Textile production positioning identification system based on visual inspection

Abstract

The invention relates to the field of textile production recognition, in particular to a textile production positioning recognition system based on visual inspection, which comprises a texture recognition end, an image detection end, a cutting position generation end, a first execution end and a knitting positioning end. The method comprises the following steps: processing lines for detecting a target detection image to obtain an image texture value of a unit element, then combining the image texture value with an image with a standard line to obtain a target cutting position, then identifying a cutting piece, generating a region image, marking a standard stitching point in the region image to obtain a region marking image, and finally marking the region marking image to obtain the target cutting position. And overlapping and processing all the area marking images to obtain a target distance value, and obtaining a sewing positioning point according to the target distance value, so that the overall spinning point of the textile is determined according to sewing positioning identification, and the production quality of the textile in the production process is kept consistent.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of textile production identification, and particularly relates to a textile production positioning identification system based on visual detection. BACKGROUND

[0002] Textile originally refers to the general term of spinning and weaving, but with the continuous development and improvement of the textile knowledge system and discipline system, especially after the emergence of non-woven textile materials and three-dimensional composite weaving technology, textile is not only traditional spinning and weaving.

[0003] The prior art CN115131353A discloses a flat screen printing textile production abnormality identification positioning method and system, which comprises the following steps: collecting textile data by using an electronic device; clustering the data to obtain different color printing areas; dividing the printing areas, obtaining the gray scale change characteristics according to the gray scale values in each area, and then obtaining the gray scale change similarity between the printing data; obtaining defect-free and defective printing data according to the gray scale change similarity; obtaining the illumination influence degree of different areas according to the gray scale difference rate of each area in the defect-free printing data; performing illumination compensation on the printing data according to the illumination influence degree to obtain the printing data without illumination influence; and identifying and positioning the abnormal production problem according to the gray scale of each area in the defective printing data without illumination influence. The above method is used for positioning the abnormal production problem of textiles, and the accuracy of production problem positioning can be improved by the above method.

[0004] However, there are the following problems: the texture of the textile product will deform after a series of process, but the influence range of the texture deformation on the textile product is within a controllable range, but due to different texture deformations, the same batch of products produced in the manufacturing process may have differences, and at this time, quality detection is performed on the batch of products, which increases the quality difference of the batch of products. SUMMARY

[0005] The purpose of the present application is to solve the problems in the background art, and the textile production positioning identification system based on visual detection is proposed.

[0006] In order to achieve the above purpose, the present application adopts the following technical scheme:

[0007] The textile production positioning identification system based on visual detection comprises a texture identification end, an image detection end, a cutting position generation end, a first execution end and a knitting positioning end.

[0008] The texture identification end is used for identifying the textile in the working area by using a visual detection device, and generating a target detection image which is transmitted to the image detection end.

[0009] The image detection end is used for detecting a target detection image, setting a unit element according to a texture of the textile, and searching the unit element in a textile database to obtain target information, then identifying a grain in the unit element and performing size detection on the grain to obtain a size related to the grain, then calculating the size of the grain, and obtaining an image texture value of the unit element according to a calculation result, and the image detection end transmits the image texture value to the cutting position generation end;

[0010] The cutting position generation end is used for setting a plane coordinate system for a standard grain in the target information, extracting a standard cutting path, obtaining a standard starting point and a coordinate of the standard starting point in the standard cutting path, then coinciding a center point of the standard grain with a center point of the standard texture, processing a coordinate position of the standard starting point to obtain a target cutting position, taking the target cutting position as a starting point, and obtaining a cutting positioning line according to the standard cutting path, and the cutting position generation end transmits the cutting positioning line to the first execution end.

[0011] The first execution end is used for cutting the textile according to the cutting positioning line to obtain a cutting piece, then a visual detection device performs visual identification on the cutting piece to obtain a region image, and the first execution end transmits the region image to the knitting positioning end.

