interaction instruction represents the first state setting instruction for the user to interact with the computer during the cutting process of the target fabric. The first interaction instruction is used to set the cutting preparation state of the target fabric. During the cutting of the target fabric, an unreasonable cutting path will lead to a decrease in cutting efficiency and poor cutting quality. Therefore, generating the first interaction instruction through the consumption of the cutting path can effectively ensure that the cutting device is in an appropriate working state before starting to cut, so as to ensure that each cut can be accurately carried out as expected.

[0141] In step 104, in response to the operation of the first interaction instruction, set the cutting item in the cutting interaction control panel to the cutting preparation state, determine the deviation cost of the left side and the right side of the target fabric on the left side of the fabric tensioning axis in the cutting preparation state, and determine the second interaction instruction for human-computer interaction in intelligent cutting of the target fabric according to the deviation cost of the left side and the deviation cost of the right side.

[0142] In specific implementation, in response to the operation of the first interaction instruction, the cutting item in the cutting interaction control panel can be set to the cutting preparation state, and the pre-cutting state represents the preparation state for cutting the target fabric.

[0143] In some embodiments, the following steps can be specifically adopted to determine the deviation cost of the left side and the right side of the target fabric on the left side of the fabric tensioning axis in the cutting preparation state, that is:

[0144] Obtain the left side monitoring image and the right side monitoring image of the target fabric at the fabric tensioning axis in the cutting preparation state;

[0145] Extract the overlapping area of the left side of the target fabric and the fabric tensioning axis in the left side monitoring image;

[0146] Extract the overlapping area of the right side of the target fabric and the fabric tensioning axis in the right side monitoring image;

[0147] Determine the deviation cost of the left side of the target fabric on the left side of the fabric tensioning axis in the cutting preparation state according to the left side overlapping area;

[0148] Determine the deviation cost of the right side of the target fabric on the right side of the fabric tensioning axis in the cutting preparation state according to the right side overlapping area.

[0149] It should be noted that in this application, the left side and the right side are only for the distinction of the scenario and have no practical significance, that is, the left side overlapping area is the overlapping area of the target fabric on the left side of the fabric tensioning axis, and the right side overlapping area is the overlapping area of the target fabric on the right side of the fabric tensioning axis.

[0150] In specific implementation, the left-side monitoring image and the right-side monitoring image of the target fabric at the fabric tensioning shaft in the cutting preparation state can be obtained through a camera. The left-side monitoring image represents the image obtained by monitoring the left side of the target fabric at the fabric tensioning shaft, and the right-side monitoring image represents the image obtained by monitoring the right side of the target fabric at the fabric tensioning shaft.

[0151] It should be noted that in this application, the fabric tensioning shaft refers to a shaft or roller device in fabric processing or cutting equipment that is used to keep the fabric moderately tensioned, flat, and stably transported. Its main function is to keep the fabric flat and fixed during processing or cutting by applying appropriate tension, preventing wrinkles, slack, and sliding, thereby improving processing accuracy and efficiency.

[0152] In specific implementation, the overlapping area between the target fabric and the left side of the fabric tensioning shaft in the left-side monitoring image and the overlapping area between the target fabric and the right side of the fabric tensioning shaft in the right-side monitoring image can be extracted through the Sobel operator in image processing technology, which will not be elaborated here. In addition, in other embodiments, the Prewitt operator can also be used for extraction, which is not limited here.

[0153] It should be noted that in this embodiment, the left-side overlapping area refers to the image area at the overlapping part of the target fabric and the left side of the fabric tensioning shaft extracted in the left-side monitoring image. The left-side overlap is the area where the target fabric and the left side of the fabric tensioning shaft cover and adjoin each other. In this embodiment, the right-side overlapping area refers to the image area at the overlapping part of the target fabric and the right side of the fabric tensioning shaft extracted in the right-side monitoring image. The right-side overlap is the area where the target fabric and the right side of the fabric tensioning shaft cover and adjoin each other.

[0154] Among them, in some embodiments, the following steps can be specifically adopted to determine the deviation cost of the target fabric on the left side of the fabric tensioning shaft in the cutting preparation state according to the left-side overlapping area, that is:

[0155] Obtain all different gray values in the left-side overlapping area;

[0156] Determine the distribution proportion of each different gray value;

[0157] Determine the deviation cost of the target fabric on the left side of the fabric tensioning shaft in the cutting preparation state according to the distribution proportion of all different gray values.

