Short fiber anomaly detection method and system based on machine learning

Abstract

The invention provides a short fiber anomaly detection method and system based on machine learning, and the method comprises the steps: carrying out the feature object extraction of a spinning strand image, and obtaining a tracing fiber region image and a background fiber region image of a spinning strand; performing connected region segmentation based on the tracer fiber region image to obtain a plurality of tracer fiber region sub-images, and extracting fiber radial breaking points corresponding to the tracer fiber region sub-images; hook form extraction is carried out on the tracer fiber area image, and a hook form image score of the tracer fiber area image is obtained; according to the hook form image score of the traced fiber area image, the drafting quality of spinning short fibers in the spinning process is detected, hook form image scoring can be carried out based on the hook form of the traced fiber image, and therefore quality detection of the short fiber drafting technology in spinning process drafting is achieved; and the accuracy of quality anomaly detection of the spun short fibers is improved.

Description

Technical Field

[0001] The present application relates to the technical field of fiber image detection, and more specifically, to a method and system for detecting short fiber anomalies based on machine learning. Background Art

[0002] Yarn structure refers to the arrangement of fibers in the yarn body, and the arrangement of fibers is closely related to the spinning process and fiber raw materials. In the traditional spinning process, the randomly arranged raw cotton fibers are arranged in a fiber arrangement after a series of processes such as opening, combing, drawing, roving, and spinning. The straightening and parallelism of the fibers in the spinning whiskers are significantly correlated with the quality of the yarn. The drafting process will cause the fibers to form a small number of forward hooks and straighten the rear hooks at the same time. Analyzing the shape and number of fiber hooks in the whiskers is important for evaluating the spinning drafting process.

[0003] In the existing spinning fiber detection method, the bending hook morphology of the fiber during the drawing process is mainly detected by manual visual inspection, that is, the image of the spinning whiskers during the spinning process is collected and amplified, and the process evaluation and abnormality detection are performed by manually observing the morphological characteristics of the fiber whiskers. This method not only consumes a large labor cost and takes a long time, but also relies on the experience of technicians, which greatly restricts the development of the spinning fiber drawing process. Summary of the invention

[0004] The present application provides a method and system for detecting short fiber anomalies based on machine learning, which can score the hook morphology image based on the hook morphology of the tracer fiber image, thereby realizing quality detection of the short fiber drafting process in the drafting process of the spinning process, and improving the accuracy of detecting quality anomalies of spinning short fibers.

[0005] In a first aspect, the present application provides a method for detecting short fiber anomalies based on machine learning, the method for detecting short fiber anomalies comprising the following steps: After the spinning process is drafted, the spinning short fiber whiskers output by the front roller are imaged to obtain the spinning whisker images; Extracting characteristic objects from the spinning whisker image to obtain a tracer fiber region image and a background fiber region image of the spinning whisker; Based on the tracer fiber region image, connected region segmentation is performed to obtain multiple tracer fiber region sub-graphs, and fracture features of each tracer fiber region sub-graph are extracted using a machine learning algorithm to obtain fiber radial fracture points corresponding to each tracer fiber region sub-graph; Acquire background fiber features of the background fiber area image, perform hook morphology extraction on the tracer fiber area image based on the background fiber features and radial fracture positions corresponding to radial fracture points of each fiber, and obtain a hook morphology image score of the tracer fiber area image; The drafting quality of the spinning staple fibers during the spinning process is detected according to the hook morphology image score of the tracer fiber region image.

[0006] In this embodiment, feature object extraction is performed on the spinning whisker image to obtain the tracer fiber region image and the background fiber region image of the spinning whisker, specifically including: Obtaining the color grade of the tracer fiber, and mapping to obtain the tracer threshold value according to the threshold interval where the color grade of the tracer fiber is located; The spinning whisker image is segmented according to the tracing threshold value to obtain a tracing fiber region image and a background fiber region image of the spinning whisker.

[0007] In this embodiment, obtaining the color grade of the tracer fiber specifically includes: using a standard sample of the tracer fiber and the background fiber to collect images, converting the sample image into a color space, and extracting the hue angle from the tracer fiber area as the color grade of the tracer fiber.

[0008] In this embodiment, performing connected region segmentation based on the tracer fiber region image to obtain multiple tracer fiber region subgraphs specifically includes: acquiring the tracer fiber region image, defining an image connectivity standard, performing connected region segmentation on the tracer fiber region image based on the image connectivity standard to obtain multiple tracer fiber region subgraphs.