[0012] The knitting positioning end is used for marking the standard stitching point in the region image according to the region image to obtain a region marking image, then coinciding all the region marking images and comparing the standard stitching point in an X-axis direction and a Y-axis direction to obtain a target distance value, then obtaining a region marking image corresponding to the target distance value, and obtaining a stitching positioning point according to distances between the region marking images.

[0013] As a further scheme of the present application, the image texture value obtaining method is as follows:

[0014] The unit element in the target detection object is extracted, the unit element is taken as a matching object, a search is performed in a textile database, corresponding standard information is obtained according to a search result, and the standard information is marked as target information;

[0015] Image analysis is performed on the unit element, a grain in the unit element is identified, and size measurement is performed on the grain, the size of the grain is marked as CHi, i=1, 2, …, I, indicating that there are I textures in the unit element, meanwhile, angles between adjacent textures are obtained, and the adjacent texture angles are marked as Ag, g=1, 2, …, G, indicating that there are G adjacent texture angles in the unit element.

[0016] The image texture value WL of the unit element is obtained, CBi represents the size of the texture i in the target information, ABg represents the angle of the angle g in the target information, and alpha1 and alpha2 are a size coefficient and an angle coefficient respectively. The image texture value WL of the unit element is obtained, CBi represents the size of the texture i in the target information, ABg represents the angle of the angle g in the target information, and alpha1 and alpha2 are a size coefficient and an angle coefficient respectively.

[0017] As a further scheme of the present application, the standard information is stored in a textile database, the textile database is provided with a textile storage unit and an image storage unit, the textile storage unit is used for storing textile information, the textile information refers to the standard information and the standard cutting path as well as the pattern and texture shape in the textile production, the standard information includes the size of the texture shape, the included angle between adjacent textures, the standard pattern and the standard cutting path, and the image storage unit is used for storing the target detection image and the area image.

[0018] As a further scheme of the present application, the target detection image and the area image stored in the image storage unit are temporarily stored, and the target detection image and the area image of the same batch of textiles will be deleted after the production of the batch of textiles is completed.

[0019] As a further scheme of the present application, the method for obtaining the cutting positioning line is as follows:

[0020] A plane coordinate system is set for the image of the standard pattern in the reference information;

[0021] The standard cutting path and the starting point of the cutting position in the standard cutting path are extracted, the starting point is marked as a standard starting point, and the position coordinates BQ(x1, y1) of the standard starting point in the standard pattern are obtained;

[0022] The center point of the unit element is obtained, the center point of the unit element is coincided with the center point of the standard pattern in the plane coordinate system, and then the unit element is placed in the plane coordinate system, and the formula The target cutting position M(x2, y2) at this time is obtained,

[0023] Then, taking the target cutting position M as the starting point, a cutting positioning line suitable for the real-time target detection image is generated for the target detection image according to the standard cutting path.

[0024] As a further scheme of the present application, the method for obtaining the stitching positioning point is as follows:

[0025] All the area images are obtained, the standard stitching points are marked in each area image respectively, and the area marking images are generated;

[0026] Detection points are set in the area marking images, all the area marking images are image-coincided, a reference area is set in the coincided image, and the standard stitching points are marked in the reference area to obtain reference stitching points;

[0027] Optionally, a detection point is taken as a target point, the distance between the target point and the corresponding reference stitching point in the region marking image is obtained, the distances between the target points in all target marking regions are sequentially measured, the maximum distance value is obtained, and the maximum distance value is marked as a target distance value;

[0028] The midpoint of the region marking image corresponding to the target distance value is connected to obtain a center line, and the midpoint of the center line is taken as a standard stitching point for correspondence to obtain a stitching positioning point.