[0158] In specific implementation, to determine the distribution proportion of each different gray value, that is: for each different gray value, obtain the frequency of occurrence of this different gray value in the left-side overlapping area, and take the ratio of the frequency to the total number of pixels in the left-side overlapping area as the distribution proportion of this different gray value, and then obtain the distribution proportion of each different gray value.

[0159] It should be noted that in this embodiment, the distribution ratio represents the distribution ratio of different gray values in the overlapping area on the left side edge.

[0160] In specific implementation, the deviation cost of the target fabric on the left side of the fabric tensioning axis in the cutting preparation state is determined according to the distribution ratio of all different gray values, that is: the distribution ratios of all different gray values are used as input parameters to be input into the information entropy model for calculation, and the output result is used as the deviation cost of the target fabric on the left side of the fabric tensioning axis in the cutting preparation state. In addition, other entropy models can also be used for processing, which is not limited here.

[0161] It should be noted that in this application, the deviation cost of the left side edge represents the degree of deviation of the target fabric on the left side of the fabric tensioning axis, that is, the greater the deviation cost of the left side edge, the greater the degree of deviation of the target fabric on the left side of the fabric tensioning axis, and the smaller the deviation cost of the left side edge, the smaller the degree of deviation of the target fabric on the left side of the fabric tensioning axis.

[0162] Among them, in some embodiments, the following steps can be specifically adopted to determine the deviation cost of the target fabric on the right side of the fabric tensioning axis in the cutting preparation state according to the overlapping area on the right side edge, that is:

[0163] Obtain all different gray values in the overlapping area on the right side edge;

[0164] Determine the distribution ratio of each different gray value;

[0165] Determine the deviation cost of the target fabric on the right side of the fabric tensioning axis in the cutting preparation state according to the distribution ratios of all different gray values.

[0166] In specific implementation, to determine the distribution ratio of each different gray value, that is: for each different gray value, obtain the frequency of occurrence of this different gray value in the overlapping area on the right side edge, and use the ratio of the frequency to the total number of pixels in the overlapping area on the right side edge as the distribution ratio of this different gray value, so as to obtain the distribution ratio of each different gray value.

[0167] It should be noted that in this embodiment, the distribution ratio represents the distribution ratio of different gray values in the overlapping area on the right side edge.

[0168] In specific implementation, the deviation cost of the target fabric on the right side of the fabric tensioning axis in the cutting preparation state is determined according to the distribution ratios of all different gray values, that is: the distribution ratios of all different gray values are used as input parameters to be input into the information entropy model for calculation, and the output result is used as the deviation cost of the target fabric on the right side of the fabric tensioning axis in the cutting preparation state. In addition, other entropy models can also be used for processing, which is not limited here.

[0169] It should be noted that in this application, the deviation cost of the right side is used to measure the degree of deviation of the target fabric on the right side of the fabric tensioning axis. That is, the greater the deviation cost of the right side, the greater the degree of deviation of the target fabric on the right side of the fabric tensioning axis, and the smaller the deviation cost of the right side, the smaller the degree of deviation of the target fabric on the right side of the fabric tensioning axis.

[0170] In some embodiments, to determine the second interaction instruction for human-computer interaction in intelligent cutting of the target fabric according to the deviation cost of the left side and the deviation cost of the right side, the following steps can be specifically adopted, that is:

[0171] Calculate the ratio of the deviation cost of the left side to the deviation cost of the right side to obtain the deviation ratio of the target fabric;

[0172] Transmit the deviation ratio to the artificial intelligence control system of the fabric automatic cutting and conveying device, and the artificial intelligence control system automatically generates the second interaction instruction for human-computer interaction when the target fabric is intelligently cut.

[0173] It should be noted that in this embodiment, the deviation ratio is used to measure the deviation amount of the target fabric at the fabric tensioning axis. That is, the greater the deviation ratio, the greater the deviation amount of the target fabric at the fabric tensioning axis, and the smaller the deviation ratio, the smaller the deviation amount of the target fabric at the fabric tensioning axis.

[0174] It should also be noted that in this application, the second interaction instruction represents the second status setting instruction for the user to interact with the computer. The second interaction instruction is used to set the cutting action status of the target fabric. During the cutting of the target fabric, due to the influence of the body receiving characteristics of the fabric and the preparatory state of the cutting device, even after adjusting the surface tension of the fabric, there will still be a situation of fabric deviation. Therefore, by generating the second interaction instruction through the deviation ratio, the cutting state of the fabric can be effectively adjusted to ensure that the cutting device can accurately proceed as expected.

