Low-power-consumption broken yarn detection method, device, equipment, medium and product

Abstract

The invention provides a low-power-consumption broken yarn detection method and device, equipment, a medium and a product, and the method comprises the steps: triggering a collection assembly of an inspection device to collect a target image corresponding to a first spinning station under the condition that the inspection device inspects to the first spinning station according to a preset inspection route and detects a target part on the first spinning station; based on yarn feature information, edge detection is carried out on the target image, an initial region of interest is obtained, and the yarn feature information is used for representing yarn physical features and yarn image features of yarn processed by the spinning equipment; filtering a non-yarn region in the initial region of interest by taking the yarn feature information as a reference to obtain a target region; performing yarn identification on the target area according to the texture features and the yarn feature information of the target area to obtain an identification result; and under the condition that the identification result indicates that no yarn exists in the target area, it is determined that the yarn breaking condition exists in the first spinning station. According to the embodiment of the invention, the accuracy of broken yarn identification can be improved.

Description

Technical Field

[0001] The present application relates to the field of detection technology, and in particular to a low-power yarn break detection method, device, equipment, medium and product. Background Art

[0002] In the textile industry, spinning machines and double twisting machines are the core processes of weaving. Their operating efficiency and stability directly determine the production capacity and quality performance of the entire production line. Among them, the common problem of yarn breakage not only leads to production interruptions and reduces production efficiency, but also requires workers to spend a lot of energy to find and repair broken ends, which greatly increases the labor intensity of workers. Therefore, it has become an urgent need for the development of the industry to use robots to realize automated inspections to replace manual yarn breakage detection. The yarn breakage detection method in the related technology relies on the image acquisition system to determine whether the yarn is broken. However, the textile production environment is complex and changeable. In order to ensure the timeliness and accuracy of the detection, the yarn breakage detection method in the related technology needs to continuously collect image data, causing the image acquisition equipment and related processing units to run continuously for a long time, which in turn causes the power consumption of the entire yarn breakage detection system to be at a high level, resulting in high power consumption of yarn breakage recognition in the related technology. Summary of the invention

[0003] The present application provides a low-power yarn break detection method, device, equipment, medium and product, which can reduce the power consumption of yarn break identification.

[0004] In a first aspect, an embodiment of the present application provides a low-power yarn break detection method, which is applied to an inspection device, the inspection device is used to inspect a spinning device, the spinning device includes a plurality of spinning stations, and the method includes:

[0005] When the inspection device inspects the first spinning station according to the preset inspection route and detects the target component on the first spinning station, the acquisition component of the inspection device is triggered to collect the target image corresponding to the first spinning station, the first spinning station is any spinning station among the multiple spinning stations, and the preset inspection route is an S-shaped route;

[0006] Based on the yarn feature information, edge detection is performed on the target image to obtain the initial region of interest. The yarn feature information is used to characterize the yarn physical characteristics and yarn image characteristics of the yarn processed by the spinning equipment;

[0007] Based on the yarn feature information, the non-yarn area in the initial region of interest is filtered to obtain the target area;

[0008] According to the texture features and yarn feature information of the target area, yarn recognition is performed on the target area to obtain a recognition result;

[0009] When the identification result indicates that no yarn exists in the target area, it is determined that a yarn breakage occurs in the first spinning station.

[0010] In a second aspect, the present application provides a low-power yarn break detection device, which is applied to an inspection device, wherein the inspection device is used to inspect a spinning device, wherein the spinning device includes a plurality of spinning stations, and the device includes:

[0011] a collection module, configured to trigger a collection component of the inspection device to collect a target image corresponding to the first spinning station when the inspection device inspects the first spinning station according to a preset inspection route and detects a target component on the first spinning station, wherein the first spinning station is any one of the multiple spinning stations and the preset inspection route is an S-shaped route;

[0012] A detection module, configured to perform edge detection on the target image based on yarn feature information to obtain an initial region of interest, wherein the yarn feature information is used to characterize yarn physical features and yarn image features of the yarn processed by the spinning device;

[0013] A filtering module, used for filtering the non-yarn area in the initial region of interest based on the yarn characteristic information to obtain a target area;

[0014] An identification module, used for performing yarn identification on the target area according to the texture features of the target area and the yarn feature information to obtain an identification result;

[0015] A determination module is used to determine whether the first spinning station has a yarn breakage condition when the identification result indicates that there is no yarn in the target area.

[0016] In a third aspect, an embodiment of the present application provides an electronic device, the electronic device comprising: a processor and a memory storing computer program instructions;

[0017] When the processor executes the computer program instructions, the low-power yarn break detection method in any one of the embodiments of the first aspect is implemented.

[0018] In a fourth aspect, an embodiment of the present application provides a computer storage medium having computer program instructions stored thereon. When the computer program instructions are executed by a processor, a low-power yarn break detection method as in any one of the embodiments in the first aspect is implemented.

[0019] In a fifth aspect, an embodiment of the present application provides a computer program product. When the instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes a low-power yarn break detection method as in any one of the embodiments in the first aspect above.

[0020] In a low-power yarn break detection method, device, equipment, medium and product provided in an embodiment of the present application, the inspection device adopts a specific trigger mechanism, and only when the first spinning station is inspected according to the S-shaped preset inspection route and the target component is detected, the acquisition component is started to collect the target image, avoiding or reducing the meaningless continuous operation of the image acquisition device, greatly reducing unnecessary energy consumption, and the S-shaped inspection route is the optimal route for inspection. The S-shaped inspection route can traverse each spinning station in an orderly, efficient and energy-saving manner according to the layout of the spinning equipment, and can cooperate with the trigger acquisition mechanism, thereby effectively reducing the power consumption of yarn break detection. In the subsequent image processing link, yarn break identification is carried out based on the predetermined yarn feature information. First, the target image is edge detected to obtain the initial area of ​​interest, and then the non-yarn area is accurately filtered based on this as a reference to obtain the target area. Finally, yarn identification is implemented based on the texture features and yarn feature information of the target area. The whole process is closely centered on the yarn features for accurate analysis, and there is no need to ensure accuracy through a large amount of redundant image data and complex calculations as in the related art. This not only reduces the operating time of the image acquisition device, but also reduces the computational burden of the processing unit, making the entire yarn break detection method more efficient and energy-saving, avoiding or reducing the high power consumption problem caused by continuous image acquisition in related technologies, and significantly reducing the power consumption of yarn break identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solution of the embodiments of the present application, the following is a brief introduction to the drawings required for use in the embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0022] Figure 1 It is a flowchart of a low-power yarn break detection method provided in an embodiment of the present application;

[0023] Figure 2 It is a structural schematic diagram of the inspection device provided in an embodiment of the present application;

[0024] Figure 3 This is one of the principle schematic diagrams of the low-power yarn break detection method provided in the embodiment of the present application;

[0025] Figure 4 This is the second principle schematic diagram of the low-power yarn break detection method provided in the embodiment of the present application;

[0026] Figure 5 This is the third principle schematic diagram of the low-power yarn break detection method provided in the embodiment of the present application;

[0027] Figure 6 This is the fourth schematic diagram of the principle of the low-power yarn break detection method provided in the embodiment of the present application

[0028] Figure 7 This is the fifth principle schematic diagram of the low-power yarn break detection method provided in the embodiment of the present application

[0029] Figure 8 It is a structural schematic diagram of a low-power yarn break detection device provided in an embodiment of the present application;

[0030] Fig. 9 It is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application.