[0029] As a further scheme of the present application, the target distance value includes an X-axis distance value and a Y-axis distance value, that is, when the distances between the target points in two adjacent region marking images are measured, the distances between the target points on the X-axis and the distances between the target points on the Y-axis are measured respectively, and the X-axis distance value and the Y-axis distance value are obtained;

[0030] The midpoints of the standard stitching points in the region marking images corresponding to the X-axis distance value and the Y-axis distance value are obtained respectively, the two midpoints are connected to obtain a center line, the midpoint of the center line is taken, and the plane coordinate position of the midpoint is obtained, and then the midpoint is taken as the midpoint of the standard stitching point, and the standard stitching point is corresponded to obtain a stitching positioning point.

[0031] As a further scheme of the present application, when the region marking images corresponding to the X-axis distance value and the Y-axis distance value are the same, the center of the reference stitching point and the midpoint of the region marking image are connected to obtain a center line, and the midpoint of the center line is taken as the midpoint of the standard stitching point, and the standard stitching point is corresponded to obtain a stitching positioning point.

[0032] As a further scheme of the present application, a second execution end is further included, which is used for receiving the stitching positioning point and sewing the cutting piece according to the stitching positioning point.

[0033] Compared with the prior art, the present application has the following advantages:

[0034] In the present application, visual detection devices are arranged in the cutting process and the sewing process of the textile, the textile in the working region is identified to obtain a target detection image, the texture of a unit element in the target detection image is matched with the texture in the standard information for matching and processing to obtain an image texture value of the unit element, the center point of the unit element is overlapped with the center point of the standard texture for processing to obtain a target cutting position, a cutting positioning line of the target detection image is obtained according to a standard cutting path, and the textile is cut according to the cutting positioning line, so that the uniform cutting positioning line is selected according to the texture of the target detection image in the cutting process of the textile, and the production error of the textile is further reduced.

[0035] The application also acquires regional images of all cutting pieces by positioning the sewing points in the textile process, superimposes all the regional images, processes the superimposed regional images to obtain sewing positioning points, and sews the cutting pieces according to the sewing positioning points, so as to determine the overall textile points of the textile according to the sewing positioning recognition, and further make the production quality of the textile consistent in the production process. BRIEF DESCRIPTION OF DRAWINGS

[0036] Figure 1 It is a schematic diagram of the system structure of the application. DETAILED DESCRIPTION

[0037] The technical solutions in the embodiments of the application will be clearly and completely described below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, rather than all the embodiments of the application.

[0038] Referring to Figure 1 The textile production positioning recognition system based on visual detection comprises a textile database, a texture recognition end, an image detection end, a cutting position generation end, a first execution end, a knitting positioning end and a second execution end.

[0039] The textile database is provided with a textile storage unit and an image storage unit. The textile storage unit is used for storing textile information. The textile information refers to the texture and the texture shape existing in the textile production process and the standard information and the standard cutting path corresponding to each texture shape. The standard information comprises the size of the texture shape and the included angle between adjacent textures in the texture shape. The standard texture is set according to the area threshold value. When the textile is processed, the textile needs to be cut to obtain cutting pieces. The contour shape of the obtained cutting pieces is taken as the standard cutting path. The image storage unit is used for storing target detection images and regional images. It needs to be further explained that the target detection images and the regional images stored in the image storage unit are temporarily stored. When the textile of the same batch is produced, the target detection images and the regional images of this batch will be deleted to improve the storage space of the image storage unit.

[0040] The textile database is bidirectionally electrically connected with the image detection end.

[0041] The texture recognition end recognizes the texture of the textile based on the textile in the processing process and generates a target detection image. Then the texture recognition end transmits the target detection image to the image detection end. The visual detection equipment used when the textile is recognized here is an AI machine recognition equipment. In this embodiment, the visual detection equipment used is a laser visual detection equipment.

[0042] In another embodiment of the present invention, a recognition area is provided in the textile equipment. The textile is placed in the recognition area, and then the visual inspection device of the texture recognition end recognizes the textile in the working area and generates a target detection image. It should be further noted that when the textile is recognized, the textile is not necessarily completely placed in the recognition area. For example, when the unfolded area of ​​the textile is larger than the recognition area, only a part of the area is placed in the recognition area. In this case, the visual inspection device only performs visual recognition on the part placed in the recognition area and generates a target detection image for the recognized part. The position not placed in the recognition area is not visually recognized.