[0009] In this embodiment, a machine learning algorithm is used to extract fracture features of each tracer fiber region sub-graph, and the fiber radial fracture points corresponding to each tracer fiber region sub-graph are obtained, which specifically include: Inputting a training data set into a support vector machine for machine learning training, wherein the training data set includes a standard tracer fiber region image and annotated breakpoints; For any tracer fiber region subgraph, an edge detection algorithm is used to extract the edge grayscale mutation position corresponding to the tracer fiber region subgraph, the shape attribute label of the connected region of the tracer fiber region subgraph is extracted, and a gray level co-occurrence matrix is ​​used to extract the local texture features of the tracer fiber region subgraph; The edge grayscale mutation position, the connected region shape attribute label and the local texture feature corresponding to the tracer fiber region sub-graph are input into the support vector machine, so as to extract the fiber radial break point of the tracer fiber region sub-graph.

[0010] In this embodiment, based on the background fiber features and the radial fracture positions corresponding to the radial fracture points of each fiber, the hook morphology of the tracer fiber region image is extracted, and the hook morphology image score of the tracer fiber region image is obtained, which specifically includes: According to the background fiber characteristics, hook morphology is extracted from each tracer fiber region sub-image in the tracer fiber region image to obtain initial hook morphology image scores corresponding to each tracer fiber region sub-image; Obtaining fiber radial fracture points corresponding to each tracer fiber region sub-image; determining radial fracture positions corresponding to each fiber radial fracture point based on the fiber radial fracture points corresponding to each tracer fiber region sub-image; According to the radial fracture positions corresponding to the radial fracture points of each fiber and the initial hook morphology image score, the radial interference degree is extracted to obtain the radial interference degree between the radial fracture positions corresponding to the radial fracture points of each fiber and the initial hook morphology image score; The initial hook morphology image scores corresponding to each tracer fiber region sub-image are fused and corrected according to the radial interference degree to obtain the hook morphology image score of the tracer fiber region image.

[0011] In this embodiment, the drafting quality of the spinning staple fibers during the spinning process is detected according to the hook morphology image score of the tracer fiber area image, which specifically includes: determining the corresponding drafting quality grade according to the scoring interval of the hook morphology image score of the tracer fiber area image.

[0012] In this embodiment, an industrial camera is used to capture images of the spinning staple fiber strands output by the front roller to obtain the spinning staple fiber strand images.

[0013] In this embodiment, before extracting feature objects from the spinning yarn image, the method further includes: performing image preprocessing on the spinning yarn image.

[0014] In a second aspect, the present application provides a short fiber anomaly detection system based on machine learning, which is used to perform a short fiber anomaly detection method based on machine learning, and the short fiber anomaly detection system includes: An image acquisition module is used to acquire images of the spinning short fiber whiskers output by the front roller after drafting in the spinning process to obtain images of the spinning whiskers; An image processing module is used to extract feature objects from the spinning whisker image to obtain a tracer fiber region image and a background fiber region image of the spinning whisker; The image processing module is further used to perform connected area segmentation based on the tracer fiber area image to obtain multiple tracer fiber area sub-graphs, and use a machine learning algorithm to extract fracture features of each tracer fiber area sub-graph to obtain fiber radial fracture points corresponding to each tracer fiber area sub-graph; An image scoring module is used to obtain background fiber features of the background fiber area image, and based on the background fiber features and radial fracture positions corresponding to radial fracture points of each fiber, perform hook morphology extraction on the tracer fiber area image to obtain a hook morphology image score of the tracer fiber area image; The fiber draft detection module is used to detect the draft quality of the spinning short fibers during the spinning process according to the hook morphology image score of the tracer fiber area image.

[0015] The technical solution provided by the embodiments disclosed in this application has the following beneficial effects: It can be seen that the present application utilizes the fact that the tracer fiber has a significant color characteristic difference from the background fiber, and extracts the tracer fiber separately through regional segmentation, so as to concentrate on analyzing the hook morphology of the tracer fiber, and extracts the fiber breaking point through a machine learning algorithm, so as to accurately obtain the specific position where the fiber breaks during the stretching process, marking the key point of the change in the stretching force. The position of the radial breaking point refines the unevenness of the fiber force, provides basic data for the subsequent hook morphology scoring, and quantifies the straightness and parallelism of the fiber through the hook morphology image scoring, and then corrects the initial hook morphology image scoring by analyzing the radial breaking point position, reducing the error caused by uneven force, thereby improving the robustness of the scoring, and finally, according to the hook morphology image scoring of the tracer fiber regional image, the drafting quality of the spinning staple fiber in the spinning process is detected, thereby avoiding subjective errors in manual detection and ensuring the consistency and reliability of the detection results.