[0031] Description of reference numerals:

[0032] 1. Inspection device; 2. Spinning equipment; 3. Sensor; 4. Light source; 5. Camera; 6. Overfeed roller; 7. Yarn; 8. Moving system; 9. Navigation system;

[0033] 801, acquisition module; 802, detection module; 803, filtering module; 804, identification module; 805, determination module; 901, processor; 902, memory; 903, communication interface; 910, bus. DETAILED DESCRIPTION

[0034] In order to more clearly understand the above-mentioned objectives, features and advantages of the present disclosure, the scheme of the present disclosure will be further described below. It should be noted that the embodiments of the present disclosure and the features in the embodiments can be combined with each other without conflict.

[0035] In the following description, many specific details are set forth to facilitate a full understanding of the present disclosure, but the present disclosure may also be implemented in other ways different from those described herein; it is obvious that the embodiments in the specification are only part of the embodiments of the present disclosure, rather than all of the embodiments.

[0036] It should be noted that, in this article, relational terms such as "first" and "second" are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the term "comprising" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article 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, article or device. In the absence of further restrictions, the elements defined by the sentence "comprising a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0037] At the front end of the weaving process, spinning machines and twisting machines often encounter yarn breakage problems during operation, which requires workers to conduct continuous inspections to find broken yarns and reconnect them. However, the harsh environment of high dust, high noise and high temperature in the workshop seriously threatens the health of the inspection workers. Therefore, it is particularly urgent to develop a robot that can replace manual inspections.

[0038] Currently, the inspection robots on the market are mainly powered by batteries, which are replaced when the battery is low to ensure 24-hour uninterrupted inspection. In order to reduce energy consumption, extend inspection time, and maximize battery efficiency, more stringent requirements are placed on the hardware usage strategy, software computing power requirements, and inspection algorithms of inspection equipment.

[0039] Although engineers have tried many yarn inspection solutions to improve the automation level of the textile industry over the years and have produced some practical improvement methods, most of these solutions have not fully considered the energy consumption issue. For example, the inspection device is set on the top of the mobile trolley, and the image sensor is rotating. The control unit continuously drives it to rotate to collect images of the workstations on both sides. However, this active detection method causes the camera and light source to run continuously, which not only increases power consumption, but also fails to ensure that each image collected can accurately capture the target position.

[0040] Another example is to use a pre-set inspection robot to shoot the working status of the two-for-one twister, obtain the working video, and extract all the frame images from it as the first frame image for analysis. However, this method requires the camera and light source to work continuously to record the video. At the same time, although the convolutional neural network algorithm it adopts can improve the accuracy of image recognition, its computationally intensive characteristics require high-performance computing hardware acceleration, which increases the power consumption of the device. In addition, some algorithm designs fail to fully consider the physical characteristics and image characteristics of the yarn, making it difficult to effectively distinguish the yarn from the background when the color of the yarn is similar to the environmental background or the yarn is moving quickly, resulting in false alarms or missed alarms. Therefore, improving the accuracy of yarn break detection and reducing power consumption are still urgent issues to be solved in the current automated inspection technology in the textile industry.

[0041] In order to solve the problems existing in the related art, the embodiments of the present application provide a low-power yarn break detection method, device, equipment, medium and product.

[0042] The present application embodiment provides a low-power yarn break detection method, device, equipment, medium and product. The following first introduces the low-power yarn break detection method provided by the present application embodiment. Figure 1 As shown, the method is applied to an inspection device 1, the inspection device 1 is used to inspect a spinning device 2, the spinning device 2 includes a plurality of spinning stations, and the method specifically includes the following steps:

[0043] S100, when the inspection device 1 inspects the first spinning station according to the preset inspection route and detects the target component on the first spinning station, the acquisition component of the inspection device 1 is triggered to collect the target image corresponding to the first spinning station, the first spinning station is any spinning station among the multiple spinning stations, and the preset inspection route is an S-shaped route.

[0044] Optionally, the inspection device 1 refers to a device that can automatically inspect the spinning equipment 2 and detect whether there is a yarn breakage in it. Specifically, the inspection device 1 can be a robot that integrates multiple sensors 3, cameras 5, light sources 4 and a mobile system 8. The robot can move in the spinning workshop and inspect each spinning station one by one. In this application, Figure 2 As shown, the core components of the inspection device 1 include a host computer (for controlling the entire system), a mobile system 8 (for enabling the robot to move within the workshop), a navigation system 9 (for helping the robot find the correct path and spinning station), a camera 5 (for collecting images of the spinning station), a light source 4 (for illuminating the yarn for the camera 5 to shoot), and a sensor 3 (for detecting the spinning station and triggering the camera 5 and the light source 4).

[0045] Specifically, the host computer is the control center of the entire inspection device 1. It is responsible for receiving and processing information from various components and issuing control instructions based on this information. The host computer can display the working status, position information and detected yarn status of the inspection device 1 in real time; it can also control the moving path and speed of the inspection device 1 in the workshop by sending instructions to the mobile system 8; it can also process and analyze the image data collected by the camera 5 to determine whether the yarn is broken; it can also record important events and test results during the detection process for subsequent analysis and improvement.

[0046] The mobile system 8 enables the inspection device 1 to move freely in the spinning workshop to inspect each spinning station. The mobile system 8 may specifically include components such as a drive motor, wheels or tracks. The mobile system 8 may adjust the moving direction and speed according to the instructions of the host computer. And under the guidance of the navigation system 9, it automatically avoids obstacles and finds the correct inspection path. The mobile system 8 may also maintain the stability of the inspection device 1 during movement to ensure that the camera 5 and the light source 4 can work normally.

[0047] The navigation system 9 helps the inspection device 1 find the correct driving path and spinning station in the spinning workshop. The navigation system 9 can construct a workshop map based on the environmental information collected by the sensor 3 and determine the position of the inspection device 1 in real time. Then, according to the workshop map and the inspection target, the optimal movement path of the inspection device 1 is planned. The navigation system 9 can also monitor obstacles in real time during the movement and guide the inspection device 1 to avoid obstacles.