[0043] The image detection unit detects textures in the received target detection image and generates image texture values ​​based on the detection results. The specific method for obtaining these image texture values ​​is as follows:

[0044] S1: Obtain the target detection image, extract the unit elements in the target detection image, then use the unit elements as matching objects to search in the textile database, obtain the corresponding standard information based on the search results, and mark them as benchmark information;

[0045] Specifically, the texture of textiles is formed by the interweaving of warp and weft threads according to certain rules. Since the weaving rules of the same textile remain unchanged during the weaving process, the texture of the textile in the same target detection area is cyclical. At this time, a certain area is extracted and marked as a unit element. It should be further noted that the specific value of the area in the unit element is set by those skilled in the art, and the unit element is a pattern formed by several cyclical textures. At the same time, the area of ​​the unit element is the same as the area of ​​the standard pattern. Furthermore, in this embodiment, when selecting the area of ​​the unit element, it is based on the image of the standard pattern, that is, the image of the standard pattern and the image of the unit element are the positions with the highest image similarity in the same marked area. The position of the marked area is set by those skilled in the art.

[0046] S2: Perform image analysis on the extracted unit elements. Image analysis includes size detection and angle measurement. The specific methods of image analysis are as follows:

[0047] Identify the textures within a unit element and measure their dimensions. Mark the obtained texture dimensions as CHi, i = 1, 2, ..., I, indicating that there are I textures in the unit element. At the same time, obtain the angle between adjacent textures and mark the angle between adjacent textures as Ag, g = 1, 2, ..., G, indicating that there are G adjacent texture angles in the unit element.

[0048] S3: Using the formula The image texture value WL of the unit element is obtained, where CBi represents the size of texture i in the benchmark information, ABg represents the angle g of the benchmark information, and α1 and α2 are the size coefficient and angle coefficient, respectively.

[0049] The image detection end then transmits the image texture value to the clipping generation end;

[0050] The cropping generation end is used to comprehensively analyze the image texture values ​​and generate cropping positioning lines based on the analysis results. The cropping generation end then transmits the cropping positioning lines to the first execution end. The specific method for generating the cropping positioning lines is as follows:

[0051] Obtain the image of the standard texture in the benchmark information, and at the same time set up a plane coordinate system to obtain the position of the point on the standard texture in the plane coordinate system;

[0052] Next, extract the standard clipping path and the starting point of the clipping position in the standard clipping path, mark it as the standard starting point, obtain the position coordinates of the standard starting point in the standard texture, and mark it as BQ(x1, y1).

[0053] Then, obtain the center point of the unit element, align the center point of the unit element with the center point of the standard texture in the planar coordinate system, and then place the unit element in the planar coordinate system using the formula. The target clipping position M(x2, y2) is obtained at this point.

[0054] Then, starting from the target cropping position M, and following the standard cropping path, a new cropping positioning line adapted to the real-time target detection image is generated for the target detection image.

[0055] The first execution end is used to cut the textile according to the cutting positioning line. After the first execution end finishes its operation, it obtains several cut pieces. Based on the vision inspection device, it acquires the image of each cut piece and marks it as a region image. Then the first execution end transmits the region image to the knitting positioning generation end.

[0056] The knitting positioning generator receives the region image and performs integrated analysis on it. Based on the results of the integrated analysis, it generates stitching positioning points. The knitting positioning generator then transmits these stitching positioning points to the second execution unit. The specific method for obtaining the stitching positioning points is as follows:

[0057] All generated region images are acquired. First, standard stitching points are marked in each region image, and region marking images are generated. The standard stitching points are set by those skilled in the art and stored in the textile storage unit.