[0016] In summary, the present application performs hook morphology image scoring based on the hook morphology of the tracer fiber image, thereby realizing quality detection of the short fiber drafting process in the drafting process of the spinning process, and improving the accuracy of abnormal quality detection of spinning short fibers. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only the embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0018] Figure 1 is a flow chart of a method for detecting short fiber anomalies based on machine learning provided in the present application; Figure 2 This is a module structure diagram of the short fiber anomaly detection system based on machine learning provided in this application. DETAILED DESCRIPTION

[0019] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0020] The embodiment of the present application provides a method and system for detecting short fiber anomalies based on machine learning, the core of which is to collect images of the spinning short fiber whiskers output by the front roller after the spinning process is drafted to obtain a spinning whisker image; extract feature objects from the spinning whisker image to obtain a tracer fiber area image and a background fiber area image of the spinning whiskers; perform connected area segmentation based on the tracer fiber area image to obtain multiple tracer fiber area subgraphs, and use a machine learning algorithm to extract fracture features from each tracer fiber area subgraph to obtain radial fiber fracture points corresponding to each tracer fiber area subgraph; The background fiber features of the background fiber area image are obtained, and based on the background fiber features and the radial fracture positions corresponding to the radial fracture points of each fiber, the hook morphology of the tracer fiber area image is extracted to obtain a hook morphology image score of the tracer fiber area image; according to the hook morphology image score of the tracer fiber area image, the drafting quality of the spinning staple fibers in the spinning process is detected. The present application can perform a hook morphology image score based on the hook morphology of the tracer fiber image, thereby realizing the quality detection of the staple fiber drafting process in the drafting of the spinning process, and improving the accuracy of the detection of abnormal quality of the spinning staple fibers.

[0021] Embodiment 1: In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods. Figure 1 As shown in the figure, this figure is an exemplary flow chart of a method for detecting short fiber anomalies based on machine learning according to this embodiment of the present application, and the method for detecting short fiber anomalies includes the following steps: In step S1, after the spinning process is drafted, the spinning staple fiber strands output by the front roller are imaged to obtain the spinning strand image.

[0022] It should be noted that the drafting of the spinning process is a key link in the spinning process, which aims to further refine and straighten the fiber whiskers through the drafting device to achieve the fineness and quality required for the final yarn. It is an important transition step from roving to yarn. In this application, the roller drafting system consists of front, middle and rear roller groups, and the fiber stretching is achieved through different rotation speeds. The spinning frame determines the drafting multiple by setting the speed ratio between the rollers, that is, the ratio of the input and output whisker thickness. The rear roller outputs roving at a slower speed, and the middle and front rollers are accelerated step by step, so that the fiber is stretched in the drafting zone. During the drafting process, the fiber is subjected to the combined action of control force and traction force, and gradually reaches a straight and parallel state.

[0023] In this embodiment, an industrial camera is used to capture images of the spinning staple fiber strands output by the front roller to obtain an image of the spinning strands; in some other embodiments, other devices or equipment capable of image capture may also be used, and this application does not limit this.

[0024] In this embodiment, a conveyor belt device is used to convey the spinning staple fiber strands output by the front roller, and an image of the spinning staple fiber strands is collected during the conveying process.

[0025] In step S2, feature object extraction is performed on the spinning whisker image to obtain a tracer fiber region image and a background fiber region image of the spinning whisker.

[0026] In this embodiment, before extracting feature objects from the spinning yarn image, the method further includes: performing image preprocessing on the spinning yarn image.

[0027] Preferably, in some embodiments, the image preprocessing of the spinning yarn image specifically includes: after grayscale processing of the spinning yarn image, using median filtering to perform noise filtering on the initial detection image.

[0028] In this embodiment, feature object extraction is performed on the spinning whisker image to obtain the tracer fiber region image and the background fiber region image of the spinning whisker, specifically including: Obtaining the color grade of the tracer fiber, and mapping to obtain the tracer threshold value according to the threshold interval where the color grade of the tracer fiber is located; The spinning whisker image is segmented according to the tracing threshold value to obtain a tracing fiber region image and a background fiber region image of the spinning whisker.