[0048] The camera 5 and the light source 4 together constitute a collection component of the inspection device 1. The camera 5 is used to collect images of the spinning station. The light source 4 is used to illuminate the yarn so that the camera 5 can capture a clear image.

[0049] The sensor 3 is used to detect the spinning station and trigger the camera 5 and the light source 4. Specifically, the sensor 3 can locate the position of the yarn by identifying specific components (such as the overfeed roller 6) on the spinning station.

[0050] Optionally, the spinning equipment 2 refers to a machine or equipment for processing fiber raw materials into yarn. In the present application, the spinning equipment 2 can be a two-for-one twister or a spinning frame, and the two-for-one twister includes multiple spinning stations, each of which has one or more yarn processing units. The spinning station is a specific area or position on the spinning equipment 2 for processing yarn. On the two-for-one twister, each spinning station includes one or more yarn processing components, such as overfeed rollers 6, etc., which work together to complete the yarn processing process. In the present application, the spinning station is the target that the inspection device 1 needs to inspect and detect. The inspection device 1 locates the position of the yarn by identifying specific components (such as overfeed rollers 6) on the spinning station, and then performs yarn break detection.

[0051] Optionally, the target component refers to a specific component on the spinning station used to locate the yarn position or trigger the collection component. In the present application, the target component can be one or more iconic components on the spinning station, such as the overfeed roller 6, the yarn guide, etc. These components have a fixed relationship with the position or movement trajectory of the yarn, so the position of the yarn can be indirectly located by identifying these components.

[0052] The acquisition component refers to the part of the inspection device 1 used to obtain the image information of the spinning station. The acquisition component is mainly composed of a camera 5 and a possibly matching light source 4. When the inspection device 1 inspects a spinning station and detects the target component on the station, the acquisition component will be triggered to start collecting the image information of the station. This image information will then be used for subsequent yarn identification and yarn breakage detection.

[0053] Optionally, in a feasible implementation of the present application, as Figure 3As shown, when the inspection device 1 moves to any spinning station (i.e., the first spinning station) according to the preset inspection route, the inspection device 1 will first use the built-in sensor 3 to identify the target component on the station. For example, this target component can be the overfeed roller 6 on the two-for-one twisting station. Since the relative position of the overfeed roller 6 and the yarn is fixed, once the sensor 3 successfully identifies the overfeed roller 6, it means that the inspection device 1 has indirectly located the position of the yarn.

[0054] It should be noted that the preset inspection route is set as an S-shaped route, which can better adapt to the layout characteristics of the spinning equipment in the textile workshop without increasing too much travel, and ensure that the inspection device 1 can approach each spinning station in turn in a relatively balanced manner, so as to detect yarn breakage in time. Moreover, the S-shaped route planning method also provides a stable foundation for the subsequent targeted image acquisition and yarn breakage detection process, making the entire detection process more orderly and helping to reduce the overall detection power consumption, because the S-shaped route reduces the unnecessary movement and waiting time of the inspection device 1, allowing the detection resources to be more accurately invested in the spinning station that actually needs to be detected, thereby improving the performance and energy efficiency of the entire yarn breakage detection system.

[0055] Next, the sensor 3 triggers the camera 5 and the light source 4 to start working through an electrical signal link. This trigger mechanism ensures that the camera 5 and the light source 4 are only started when needed, thereby reducing power consumption and improving the response speed of the system. After the camera 5 and the light source 4 are triggered, the light source 4 will start working before the camera 5, illuminating the yarn to ensure that the camera 5 can capture a clear and accurate image. Subsequently, the camera 5 performs image acquisition according to the preset callback image acquisition mode.

[0056] In the callback image acquisition mode, by setting a callback function, the camera 5 only acquires images when it is really needed. The significant advantage of this method is that it greatly reduces the power consumption of the camera 5, because the camera 5 no longer performs endless image data acquisition tasks, but triggers image acquisition events based on the specific needs of the actual application. This is completely different from the traditional inspection mode, which often continuously acquires images and requires the application layer to use multi-threading technology to process these continuous image data, which not only leads to fierce competition for system CPU resources, but also comes with high image data processing costs. In contrast, the callback image acquisition mode cleverly avoids the frequent switching of multi-threads, thereby reducing the burden on the system.

[0057] Each time the camera 5 passes through the inspected station and is identified by the sensor 3, it will be hard-triggered to collect images. This hard-triggering method ensures that the camera 5 collects the target image at the right time, thereby improving the accuracy and reliability of image collection. At the same time, since each station only needs to calculate and process one image, the response speed and real-time performance of the system are also significantly improved.

[0058] Optionally, in a feasible implementation of the present application, in order to identify the target component, the inspection device 1 can use a variety of sensors 3 to achieve accurate detection. For example, since the position of the overfeed roller 6 on the two-for-one twister is fixed, the sensor 3 can determine its identity by measuring the distance or angle relative to the overfeed roller 6. Specifically, a high-precision position sensor 3 can be used, such as a photoelectric switch, a laser distance sensor 3, etc. When the sensor 3 detects a distance or angle that matches the preset position, it can be confirmed that the overfeed roller 6 has been identified.

[0059] In other embodiments, the overfeed roller 6 has specific physical features such as shape, size or material. The sensor 3 can determine whether it is an overfeed roller 6 by identifying these features. For example, the image recognition sensor 3 is used to capture the image of the overfeed roller 6 and compare it with a preset image template; or the shape sensor 3 is used to detect the outline and size of the overfeed roller 6 and match it with preset parameters.

[0060] In some embodiments, in order to simplify the identification process, specific marks such as a QR code, a barcode, a magnetic mark, etc. can be added to the overfeed roller 6. The sensor 3 can quickly identify the overfeed roller 6 by reading these marks. For example, a QR code reader is used to scan the QR code on the overfeed roller 6, or a magnetic sensor 3 is used to detect the presence of a magnetic mark. In summary, the method by which the sensor 3 identifies whether it is an overfeed roller 6 depends on the physical characteristics, working environment, and application requirements of the overfeed roller 6. By selecting a suitable position identification, feature identification, or mark identification method, and a suitable sensor 3 type, accurate and reliable identification can be achieved.

[0061] S200, performing edge detection on the target image based on yarn feature information to obtain an initial region of interest, wherein the yarn feature information is used to characterize yarn physical features and yarn image features of the yarn processed by the spinning device 2.

[0062] Optionally, yarn feature information can be used to describe the physical state of the yarn or to describe the appearance of the yarn in the image. Yarn physical features refer to the physical properties of the yarn itself, such as diameter, material, elasticity, strength, etc. Yarn image features are the appearance of the yarn in the image, including color, brightness, texture, shape, etc. Yarn image features are extracted from the image through image processing technology, and they reflect the visual characteristics of the yarn in the image.