[0058] Detection points are set in the region-marked image, where the detection points are any points selected on the standard suture points. In this embodiment, 10 detection points are set, and the detection points are evenly distributed in the standard suture points.

[0059] Then, all the region-marked images are superimposed, and a reference region is set in the superimposed image. The standard suture point is marked in the reference region to obtain the reference suture point. At this time, after the images are superimposed, any detection point is selected as the target point, and the distance between the target point in the region-marked image and the corresponding reference suture point is obtained. The distance between the target points in all target-marked regions is measured in turn, the maximum distance value is obtained, and it is marked as the target distance value.

[0060] It should be further explained that the target distance value here includes the X-axis distance value and the Y-axis distance value. That is, when measuring the distance between target points in two adjacent region labeled images, the distance between the target points on the X-axis and the distance on the Y-axis are measured respectively, and the X-axis distance value and the Y-axis distance value are obtained at this time.

[0061] The midpoints of the standard suture points in the region marker images corresponding to the X-axis distance values ​​and the Y-axis distance values ​​are obtained respectively. These two midpoints are connected to form a midline. The midpoint of the midline is then taken, and its planar coordinates are obtained. This midpoint is then used as the midpoint of the standard suture point, and the standard suture point is matched to obtain the suture positioning point. It should be further noted that when the region marker images corresponding to the X-axis distance values ​​and the Y-axis distance values ​​are the same, the center of the reference suture point is connected to the midpoint of the region marker image to obtain a midline. The midpoint of this midline is then used as the midpoint of the standard suture point, and the suture positioning point is matched to the standard suture point.

[0062] The second execution end is used to receive the sewing positioning point and sew the cut pieces according to the sewing positioning point. In this way, the overall weaving point of the textile is determined according to the sewing positioning identification, thereby ensuring that the production quality of the textile remains consistent throughout the production process.

[0063] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A textile production positioning and identification system based on visual inspection, characterized in that, Includes a texture recognition end, an image detection end, a cutting position generation end, a first execution end, and a knitting positioning end; The texture recognition end is used to identify textiles in the work area using visual inspection equipment, generate target detection images, and transmit them to the image detection end; The image detection end is used to detect the target image. First, based on the texture of the textile, unit elements are set and the unit elements are retrieved from the textile database to obtain the benchmark information. Then, the texture in the unit elements is identified and the size of the texture is detected to obtain the size related to the texture. After that, the size of the texture is calculated and the image texture value of the unit element is obtained based on the calculation result. The image detection end transmits the image texture value to the cutting position generation end. The cropping generator is used to set a planar coordinate system for the standard texture based on the image of the standard texture in the benchmark information, extract the standard cropping path, obtain the standard starting point and its coordinates in the standard cropping path, then make the center point of the standard texture coincide with the center point of the unit element, process the coordinate position of the standard starting point to obtain the target cropping position, and use the target cropping position as the starting point to obtain the cropping positioning line according to the standard cropping path. The cropping generator transmits the cropping positioning line to the first execution end. The first execution end is used to cut the textile according to the cutting positioning line to obtain the cut piece. Then, the visual inspection device performs visual recognition on the cut piece to obtain the area image. The first execution end transmits the area image to the knitting positioning end. The knitting positioning end is used to mark the standard seam points in the area image based on the area image to obtain the area mark image. Then, all the area mark images are superimposed and compared with the standard seam points in the X-axis and Y-axis directions to obtain the target distance value. Then, the area mark image corresponding to the target distance value is obtained, and the seam positioning point is obtained based on the distance between the area mark images.