[0029] In this embodiment, obtaining the color grade of the tracer fiber specifically includes: using a standard sample of the tracer fiber and the background fiber to collect images, converting the sample image into the HSV color space, and extracting the hue angle from the tracer fiber area as the color grade of the tracer fiber.

[0030] It should be noted that the tracer fiber uses fluorescent fiber or dyed fiber with similar performance to the main fiber but obvious color contrast. The tracer fiber is evenly mixed into the fiber raw material in a certain proportion to ensure that the distribution of the tracer fiber is random and can represent the overall fiber behavior. In specific implementation, grayscale mapping can be performed according to the threshold interval of the color level of the tracer fiber according to a preset grayscale mapping table to obtain the corresponding grayscale value as the tracer threshold, and then the difference in optical properties between the tracer fiber and the background fiber is used to perform regional segmentation according to the tracer threshold, that is, pixel points with grayscale values ​​higher than the tracer threshold are classified as tracer fiber areas, and other pixels are classified as background fiber areas, thereby obtaining the tracer fiber area image and the background fiber area image of the spinning strands.

[0031] In step S3, connected region segmentation is performed based on the tracer fiber region image to obtain multiple tracer fiber region sub-graphs, and fracture features of each tracer fiber region sub-graph are extracted using a machine learning algorithm to obtain fiber radial fracture points corresponding to each tracer fiber region sub-graph.

[0032] In this embodiment, performing connected region segmentation based on the tracer fiber region image to obtain multiple tracer fiber region subgraphs specifically includes: acquiring the tracer fiber region image, defining an image connectivity standard, performing connected region segmentation on the tracer fiber region image based on the image connectivity standard to obtain multiple tracer fiber region subgraphs.

[0033] In specific implementation, the image connectivity standard can be defined as a 4-connectivity standard, that is, each pixel is associated with its upper, lower, left, and right adjacent pixels, and then the Flood Fill algorithm is used to mark the connected areas in the connected areas of the tracer fiber area image, and each separate connected area is used as a tracer fiber area subgraph, thereby obtaining multiple tracer fiber area subgraphs.

[0034] It should be noted that, in the present application, filaments with equally spaced weak rings in the length direction are fed from the front roller of the roving machine to be spun together with the slivers into roving, and the roving is drafted in the spun yarn drafting device, and the weak ring filaments are broken at the weak rings during the drafting process to become sequential tracer short fibers distributed in the spun yarn whiskers, and the tracer fibers are distributed in different radial sections. The fiber radial breaking points described in the present application are the fiber breaking points at different radial cross-sectional positions of the tracer fibers. In the present embodiment, a machine learning algorithm is used to extract the breaking features of each tracer fiber area sub-graph, and the fiber radial breaking points corresponding to each tracer fiber area sub-graph respectively include: Inputting a training data set into a support vector machine for machine learning training, wherein the training data set includes a standard tracer fiber region image and annotated breakpoints; For any tracer fiber region subgraph, an edge detection algorithm is used to extract the edge grayscale mutation position corresponding to the tracer fiber region subgraph, the shape attribute label of the connected region of the tracer fiber region subgraph is extracted, and a gray level co-occurrence matrix is ​​used to extract the local texture features of the tracer fiber region subgraph; The edge grayscale mutation position, the connected region shape attribute label and the local texture feature corresponding to the tracer fiber region sub-graph are input into the support vector machine, so as to extract the fiber radial break point of the tracer fiber region sub-graph.

[0035] In specific implementation, the Canny edge detection algorithm can be used to extract the edge grayscale value change position or amplitude corresponding to the tracer fiber area sub-image to form a feature vector, and the rectangularity of the tracer fiber area sub-image is extracted as the shape attribute label of the connected area. Then, the gray-level co-occurrence matrix is ​​used to extract the image contrast of the tracer fiber area sub-image as the local texture feature. The above three features are spliced ​​into a complete feature vector and then input into a support vector machine. Support vector regression is used to predict the radial fracture point position of the fiber in the tracer fiber area sub-image.

[0036] In some other embodiments, the image features of the fiber breakage site may be input into a machine learning algorithm for training, thereby learning the image features of the fiber breakage site and identifying the radial break points of the fibers in the tracer fiber region sub-image, which is not limited in the present application.