[0063] Optionally, in a feasible implementation of the present application, in order to improve the image quality of the target image, reduce noise interference, and enhance the clarity of the yarn edge, the inspection device 1 can perform Gaussian filtering preprocessing on the collected target image.

[0064] Specifically, we can first construct a 3x3 Gaussian kernel. Each element value of this Gaussian kernel can be calculated according to the Gaussian function to ensure that the central element value is the largest, while the surrounding element values ​​gradually decrease according to the distance from the center, forming a weight distribution with high in the middle and low around. This design enables the Gaussian kernel to give a higher weight to the central pixel when processing the image, while considering the influence of its neighboring pixels, but the weight of the neighboring pixels decreases with the increase of distance.

[0065] In actual operation, the inspection device 1 will perform a convolution operation on the constructed Gaussian kernel with each pixel and its neighborhood in the target image. Specifically, each element of the Gaussian kernel is multiplied by the pixel value at the corresponding position in the target image, and then these products are summed, and the sum value obtained will be assigned to the center pixel of the current sliding matrix in the target image. This convolution process will be performed one by one along each pixel of the target image until the entire target image is covered. In this way, each pixel in the target image will be affected by its surrounding pixels (according to the weight distribution of the Gaussian kernel), thereby achieving image smoothing.

[0066] It should be noted that the smoothing effect of Gaussian filtering is closely related to the size of the Gaussian kernel and the standard deviation parameter. In this application, a Gaussian kernel of size 3x3 is selected to achieve a moderate smoothing effect while ensuring computational efficiency. The standard deviation parameter determines the weight distribution range of the Gaussian kernel, that is, the neighborhood range that affects the central pixel. By setting these parameters reasonably, the inspection device 1 can retain the clarity and detail features of the yarn edge while removing image noise.

[0067] After Gaussian filter preprocessing, the noise in the target image is effectively suppressed, and the yarn edge becomes smoother and clearer. This lays a solid foundation for subsequent steps such as yarn edge detection, contour extraction, and broken yarn identification. Therefore, Gaussian filter preprocessing is an indispensable part of the inspection device 1 to achieve accurate broken yarn detection, and it is also the key to improving detection efficiency and accuracy. Through the meticulous implementation of this step, the inspection device 1 can more accurately capture the state changes of the yarn, issue a yarn break warning in time, and provide strong technical support for the automation and intelligence of spinning production.

[0068] After Gaussian filtering the target image, the inspection device 1 can use one or more edge detection algorithms to accurately identify subtle changes in the edges of the yarns in the image. These edge detection algorithms will carefully analyze the grayscale value or color value of each pixel in the target image, looking for points where the values ​​have significantly changed, which usually mark the edges of the objects. In order to more accurately locate the edges of the yarns, the inspection device 1 will dynamically adjust the edge detection algorithm based on the physical and image characteristics of the yarns, such as adjusting the detection sensitivity, gradient threshold or applying a specific filter to ensure that the algorithm can accurately capture the edges of the yarns while effectively suppressing image noise and background interference.

[0069] In addition, the inspection device 1 can also use morphological operations (such as expansion, corrosion, opening and closing operations, etc.) to further optimize the edge detection results. These operations can enhance the continuity of the yarn edge, fill in the tiny breaks on the edge, and remove irrelevant fine structures. Through these fine edge detection steps, the inspection device 1 can accurately extract the edge areas that match the yarn feature information from the target image. These areas are the initial regions of interest (ROI). They not only contain the main contours of the yarn in the image, but also cover some related areas around the yarn, providing image data for subsequent non-yarn area filtering and yarn identification steps. It should be noted that although a variety of image processing techniques and algorithms are mentioned in the above description, in actual applications, the inspection device 1 will flexibly select and combine these techniques and algorithms according to factors such as the specific characteristics of the yarn, image quality, and detection requirements to achieve the best edge detection effect. At the same time, the parameter settings of edge detection (such as gradient threshold, size and intensity of morphological operations, etc.) can also be dynamically adjusted and optimized according to actual conditions.

[0070] S300, based on the yarn characteristic information, filtering the non-yarn area in the initial region of interest to obtain a target area.

[0071] Optionally, in a feasible implementation of the present application, after successfully acquiring the initial region of interest containing the yarn edge, the inspection device 1 will further use predefined yarn feature information to accurately distinguish the yarn area from other non-yarn areas in the initial region of interest.

[0072] Specifically, the inspection device 1 can use one or more advanced image processing and classification techniques, such as color space conversion, threshold segmentation, machine learning classifiers (such as support vector machines, random forests or deep learning models, etc.) or morphological operations, to analyze and compare the degree of match between each pixel or pixel block in the initial area of ​​interest and the yarn feature information.

[0073] During the processing, the inspection device 1 can first apply the color space conversion technology to convert the initial region of interest from the original color space (such as RGB) to a color space that is more suitable for yarn feature extraction and classification (such as hue, saturation, brightness (Hue Saturation Value, HSV), Lab, etc.), so as to better capture the color characteristics of the yarn. Then, using the threshold segmentation technology, the initial region of interest is segmented into two parts, the foreground (which can be yarn) and the background, according to the characteristics of the yarn brightness or color contrast. In order to further improve the accuracy of classification, the inspection device 1 can also use machine learning classifiers. These classifiers have been trained and optimized with a large amount of training data containing yarn and non-yarn samples, and can accurately identify the yarn area and non-yarn area in the initial region of interest. In addition, morphological operations (such as expansion, corrosion, opening and closing operations, etc.) are also used to optimize the classification results, by enhancing the coherence of the yarn area, filling the small breaks on the edge, and removing irrelevant small structures, to further improve the accuracy and robustness of yarn recognition.

[0074] Finally, through this series of sophisticated image processing and classification steps, the inspection device 1 can successfully filter out the non-yarn area from the initial area of ​​interest, and obtain a target area containing only the yarn and its surrounding key features. This target area not only provides clearer and more accurate input data for the subsequent yarn identification step, but also significantly reduces the amount of calculation, and improves the efficiency and accuracy of the entire yarn break detection process.

[0075] S400, performing yarn recognition on the target area according to the texture feature of the target area and the yarn feature information to obtain a recognition result.

[0076] Optionally, in a feasible implementation of the present application, yarn recognition in the target area is achieved by comparing and matching the texture features of the target area with predefined yarn feature information. First, the inspection device 1 extracts the texture features of the target area, which include but are not limited to the local variation pattern of pixel values, edge density, directionality and periodicity of texture, etc. These features can reflect the unique performance of yarn in the image, such as the thickness of yarn, density of texture, glossiness, etc. In order to extract these features, the inspection device 1 can use texture analysis techniques such as Gray-Level Co-occurrence Matrix (GLCM), Local Binary Pattern (LBP), Gabor filter or convolutional neural network in deep learning.