[0175] In step 105, in response to the operation of the second interaction instruction, control the cutting execution mechanism to automatically cut the target fabric, and the user control interface of the fabric automatic cutting and conveying device displays the cutting result.

[0176] In some embodiments, in response to the operation of the second interaction instruction, to control the cutting execution mechanism to automatically cut the target fabric, and the user control interface of the fabric automatic cutting and conveying device displays the cutting result, the following steps can be specifically adopted, that is:

[0177] When the computer receives the second interaction instruction, control the cutting execution mechanism to automatically cut the target fabric according to the preset cutting trajectory, and finally the user control interface of the fabric automatic cutting and conveying device displays the cutting result.

[0178] It should be noted that in this embodiment, the cutting execution mechanism represents the key components responsible for performing the cutting operation, including a series of mechanical and electrical components. Through the collaborative work of the cutting execution mechanism, the precise cutting of the target fabric can be effectively achieved. In addition, in this embodiment, the user interface represents an interactive interface for displaying the results and relevant information of the cutting operation.

[0179] It should also be noted that in this embodiment, the preset cutting trajectory represents the cutting path of the target fabric set in advance, which can be specifically set according to actual needs and will not be elaborated here.

[0180] In addition, on the other hand of this application, in some embodiments, this application provides an artificial intelligence control system for a fabric automatic cutting and conveying device. The artificial intelligence control system of the fabric automatic cutting and conveying device includes a cutting control unit. Refer to Figure 4 , this figure is a schematic diagram of the exemplary hardware and / or software of the cutting control unit shown according to some embodiments of this application. The cutting control unit 200 includes: a collection module 201, a processing module 202, and an execution module 203, which are described as follows:

[0181] Collection module 201. In this application, the collection module 201 is mainly used to respond to the operation of the fabric production instruction, collect the intelligent monitoring images of the target fabric in the cutting area, and determine the cutting execution block where the target fabric is located in the cutting area based on the intelligent monitoring images.

[0182] Processing module 202. In this application, the processing module 202 is mainly used to determine the multi-corner shape characteristics and wrinkle shape characteristics of the target fabric in the intelligent cutting state according to the cutting execution block, and then determine the fabric loose signal of the target fabric in the intelligent cutting state through the multi-corner shape characteristics and the wrinkle shape characteristics.

[0183] The processing module 202 is also used to obtain the surface flattening data of the target fabric when adjusting the surface tension based on the fabric loose signal, determine all the characteristic corner points of the target fabric after adjusting the surface tension, and then determine multiple intelligent cutting landing points of the target fabric according to all the characteristic corner points. The first interaction instruction of the human-machine interaction of the target fabric in intelligent cutting is determined through the surface flattening data and all the intelligent cutting landing points.

[0184] In addition, the processing module 202 is also used to respond to the operation of the first interaction instruction, set the cutting item in the cutting interaction control panel to the cutting preparation state, determine the deviation cost of the left side and the right side of the target fabric on the left side of the fabric tensioning axis in the cutting preparation state, and determine the second interaction instruction of the human-machine interaction of the target fabric in intelligent cutting according to the deviation cost of the left side and the deviation cost of the right side.

[0185] Execution module 203. In this application, the execution module 203 is mainly used to respond to the operation of the second interaction instruction, control the cutting execution mechanism to automatically cut the target fabric, and display the cutting result on the user control interface of the fabric automatic cutting and conveying device.

[0186] In addition, this application also provides a computer device, which includes a memory and a processor. The memory stores code, and the processor is configured to obtain the code and execute the artificial intelligence control method of the above-mentioned fabric automatic cutting and conveying device.

[0187] In some embodiments, refer to Figure 5 , this figure is a schematic structural diagram of a computer device applying the artificial intelligence control method of the fabric automatic cutting and conveying device according to some embodiments of this application. The artificial intelligence control method of the fabric automatic cutting and conveying device in the above embodiments can be implemented by Figure 5 the computer device shown. The computer device 300 includes at least one processor 301, a communication bus 302, a memory 303, and at least one communication interface 304.

[0188] The processor 301 can be a general-purpose central processing unit (CPU), or an application-specific integrated circuit (ASIC), or one or more for controlling the execution of the artificial intelligence control method of the fabric automatic cutting and conveying device in this application.