[0037] In step S4, the background fiber features of the background fiber area image are obtained, and based on the background fiber features and the radial fracture positions corresponding to the radial fracture points of each fiber, the hook morphology of the tracer fiber area image is extracted to obtain a hook morphology image score of the tracer fiber area image.

[0038] In this embodiment, obtaining the background fiber features of the background fiber area image specifically includes: obtaining the fiber arrangement direction of the background area image as the background fiber feature of the background fiber area image. In specific implementation, the direction vector corresponding to the fiber arrangement direction of the background area image can be used as the background fiber feature.

[0039] It should be noted that the different radial cross-sectional distributions of the tracer fibers can easily interfere with the accuracy parameters of the hook morphology image score when the forces at different radial positions are uneven. The present application determines the radial fracture position corresponding to the radial fracture point of each fiber, and determines the radial interference degree based on the degree of connection between the radial fracture position and the initial hook morphology image score. The final hook morphology image score is corrected based on the radial interference degree, thereby reducing the detection interference caused by uneven forces at different radial distribution positions and improving the accuracy of short fiber hook morphology detection.

[0040] In this embodiment, based on the background fiber features and the radial fracture positions corresponding to the radial fracture points of each fiber, the hook morphology of the tracer fiber region image is extracted, and the hook morphology image score of the tracer fiber region image is obtained, which specifically includes: According to the background fiber characteristics, hook morphology is extracted from each tracer fiber region sub-image in the tracer fiber region image to obtain initial hook morphology image scores corresponding to each tracer fiber region sub-image; Obtaining fiber radial fracture points corresponding to each tracer fiber region sub-image; determining radial fracture positions corresponding to each fiber radial fracture point based on the fiber radial fracture points corresponding to each tracer fiber region sub-image; The radial interference degree is extracted according to the radial fracture position and initial hook morphology image score corresponding to each radial fracture point of each fiber, and the radial interference degree corresponding to each tracer fiber region sub-image is obtained; The initial hook morphology image scores corresponding to each tracer fiber region sub-image are fused and corrected according to the radial interference degree to obtain the hook morphology image score of the tracer fiber region image.

[0041] Optionally, in some embodiments, the hook morphology is extracted for each tracing fiber area sub-image in the tracing fiber area image according to the background fiber feature, and the initial hook morphology image score corresponding to each tracing fiber area sub-image is obtained, which specifically includes: for any tracing fiber area sub-image, the fiber curvature and the number of hook points of the tracing fiber area sub-image are extracted according to the direction vector corresponding to the fiber arrangement direction in the background fiber feature, and the initial hook morphology image score corresponding to the tracing fiber area sub-image is generated based on the fiber curvature and the number of hook points, and then the initial hook morphology image score corresponding to other tracing fiber area sub-images is determined in the same way.

[0042] In a specific implementation, the pixel coordinates corresponding to the radial fracture points of the fibers corresponding to the respective tracer fiber region sub-images are used as radial fracture positions.

[0043] Optionally, in some embodiments, radial interference extraction is performed according to the radial fracture positions and initial hook morphology image scores corresponding to the radial fracture points of each fiber, and the radial interference corresponding to each tracer fiber region sub-image is obtained, specifically including: Obtain the radial break positions corresponding to the radial break points of each fiber and assign corresponding sequence tags; Determine the projection distance of each fiber radial break point on the direction vector of the background fiber feature as the radial distance, and obtain a radial distance sequence composed of the radial distances corresponding to each fiber radial break point; The initial hook morphology image scores corresponding to each tracing fiber area sub-image are obtained, and the interference degree detection windows corresponding to each tracing fiber area sub-image are determined according to the sequence labels of the fiber radial breakpoints corresponding to each tracing fiber area sub-image. Furthermore, within the interference degree detection windows corresponding to each tracing fiber area sub-image, the radial interference degree corresponding to each tracing fiber area sub-image is determined by the radial distance corresponding to each fiber radial breakpoint and the initial hook morphology image scores.

[0044] In specific implementation, the initial hook morphology image scores corresponding to each tracing fiber area subgraph can be sorted according to the sequence labels of the fiber radial breakpoints corresponding to each tracing fiber area subgraph, and then according to the window size of the preset interference degree detection window, the initial hook morphology image scores and radial distances of the tracing fiber area subgraphs corresponding to the same number of adjacent sequence labels are used as window elements, and the Pearson correlation coefficient between the initial hook morphology image scores and the radial distance is used as the radial interference degree corresponding to the tracing fiber area subgraph.