[0077] After extracting the texture features of the target area, the inspection device 1 can compare and match these texture features with the predefined yarn feature information library. The yarn feature information library is obtained through training and learning of a large amount of sample data in the early stage, and it contains the typical texture features of various yarn types in the image. The comparison and matching process involves the calculation of feature vectors, that is, converting the extracted texture features into mathematical vector form for quantitative comparison. Then, a suitable distance measurement method is selected, such as Euclidean distance, Manhattan distance or cosine similarity, to measure the similarity or difference between the texture feature vector and the yarn feature vector. On this basis, a classification algorithm, such as K nearest neighbor, support vector machine or decision tree, is applied to classify the texture features of the target area and determine the yarn type or state to which it belongs.

[0078] It should be noted that relying solely on contrast matching may not be accurate enough, because factors such as noise and lighting changes in the image may interfere with the recognition results. Therefore, the inspection device 1 can also adopt some optimization strategies to improve the accuracy and robustness of recognition. These strategies include using machine learning or deep learning models to further abstract and combine texture features to form more distinguishing feature representations; and considering the contextual information of the yarn in the image, such as position, direction, etc., which can help the inspection device 1 identify the yarn more accurately.

[0079] Finally, through the application of comparison matching and optimization strategies, the inspection device 1 can obtain a recognition result, which indicates whether there is yarn in the target area and the type or state of the yarn. If the recognition result indicates that there is no yarn in the target area, the inspection device 1 will further determine whether the situation is a yarn breakage situation, that is, whether the yarn fails to appear in the target area due to some reason (such as breakage, falling off or blockage). If it is confirmed to be a yarn breakage situation, the inspection device 1 will issue a corresponding alarm or signal so that the operator can take timely measures to deal with it.

[0080] S500: When the recognition result indicates that no yarn exists in the target area, determining that a yarn breakage occurs in the first spinning station.

[0081] Optionally, in a feasible implementation of the present application, if the recognition result clearly indicates that there is no yarn in the target area, the inspection device 1 will immediately start a series of subsequent judgment logics.

[0082] First, the inspection device 1 will conduct further analysis in combination with the operating logic of the spinning equipment 2. The spinning equipment 2 has a complex control system for monitoring and managing the operating status of each spinning station. The inspection device 1 will communicate with this control system to obtain the operating status information of the current spinning station, such as spinning speed, tension, etc. If the recognition result shows that there is no yarn in the target area, but the spinning station is in normal operation (such as normal spinning speed, stable tension, etc.), then the inspection device 1 will preliminarily determine that the spinning station may have a yarn breakage situation.

[0083] In order to further improve the accuracy of the judgment, the inspection device 1 can also use other auxiliary means for verification. For example, the inspection device 1 can also use the sensor 3 data of the spinning equipment 2, such as the yarn tension sensor 3, the yarn break sensor 3, etc., to assist in the judgment. If the information from multiple sources all points to the yarn break situation, the inspection device 1 will eventually confirm that the spinning station has a yarn break problem.

[0084] Once the yarn breakage is confirmed, the inspection device 1 will immediately send out an alarm or signal to notify the operator or the control system of the spinning equipment 2. This ensures that the operator can respond quickly and take necessary measures to solve the problem, such as stopping the machine for inspection, replacing the yarn, etc. At the same time, the inspection device 1 will also record the yarn breakage in the log for subsequent analysis and tracking.

[0085] In a low-power yarn break detection method provided in an embodiment of the present application, the inspection device 1 adopts a specific trigger mechanism. Only when the first spinning station is inspected according to the S-shaped preset inspection route and the target component is detected, the acquisition component is started to collect the target image, thereby avoiding or reducing the meaningless continuous operation of the image acquisition device, greatly reducing unnecessary energy consumption, and the S-shaped inspection route is the optimal route for inspection. The S-shaped inspection route can traverse each spinning station in an orderly, efficient and energy-saving manner according to the layout of the spinning equipment, and can cooperate with the trigger acquisition mechanism to effectively reduce the power consumption of yarn break detection. In the subsequent image processing link, yarn break identification is carried out based on the predetermined yarn feature information. First, the target image is edge detected to obtain the initial area of ​​interest, and then the non-yarn area is accurately filtered based on this as a reference to obtain the target area. Finally, yarn identification is implemented based on the texture features and yarn feature information of the target area. The whole process is closely centered on the yarn features for accurate analysis, and there is no need to ensure accuracy through a large amount of redundant image data and complex calculations as in the related art. This not only reduces the operating time of the image acquisition device, but also reduces the computational burden of the processing unit, making the entire yarn break detection method more efficient and energy-saving, avoiding or reducing the high power consumption problem caused by continuous image acquisition in related technologies, and significantly reducing the power consumption of yarn break identification.

[0086] In one embodiment, the edge detection of the target image is performed based on the yarn feature information to obtain an initial region of interest, including:

[0087] Obtaining the gradient value corresponding to each pixel in the target image;

[0088] Determine a pixel point corresponding to a first value as an edge point, wherein the first value is a gradient value greater than a preset gradient threshold value among the gradient values ​​corresponding to the pixel points, and the preset gradient threshold value is determined according to the yarn characteristic information;

[0089] When the distance between two target points is less than a first preset threshold, the two target points are determined as connected edge points, and the two target points are two edge points located in the same connected neighborhood, and the first preset threshold is determined according to the yarn feature information;

[0090] Based on the contour shape of the processed yarn, contour fitting is performed on the connected edge points to form the initial region of interest.

[0091] Optionally, in a specific implementation of the present application, first, the inspection device 1 can use image processing technology, such as Sobel operator, Canny edge detector, etc., to calculate the gradient value of each pixel in the target image. The gradient value reflects the speed of the change of the brightness of the pixel in the image and is an important basis for edge detection. These gradient values ​​will be used to determine which pixels may be located on the edge of the yarn.

[0092] Next, the inspection device 1 presets a gradient threshold based on the yarn feature information (such as yarn thickness, color contrast, etc.). This gradient threshold is used to distinguish which pixel points have gradient values ​​large enough to be considered edge points. Specifically, the inspection device 1 traverses the gradient values ​​of all pixel points and marks the pixel points corresponding to the gradient values ​​greater than the preset gradient threshold as edge points. These edge points preliminarily outline the contour of the yarn in the image.