2. The textile production positioning and identification system based on visual detection according to claim 1, characterized in that, The method for obtaining image texture values ​​is as follows: Extract unit elements from the target detection object, use the unit elements as matching objects, search in the textile database, obtain the corresponding standard information based on the search results, and mark them as benchmarking information; Image analysis is performed on the unit element to identify the texture within the unit element and measure the size of the texture. The obtained texture size is marked as CHi, i = 1, 2, ..., I, indicating that there are I textures in the unit element. At the same time, the included angle between adjacent textures is obtained and marked as Ag, g = 1, 2, ..., G, indicating that there are G adjacent texture included angles in the unit element. use The image texture value WL of the unit element is obtained, where CBi represents the size of texture i in the benchmark information, ABg represents the angle g in the benchmark information, and α1 and α2 are the size coefficient and angle coefficient, respectively.

3. The textile production positioning and identification system based on visual detection according to claim 1, characterized in that, Standard information is stored in a textile database, which includes textile storage units and image storage units. The textile storage units are used to store textile information, which refers to the texture and shape of textile production, as well as standard information and standard cutting paths. Standard information includes the size of the texture shape, the angle between adjacent textures, standard textures, and standard cutting paths. The image storage units are used to store target detection images and region images.

4. The textile production positioning and identification system based on visual detection according to claim 3, characterized in that, The target detection images and region images stored in the image storage unit are temporary. Once the same batch of textiles is produced, the target detection images and region images for that batch will be deleted.

5. The textile production positioning and identification system based on visual detection according to claim 2, characterized in that, The method for obtaining the trimming positioning lines is as follows: Set the image of the standard texture in the benchmarking information to a plane coordinate system; Extract the standard clipping path and the starting point of the clipping position in the standard clipping path, mark the starting point as the standard starting point, and obtain the position coordinates BQ(x1, y1) of the standard starting point in the standard texture; Obtain the center point of the unit element, align the center point of the unit element with the center point of the standard texture in the planar coordinate system, and then place the unit element in the planar coordinate system using the formula. The target clipping position M(x2, y2) is obtained at this point. Then, starting from the target cropping position M, and following the standard cropping path, a new cropping positioning line is generated for the target detection image to adapt to the real-time target detection image.

6. The textile production positioning and identification system based on visual detection according to claim 1, characterized in that, The method for obtaining the suture positioning points is as follows: Acquire all region images, mark the standard suture points in each region image, and generate region-marked images; Detection points are set in the region-marked images, all region-marked images are superimposed, a reference region is set in the superimposed image, and the standard suture point is marked in the reference region to obtain the reference suture point; Choose any detection point as the target point, obtain the distance between the target point and the corresponding reference stitching point in the region labeling image, measure the distance between the target points in all target labeling regions in turn, obtain the maximum distance value, and mark it as the target distance value; Connect the midpoints of the region marker images corresponding to the target distance values ​​to obtain the midline. Take the midpoint of the midline as the standard suture point to obtain the suture positioning point.

7. The textile production positioning and identification system based on visual detection according to claim 6, characterized in that, The target distance value includes the X-axis distance value and the Y-axis distance value. That is, when measuring the distance between target points in two adjacent region-marked images, the distance between the target points on the X-axis and the distance on the Y-axis are measured respectively, and the X-axis distance value and the Y-axis distance value are obtained at this time. The midpoints of the standard suture points in the region marker image corresponding to the X-axis distance value and the midpoints of the region marker image corresponding to the Y-axis distance value are obtained respectively. The two midpoints are connected to obtain the midline. Then, the midpoint of the midline is taken and the planar coordinate position of the midpoint is obtained. Then, this midpoint is used as the midpoint of the standard suture point, and the standard suture point is matched to obtain the suture positioning point.

8. The textile production positioning and identification system based on visual detection according to claim 7, characterized in that, When the X-axis distance value and the Y-axis distance value correspond to the same region marker image, the center of the reference suture point is connected to the midpoint of the region marker image to obtain the midline. The midpoint of the midline is then taken as the midpoint of the standard suture point and compared with the standard suture point to obtain the suture positioning point.

9. The textile production positioning and identification system based on visual detection according to claim 1, characterized in that, It also includes a second execution end for receiving the sewing positioning point and sewing the cut pieces according to the sewing positioning point.

Images