[0189] The communication bus 302 can be used to transmit information between the above components.

[0190] The memory 303 can be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, a random access memory (RAM), or other types of dynamic storage devices that can store information and instructions. It can also be an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM), or other optical disc storage, optical disc storage (including compact discs, laser discs, optical discs, digital versatile discs, Blu-ray discs, etc.), a magnetic disk, or other magnetic storage devices, or any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory 303 can exist independently and be connected to the processor 301 through the communication bus 302. The memory 303 can also be integrated with the processor 301.

[0191] Among them, the memory 303 is used to store the program code for executing the solution of this application and is controlled by the processor 301 to execute. The processor 301 is used to execute the program code stored in the memory 303. The program code can include one or more software modules. The determination of the artificial intelligence control method of the fabric automatic cutting and conveying device in the above embodiment can be implemented by one or more software modules in the program code in the processor 301 and the memory 303.

[0192] The communication interface 304 uses any device such as a transceiver to communicate with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area networks (WLAN), etc.

[0193] In a specific implementation, as an embodiment, the computer device can include multiple processors, and each of these processors can be a single-core (single-CPU) processor or a multi-core (multi-CPU) processor. Here, the processor can refer to one or more devices, circuits, and / or processing cores for processing data (such as computer program instructions).

[0194] The computer device described above may be a general-purpose computer device or a special-purpose computer device. In a specific implementation, the computer device may be a desktop computer, a laptop computer, a network server, a personal digital assistant (PDA), a mobile phone, a tablet computer, a wireless terminal device, a communication device, or an embedded device. The embodiments of the present application do not limit the type of the computer device.

[0195] In addition, the present application further provides a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, the artificial intelligence control method of the automatic cloth cutting and conveying device described above is implemented.

[0196] Although the preferred embodiments of the present application have been described, those skilled in the art can make additional changes and modifications once they learn the basic creative concept. Therefore, the appended claims are intended to be construed to include the preferred embodiments as well as all changes and modifications falling within the scope of the present application.

[0197] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application is also intended to include these modifications and variations.

Claims

1. An artificial intelligence control method for an automatic cutting and conveying device of fabric, characterized in that, The steps include: In response to the operation of the fabric production instruction, collect the intelligent monitoring image of the target fabric in the cutting area, and determine the cutting execution block where the target fabric is located in the cutting area based on the intelligent monitoring image; Determine the multi-corner shape feature and the wrinkle shape feature of the target fabric in the intelligent cutting state according to the cutting execution block, and further determine the fabric loosening signal of the target fabric in the intelligent cutting state through the multi-corner shape feature and the wrinkle shape feature; Obtain the surface flattening data of the target fabric when adjusting the surface tension based on the fabric loosening signal, determine all the feature corner points of the target fabric after adjusting the surface tension, and then determine multiple intelligent cutting drop points of the target fabric according to all the feature corner points. Determine the first interaction instruction for human-computer interaction of the target fabric in intelligent cutting through the surface flattening data and all the intelligent cutting drop points; In response to the operation of the first interaction instruction, set the cutting item in the cutting interaction control panel to the cutting preparation state, determine the deviation cost of the left side and the deviation cost of the right side of the target fabric at the left side of the fabric tensioning axis in the cutting preparation state, and determine the second interaction instruction for human-computer interaction of the target fabric in intelligent cutting according to the deviation cost of the left side and the deviation cost of the right side; In response to the operation of the second interaction instruction, control the cutting execution mechanism to automatically cut the target fabric, and display the cutting result on the user control interface of the fabric automatic cutting and conveying device.

2. The method according to claim 1, characterized in that Determining the cutting execution block where the target fabric is located in the cutting area based on the intelligent monitoring image specifically includes: Obtain the preset matching sample of the target fabric in the cutting area; Perform sliding matching on the intelligent monitoring image by the preset matching sample to obtain multiple sliding matching areas, and determine the matching degree of each sliding matching area; Extract the sliding matching area corresponding to the maximum matching degree, and use the sliding matching area as the cutting execution block where the target fabric is located in the cutting area.

3. The method according to claim 1, characterized in that, Determining the multi-corner shape feature and the wrinkle shape feature of the target fabric in the intelligent cutting state according to the cutting execution block specifically includes: Extract the cutting execution edge of the cutting execution block; Perform multi-corner point fitting on the cutting execution edge to obtain the multi-corner shape feature of the target fabric in the intelligent cutting state; Perform local feature description on each pixel point in the cutting execution block to obtain the feature descriptor corresponding to each pixel point; Determine the texture feature map of the target fabric from the feature descriptors of each pixel point; Determine the wrinkle shape feature of the target fabric in the intelligent cutting state according to the texture feature map.