[0093] However, relying solely on gradient values ​​to determine edge points may result in some isolated edge points or disconnected edge lines. In order to solve this problem, the inspection device 1 can further analyze the connectivity between these edge points. Specifically, traverse each edge point in the edge image and check other edge points in its neighborhood. If the distance between two edge points is less than a preset distance threshold (this threshold is related to the thickness of the yarn), they are regarded as connected edge points, that is, they are located in the same connected neighborhood. In this way, the inspection device 1 can connect isolated edge points to form a continuous edge line. Subsequently, morphological operations (such as expansion, filling holes, etc.) are used to improve the edge contour to form a continuous yarn contour. Finally, the initial area of ​​interest suspected to be the yarn is determined based on the circumscribed rectangle or minimum enclosing rectangle of this contour.

[0094] In these optional embodiments, the edge of the yarn is extracted from the target image based on the yarn feature information, thereby forming an initial region of interest. By calculating the gradient value and setting a reasonable gradient threshold, the yarn and the background can be effectively distinguished and noise interference can be reduced. At the same time, connecting adjacent edge points and fitting the contour can accurately reflect the actual shape of the yarn and improve the accuracy of the region of interest. This process not only improves the accuracy of edge detection, but also lays a solid foundation for the subsequent broken yarn analysis step, which helps to improve the efficiency and accuracy of the overall processing.

[0095] In one embodiment, the filtering of the non-yarn area in the initial region of interest based on the yarn characteristic information to obtain the target area includes:

[0096] Performing connected pixel region identification on the initial region of interest to obtain multiple identification regions;

[0097] When the boundary curvature of the first region is greater than a preset curvature threshold, or the area of ​​the first region is greater than a preset area threshold, or the aspect ratio of the first region is less than a second preset threshold, the first region is filtered from the initial region of interest to obtain the target region, the first region is any one of the multiple identification regions, and the preset curvature threshold, the preset area threshold and the second preset threshold are determined based on the yarn feature information.

[0098] Optionally, in a specific implementation of the present application, first, a connected pixel area recognition operation is performed on the initial region of interest, which can identify and segment all connected pixel sets to obtain Blob areas (i.e., recognition areas), each of which represents a potential yarn or interference area.

[0099] Then, for each identified Blob region, key features such as its boundary curvature, area, and aspect ratio are calculated. Among them, the boundary curvature reflects the smoothness of the boundary of the region, the area measures the size of the region, and the aspect ratio reveals the shape characteristics of the region. Then, based on the known characteristic information of the yarn, a series of thresholds are preset, including curvature threshold, area threshold, and aspect ratio threshold. These thresholds are used to evaluate whether each Blob region meets the typical characteristics of the yarn.

[0100] Specifically, if the boundary curvature of a Blob region is too large, it may indicate that its edge is irregular and does not conform to the smooth characteristics of the yarn, because the two ends of the working yarn are straight; if the area exceeds the preset range, it may mean that the region contains non-yarn interference; if the aspect ratio is too small, it may indicate that the shape of the region is too flat and does not conform to the slender shape of the yarn. Based on these evaluation criteria, the Blob regions that do not meet the conditions are filtered out from the initial region of interest, and only those Blob regions that meet the yarn characteristics are retained as target regions. This process not only improves the accuracy of target region extraction, but also effectively reduces noise and interference in subsequent processing, providing a reliable basis for further analysis and processing of yarns.

[0101] In these optional embodiments, the accuracy of target region extraction is significantly improved by accurately identifying and filtering non-yarn regions in the initial region of interest. Connected pixel region recognition technology is used, combined with the curvature, area, and aspect ratio thresholds set by yarn feature information, to effectively eliminate interference regions with rugged edges, abnormal areas, or inconsistent shapes. This process not only reduces noise and false detection, but also ensures a close match between the target region and the yarn features, providing more reliable data support for subsequent yarn analysis, defect detection, and other steps, and improving the efficiency and accuracy of the overall processing.

[0102] In one embodiment, performing yarn recognition on the target area according to the texture features of the target area and the yarn feature information to obtain a recognition result includes:

[0103] Traversing each pixel in the target area at a preset distance and a preset direction to obtain a plurality of pixel pairs, wherein the distance between two pixels in a pair of pixel pairs is the preset distance, and the direction between two pixels in a pair of pixel pairs is the preset direction;

[0104] Determining a gray level co-occurrence matrix of the target area according to the gray values ​​of the plurality of pixel pairs;

[0105] In the case where the texture features of the gray-level co-occurrence matrix meet the target preset conditions, the recognition result is determined as the existence of yarn in the target area, the texture features of the gray-level co-occurrence matrix include energy value, contrast, correlation value, and entropy value, and the target preset conditions are determined according to the yarn feature information;

[0106] When the texture feature of the gray level co-occurrence matrix does not satisfy the target preset condition, the recognition result is determined as no yarn exists in the target area.

[0107] Optionally, in a specific implementation of the present application, first, a set of preset distances and preset directions are set, and the selection of these parameters can be based on the characteristic information of the yarn to ensure that the key information of the yarn texture can be captured. Then, each pixel in the target area is traversed, and the corresponding pixel pairs are found according to the preset distance and direction. These pixel pairs constitute the basic data for analyzing the texture features.

[0108] After obtaining all pixel pairs, the grayscale values ​​of each pair of pixels are calculated, and the grayscale co-occurrence matrix of the target area is constructed based on these grayscale values. The grayscale co-occurrence matrix is ​​a two-dimensional array whose element values ​​represent the frequency of occurrence of pixel pairs with specific grayscale value combinations in a specific direction and distance. This grayscale co-occurrence matrix provides detailed statistical information about the image texture and is the basis for subsequent texture feature extraction.

[0109] Next, four key texture features are extracted from the grayscale co-occurrence matrix: energy value, contrast, correlation value and entropy value. The energy value reflects the uniformity of the grayscale distribution of the image and the coarseness of the texture. A high energy value means that the image texture is relatively uniform and rough. The contrast reveals the speed of the local grayscale change in the image. A high contrast indicates that there are obvious changes in light and dark in the image. The correlation value measures the degree of linear relationship between the grayscale values ​​in the image. A high correlation means that the grayscale value changes in the image have a stronger regularity. The entropy value reflects the randomness of the grayscale value distribution in the image. A high entropy value indicates that the image texture is more complex and chaotic.

[0110] In order to determine whether there is yarn in the target area, a series of target preset conditions can be set for the above four texture features based on the known yarn feature information. These conditions are based on the analysis of a large number of yarn images and experimental data, and are intended to ensure the accuracy and reliability of the recognition results. For example, for energy values, a range can be set, and it is believed that the energy value within this range represents the uniformity of the yarn texture; for contrast, a threshold can also be set, and it is believed that the contrast exceeding this threshold indicates that there is a significant change in light and dark in the yarn; for correlation values ​​and entropy values, corresponding judgment criteria are also set.