[0045] Optionally, in some embodiments, the initial hook morphology image scores corresponding to each tracing fiber area sub-image are fused and corrected according to the radial interference degree. In the process of obtaining the hook morphology image score of the tracing fiber area image, the initial hook morphology image scores corresponding to each tracing fiber area sub-image are weightedly fused according to the radial interference degree to obtain the final hook morphology image score.

[0046] In step S5, the drafting quality of the spinning staple fibers in the spinning process is detected according to the hook morphology image score of the tracer fiber region image.

[0047] In this embodiment, the drafting quality of the spinning staple fibers during the spinning process is detected according to the hook morphology image score of the tracer fiber area image, specifically including: determining the corresponding drafting quality grade according to the scoring interval of the hook morphology image score of the tracer fiber area image, and performing quality assessment according to the drafting quality grade.

[0048] In specific implementation, the hook morphology image score can be mapped to the corresponding drafting quality grade according to a preset score mapping table.

[0049] Embodiment 2: This application provides a short fiber abnormality detection system based on machine learning, referring to Figure 2 As shown, this figure is a schematic diagram of a short fiber abnormality detection system according to this embodiment of the present application, and the short fiber abnormality detection system includes: The image acquisition module 100 is used to acquire an image of the spinning staple fiber strands output by the front roller after the spinning process is drafted, so as to obtain an image of the spinning staple fiber strands; An image processing module 200 is used to extract feature objects from the spinning whisker image to obtain a tracer fiber region image and a background fiber region image of the spinning whisker; The image processing module 200 is further used to perform connected region segmentation based on the tracer fiber region image to obtain multiple tracer fiber region sub-graphs, and use a machine learning algorithm to extract fracture features of each tracer fiber region sub-graph to obtain fiber radial fracture points corresponding to each tracer fiber region sub-graph; An image scoring module 300 is used to obtain background fiber features of the background fiber region image, and based on the background fiber features and radial fracture positions corresponding to radial fracture points of each fiber, perform hook morphology extraction on the tracer fiber region image to obtain a hook morphology image score of the tracer fiber region image; The fiber draft detection module 400 is used to detect the draft quality of the spinning short fibers in the spinning process according to the hook morphology image score of the tracer fiber area image.

[0050] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems) and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0051] A person skilled in the art may understand that all or part of the steps in the various methods of the above embodiments may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, the storage medium including a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically-erasable programmable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0052] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

Claims

1. A method for detecting short fiber anomalies based on machine learning, characterized in that: The short fiber abnormality detection method comprises the following steps: After the spinning process is drafted, the spinning short fiber whiskers output by the front roller are imaged to obtain the spinning whisker images; Extracting characteristic objects from the spinning whisker image to obtain a tracer fiber region image and a background fiber region image of the spinning whisker; Based on the tracer fiber region image, connected region segmentation is performed to obtain multiple tracer fiber region sub-graphs, and fracture features of each tracer fiber region sub-graph are extracted using a machine learning algorithm to obtain fiber radial fracture points corresponding to each tracer fiber region sub-graph; Acquire background fiber features of the background fiber area image, perform hook morphology extraction on the tracer fiber area image based on the background fiber features and radial fracture positions corresponding to radial fracture points of each fiber, and obtain a hook morphology image score of the tracer fiber area image; The drafting quality of the spinning staple fibers during the spinning process is detected according to the hook morphology image score of the tracer fiber region image.

2. The method for detecting short fiber abnormalities based on machine learning according to claim 1, characterized in that: Extracting feature objects from the spinning whisker image to obtain the tracer fiber region image and the background fiber region image of the spinning whisker specifically includes: Obtaining the color grade of the tracer fiber, and mapping to obtain the tracer threshold value according to the threshold interval where the color grade of the tracer fiber is located; The spinning whisker image is segmented according to the tracing threshold value to obtain a tracing fiber region image and a background fiber region image of the spinning whisker.

3. A method for detecting short fiber abnormalities based on machine learning as claimed in claim 2, characterized in that: Obtaining the color grade of the tracer fiber specifically includes: using a standard sample of the tracer fiber and background fiber blended to collect images, converting the sample image into a color space, and extracting a hue angle from the tracer fiber region as the color grade of the tracer fiber.