4. The method according to claim 3, wherein Performing local feature description on each pixel point in the cutting execution block to obtain the feature descriptor corresponding to each pixel point specifically includes: Select a pixel point in the cutting execution block as the selected pixel point, and obtain all the adjacent pixel points of the selected pixel point; Compare and describe the pixel value of the selected pixel point with the pixel values of each adjacent pixel point respectively to obtain the feature descriptor corresponding to the selected pixel point; Continue to determine the feature descriptors corresponding to the remaining pixel points in the cutting execution block.

5. The method according to claim 1, characterized in that, Determining the fabric looseness signal of the target fabric in the intelligent cutting state through the multi-angle morphological feature and the fold morphological feature specifically includes: Obtaining a preset multi-angle morphology influence factor and a fold morphology influence factor; Determining the fabric looseness of the target fabric according to the multi-angle morphology influence factor, the fold morphology influence factor, the multi-angle morphological feature and the fold morphological feature; Generating a fabric looseness signal of the target fabric in the intelligent cutting state through the fabric looseness.

6. The method according to claim 1, wherein Determining all the characteristic corner points of the target fabric after surface tension adjustment specifically includes: Collecting a fabric monitoring image of the target fabric after surface tension adjustment; Selecting a pixel point from the fabric monitoring image as the selected pixel point, and determining the horizontal gray level change amount and the vertical gray level change amount of the selected pixel point; Constructing a gray level autocorrelation matrix of the selected pixel point according to the horizontal gray level change amount and the vertical gray level change amount; Determining the corner point response degree corresponding to the selected pixel point according to the gray level autocorrelation matrix; Continuing to determine the corner point response degrees corresponding to the remaining pixel points in the fabric monitoring image; Comparing the corner point response degree corresponding to each pixel point with a preset corner point response degree, extracting all the corner point response degrees greater than the preset corner point response degree, and then taking the pixel points corresponding to the extracted corner point response degrees as the characteristic corner points of the target fabric after surface tension adjustment.

7. The method according to claim 1, wherein Collecting an intelligent monitoring image of the target fabric in the cutting area through a camera installed on the fabric automatic cutting and conveying device.

8. An artificial intelligence control system for an automatic cloth cutting and conveying device, the artificial intelligence system of the automatic cloth cutting and conveying device includes a cutting control unit, characterized in that, The cutting control unit includes: A collection module, configured to respond to an operation of a fabric production instruction, collect an intelligent monitoring image of the target fabric in the cutting area, and determine a cutting execution block where the target fabric is located in the cutting area based on the intelligent monitoring image; A processing module, configured to determine the multi-angle morphological feature and the fold morphological feature of the target fabric in the intelligent cutting state according to the cutting execution block, and then determine the fabric looseness signal of the target fabric in the intelligent cutting state through the multi-angle morphological feature and the fold morphological feature; The processing module is further configured to obtain surface flattening data of the target fabric during surface tension adjustment based on the fabric looseness signal, determine all the characteristic corner points of the target fabric after surface tension adjustment, and then determine multiple intelligent cutting landing points of the target fabric according to all the characteristic corner points, and determine a first interaction instruction for human-machine interaction of the target fabric in intelligent cutting through the surface flattening data and all the intelligent cutting landing points; The processing module is further configured to respond to an operation of the first interaction instruction, set the cutting item in the cutting interaction control panel to a cutting preparation state, determine the deviation cost of the target fabric on the left side of the fabric tensioning axis and the deviation cost on the right side in the cutting preparation state, and determine a second interaction instruction for human-machine interaction of the target fabric in intelligent cutting according to the deviation cost on the left side and the deviation cost on the right side; An execution module, configured to respond to an operation of the second interaction instruction, control a cutting execution mechanism to automatically cut the target fabric, and display a cutting result on a user control interface of the fabric automatic cutting and conveying device.

9. A computer device, characterized in that, The computer device includes a memory and a processor. The memory stores code, and the processor is configured to obtain the code and execute the artificial intelligence control method of the automatic cloth cutting and conveying device according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the artificial intelligence control method of the automatic cloth cutting and conveying device according to any one of claims 1 to 7.

Images