[0111] In the yarn recognition stage, the texture features of the target area are compared one by one with the preset target conditions. If all texture features meet the conditions, it is judged that there is yarn in the target area, and the recognition result is marked as "yarn exists". On the contrary, if any texture feature does not meet the conditions, or there is a significant deviation between the texture features of the target area and the yarn feature information, it is judged that there is no yarn in the target area, and the recognition result is marked as "yarn does not exist".

[0112] Optionally, in another feasible implementation of the present application, in addition to the gray level co-occurrence matrix method, a yarn recognition method based on deep learning can also be used. This method uses a deep learning algorithm, such as a convolutional neural network, to extract and classify features of the image of the target area. Compared with the gray level co-occurrence matrix, the deep learning algorithm can automatically learn and extract high-level semantic features in the image without manually designing feature extraction rules. This means that the deep learning algorithm can more flexibly adapt to different yarn types and texture features, thereby improving the accuracy and generalization of recognition.

[0113] In the deep learning solution, a large number of yarn images need to be annotated and preprocessed first to build a training data set. Then, these data sets are used to train the convolutional neural network so that it can learn the texture features and classification rules of the yarn. During the training process, the algorithm will continuously optimize the network parameters to improve the recognition accuracy of the yarn. Once the training is completed, the convolutional neural network can quickly and accurately identify the yarn in the new target area image.

[0114] In addition to deep learning, yarn recognition can also be performed based on multi-scale feature fusion. This method combines image features at different scales to capture the texture information of the yarn more comprehensively. In specific implementation, the target area image can be first decomposed at multiple scales to obtain sub-images at different scales. Then, feature extraction is performed on each sub-image separately, and these features are fused to obtain a richer description of the yarn texture features. This method can more effectively cope with the complexity and diversity of yarn textures and improve the accuracy and robustness of recognition.

[0115] In these optional embodiments, by capturing the texture features of the target area and using the grayscale co-occurrence matrix and its energy value, contrast, correlation value, entropy value and other texture feature parameters, the accurate identification of the yarn is effectively achieved. This not only improves the accuracy and stability of the identification, but also has strong adaptability and robustness, and can cope with the yarn detection needs in complex environments such as different lighting and noise. At the same time, the implementation of this method is relatively simple, the calculation efficiency is high, and the power consumption of broken yarn detection is reduced.

[0116] In one embodiment, when the inspection device 1 inspects the first spinning station and detects a target component on the first spinning station, before triggering the acquisition component of the inspection device 1 to acquire the target image corresponding to the first spinning station, the method further includes:

[0117] The inspection device 1 is controlled to move and inspect on a preset path according to a preset inspection route. The preset path includes multiple main paths and multiple branch paths. The multiple main paths are connected to the multiple branch paths. The main path is composed of channels between the spinning equipment 2, and the branch path is composed of aisles at the first end and the second end of the spinning equipment 2 along a preset direction. The inspection device 1 moves along the main path to any branch path.

[0118] Optionally, in a specific implementation of the present application, the inspection strategy of the inspection device 1 is carefully designed as an efficient and flexible path planning scheme, aiming to ensure full coverage of the spinning stations while maximizing the inspection efficiency and minimizing the path movement. Figure 4 and Figure 5 As shown, the preset path is clearly divided into a branch path and a main path. The branch path is mainly composed of spacious passages between the spinning equipment 2 (such as the two-for-one twisting equipment), which ensure that the inspection device 1 can smoothly shuttle between the various spinning stations and perform close-up and detailed inspections. The main path utilizes the aisles of the spinning equipment 2 along the preset direction (such as the front and rear ends of the equipment), which serve as bridges connecting the various branch paths, so that the inspection device 1 can efficiently switch between different branches.

[0119] In terms of planning the inspection route (i.e., the preset inspection route), the inspection device 1 adopts an S-shaped module movement strategy. The implementation process of this strategy is as follows: Figure 6As shown, the inspection device 1 starts from a preset starting point (or origin), and first moves along the main path to the front end of the nearest branch path. After entering the branch path, the inspection device 1 inspects the spinning stations in the branch path one by one in the order of entering from the front end and exiting from the rear end. After completing the inspection of the current branch, the inspection device 1 does not return directly to the main road, but continues to follow the logic of the S-shaped route, entering from the rear end of the next adjacent branch, exiting from the front end, and so on, until the inspection task of all branches is completed. This S-shaped movement mode not only avoids unnecessary path duplication, but also ensures that the inspection device 1 can evenly cover each spinning station, thereby improving the efficiency and comprehensiveness of the inspection.

[0120] In these optional embodiments, by controlling the inspection device 1 to move and inspect on a preset path including multiple main paths and multiple branch paths according to a preset inspection route, the inspection device can flexibly enter any branch path along the main path. The setting of this step ensures that the inspection device 1 can fully and efficiently cover all spinning stations. Not only the degree of automation of the inspection is improved, but also the inspection path is optimized, the possibility of missed inspection is reduced, and the inspection efficiency and accuracy are improved, which provides strong support for the stability and quality control of spinning production.

[0121] In one embodiment, controlling the inspection device 1 to move and inspect on a preset path according to a preset inspection route includes:

[0122] When the inspection device 1 moves to a first branch road and a first obstacle is detected on the first branch road, the inspection device 1 is controlled to exit the first branch road in the opposite direction of the current driving direction and move to the next branch road of the first branch road in the preset inspection route, wherein the first branch road is any branch road among the multiple branch road paths;

[0123] When the inspection device 1 moves to a first main road and a second obstacle is detected on the first main road, obtaining a feasibility of the inspection device 1 to bypass the second obstacle according to the preset inspection route, wherein the first main road is any one of the multiple main road paths;

[0124] When the feasibility degree is greater than or equal to a third preset threshold, controlling the inspection device 1 to bypass the second obstacle along a supplementary route, where the supplementary route is used to bypass the second obstacle;

[0125] When the feasibility degree is less than the third preset threshold, the inspection device 1 is controlled to return to the branch path of the last movement of the inspection device 1, and move from the target entrance of the branch path of the last movement to the next branch path, and the next branch path is the branch path of the next movement of the inspection device 1 according to the preset inspection route.

[0126] Optionally, in a specific implementation of the present application, Figure 7 As shown, first, when the inspection device 1 moves to the first branch in the preset path (the "first branch" here refers to any one of the multiple branch paths) for inspection, the inspection device 1 will use its own sensor 3 or camera and other detection means to perceive the surrounding environment. If a first obstacle is detected on the first branch, this may hinder the normal progress of the inspection device 1. At this time, a reverse movement instruction will be immediately triggered to control the inspection device 1 to gradually retreat in the opposite direction of the current driving direction until it completely exits the first branch. This step ensures that the inspection device 1 will not force its way through when encountering an obstacle, avoiding or reducing possible collisions or damage. After exiting the first branch, the inspection device 1 will automatically adjust its direction according to the navigation information in the preset inspection route and move to the next branch of the first branch in the preset inspection route.