4. The method for detecting short fiber abnormalities based on machine learning according to claim 1, characterized in that: Performing connected region segmentation based on the tracer fiber region image to obtain multiple tracer fiber region subgraphs specifically includes: acquiring the tracer fiber region image, defining an image connectivity standard, performing connected region segmentation on the tracer fiber region image based on the image connectivity standard to obtain multiple tracer fiber region subgraphs.

5. The method for detecting short fiber abnormalities based on machine learning according to claim 4, characterized in that: The machine learning algorithm is used to extract the fracture features of each tracer fiber region sub-image, and the fiber radial fracture points corresponding to each tracer fiber region sub-image are obtained, including: Inputting a training data set into a support vector machine for machine learning training, wherein the training data set includes a standard tracer fiber region image and annotated breakpoints; For any tracer fiber region subgraph, an edge detection algorithm is used to extract the edge grayscale mutation position corresponding to the tracer fiber region subgraph, the shape attribute label of the connected region of the tracer fiber region subgraph is extracted, and a gray level co-occurrence matrix is ​​used to extract the local texture features of the tracer fiber region subgraph; The edge grayscale mutation position, the connected region shape attribute label and the local texture feature corresponding to the tracer fiber region sub-graph are input into the support vector machine, so as to extract the fiber radial break point of the tracer fiber region sub-graph.

6. The method for detecting short fiber abnormalities based on machine learning according to claim 1, characterized in that: Based on the background fiber features and the radial fracture positions corresponding to the radial fracture points of each fiber, the hook morphology of the tracer fiber region image is extracted to obtain the hook morphology image score of the tracer fiber region image, which specifically includes: According to the background fiber characteristics, hook morphology is extracted from each tracer fiber region sub-image in the tracer fiber region image to obtain initial hook morphology image scores corresponding to each tracer fiber region sub-image; Obtaining fiber radial fracture points corresponding to each tracer fiber region sub-image; determining radial fracture positions corresponding to each fiber radial fracture point based on the fiber radial fracture points corresponding to each tracer fiber region sub-image; According to the radial fracture positions corresponding to the radial fracture points of each fiber and the initial hook morphology image score, the radial interference degree is extracted to obtain the radial interference degree between the radial fracture positions corresponding to the radial fracture points of each fiber and the initial hook morphology image score; The initial hook morphology image scores corresponding to each tracer fiber region sub-image are fused and corrected according to the radial interference degree to obtain the hook morphology image score of the tracer fiber region image.

7. The method for detecting short fiber abnormalities based on machine learning according to claim 1, characterized in that: According to the hook morphology image score of the tracer fiber area image, the drafting quality of the spinning staple fibers in the spinning process is detected, which specifically includes: determining the corresponding drafting quality grade according to the scoring interval of the hook morphology image score of the tracer fiber area image.

8. The method for detecting short fiber abnormalities based on machine learning according to claim 1, characterized in that: An industrial camera is used to collect images of the spinning staple fiber whiskers output by the front roller to obtain the spinning whisker images.

9. The method for detecting short fiber abnormalities based on machine learning according to claim 1, characterized in that: Before extracting the characteristic objects from the spinning yarn image, the method further includes: performing image preprocessing on the spinning yarn image.

10. A machine learning-based short fiber anomaly detection system, used to execute a machine learning-based short fiber anomaly detection method according to any one of claims 1 to 9, characterized in that: The short fiber abnormality detection system comprises: An image acquisition module is used to acquire images of the spinning short fiber whiskers output by the front roller after drafting in the spinning process to obtain images of the spinning whiskers; An image processing module is used to extract feature objects from the spinning whisker image to obtain a tracer fiber region image and a background fiber region image of the spinning whisker; The image processing module is further used to perform connected area segmentation based on the tracer fiber area image to obtain multiple tracer fiber area sub-graphs, and use a machine learning algorithm to extract fracture features of each tracer fiber area sub-graph to obtain fiber radial fracture points corresponding to each tracer fiber area sub-graph; An image scoring module is used to obtain background fiber features of the background fiber area image, and based on the background fiber features and radial fracture positions corresponding to radial fracture points of each fiber, perform hook morphology extraction on the tracer fiber area image to obtain a hook morphology image score of the tracer fiber area image; The fiber draft detection module is used to detect the draft quality of the spinning short fibers during the spinning process according to the hook morphology image score of the tracer fiber area image.

Images