[0127] However, if Figure 7 As shown, when the inspection device 1 moves on the main road path (the "first main road" here refers to any one of the multiple main roads), if a second obstacle is detected, the situation becomes more complicated. At this time, the inspection device 1 needs to rely on its built-in navigation system 9 to evaluate the feasibility of bypassing the obstacle. This evaluation process specifically involves a comprehensive consideration of the obstacle's position, size, shape, and the inspection device 1's own moving speed and steering ability.

[0128] If the navigation system 9 determines that the feasibility of bypassing the obstacle is greater than or equal to the third preset threshold (this threshold is preset according to the actual situation and inspection requirements), the inspection device 1 will bypass the obstacle along a supplementary route. This supplementary route is pre-planned to ensure that the inspection device 1 can continue to inspect along the preset inspection route after bypassing the obstacle. In the process of bypassing the obstacle, the inspection device 1 can adopt a smooth and efficient path planning strategy to minimize the detour time and distance.

[0129] However, if the navigation system 9 determines that the feasibility of bypassing the obstacle is less than the third preset threshold, that is, the bypass is too difficult or impossible to achieve, the inspection device 1 will adopt another strategy: return to the branch path of the last movement, and go to the next branch path from the other exit of the branch (i.e., the target entrance). This strategy ensures that the inspection device 1 can continue the inspection task when encountering an obstacle that cannot be bypassed, but bypasses the obstacle by selecting a different path. In the process of returning to the previous branch and going from another exit to the next branch, the inspection device 1 can use its own navigation system 9 to update the location information in real time and plan the best path.

[0130] In these optional embodiments, when encountering an obstacle in the first branch, the inspection device 1 can quickly reverse and exit and turn to the next branch to avoid stagnation. When the main road encounters an obstacle, by evaluating the feasibility of detour, the inspection device 1 can intelligently choose to detour or return to the previous branch and enter the next branch from another entrance, effectively coping with complex environments. This strategy not only improves the inspection efficiency, but also reduces the inspection interruption caused by obstacles, providing reliable guarantee for the continuous monitoring of the spinning station.

[0131] Figure 8 A schematic structural diagram of a low-power yarn break detection device provided in another embodiment of the present application is shown. For ease of explanation, only the parts related to the embodiment of the present application are shown.

[0132] Reference Figure 8 The low-power yarn break detection device is applied to an inspection device, the inspection device is used to inspect a spinning device, the spinning device includes a plurality of spinning stations, and the low-power yarn break inspection device may include:

[0133] The acquisition module 801 is used for triggering the acquisition component of the inspection device to acquire a target image corresponding to the first spinning station when the inspection device inspects the first spinning station according to a preset inspection route and detects a target component on the first spinning station, wherein the first spinning station is any one of the multiple spinning stations and the preset inspection route is an S-shaped route;

[0134] A detection module 802 is used to perform edge detection on the target image based on yarn feature information to obtain an initial region of interest, wherein the yarn feature information is used to characterize the yarn physical features and yarn image features of the yarn processed by the spinning device;

[0135] A filtering module 803 is used to filter the non-yarn area in the initial region of interest based on the yarn characteristic information to obtain a target area;

[0136] An identification module 804 is used to perform yarn identification on the target area according to the texture features of the target area and the yarn feature information to obtain an identification result;

[0137] The determination module 805 is used to determine whether the first spinning station has a yarn breakage condition when the identification result indicates that there is no yarn in the target area.

[0138] In one embodiment, the detection module 802 may include:

[0139] A first acquisition submodule is used to acquire the gradient value corresponding to each pixel point in the target image;

[0140] A first determination submodule, configured to determine a pixel point corresponding to a first value as an edge point, wherein the first value is a gradient value greater than a preset gradient threshold value among the gradient values ​​corresponding to the pixel points, and the preset gradient threshold value is determined according to the yarn characteristic information;

[0141] A second determination submodule is used to determine the two target points as connected edge points when the distance between the two target points is less than a first preset threshold, and the two target points are two edge points located in the same connected neighborhood, and the first preset threshold is determined according to the yarn feature information;

[0142] The fitting submodule is used to perform contour fitting on the connected edge points based on the contour shape of the processed yarn to form the initial region of interest.

[0143] In one embodiment, the filtering module 803 may include:

[0144] A first identification submodule is used to identify connected pixel regions of the initial region of interest to obtain multiple identification regions;

[0145] A filtering submodule is used to filter the first area from the initial area of ​​interest to obtain the target area when the boundary curvature of the first area is greater than a preset curvature threshold, or the area of ​​the first area is greater than a preset area threshold, or the aspect ratio of the first area is less than a second preset threshold, wherein the first area is any one of the multiple identification areas, and the preset curvature threshold, the preset area threshold and the second preset threshold are determined according to the yarn feature information.

[0146] In one embodiment, the identification module 804 may include:

[0147] A third determination submodule is used to traverse each pixel in the target area at a preset distance and a preset direction to obtain a plurality of pixel pairs, wherein the distance between two pixels in a pair of pixel pairs is the preset distance, and the direction between two pixels in a pair of pixel pairs is the preset direction;

[0148] A fourth determination submodule, configured to determine a gray level co-occurrence matrix of the target area according to the gray values ​​of the plurality of pixel pairs;

[0149] A fifth determination submodule is used to determine the recognition result as the presence of yarn in the target area when the texture features of the gray level co-occurrence matrix meet the target preset conditions, the texture features of the gray level co-occurrence matrix include energy value, contrast, correlation value, and entropy value, and the target preset conditions are determined according to the yarn feature information;

[0150] The sixth determination submodule is used to determine the recognition result as no yarn exists in the target area when the texture feature of the gray level co-occurrence matrix does not meet the target preset condition.

[0151] In one embodiment, the low power consumption yarn breakage detection device may further include:

[0152] A control module is used to control the inspection device to move and inspect on a preset path according to a preset inspection route. The preset path includes multiple main paths and multiple branch paths. The multiple main paths are connected to the multiple branch paths. The main path is composed of channels between the spinning equipment, and the branch path is composed of aisles at the first end and the second end of the spinning equipment along a preset direction. The inspection device moves along the main path to any branch path.

[0153] In one embodiment, the control module may include:

[0154] a first control submodule, configured to control the inspection device to exit the first branch in the opposite direction of the current driving direction and move to the next branch of the first branch in the preset inspection route when the inspection device moves to the first branch and a first obstacle is detected on the first branch, wherein the first branch is any branch path among the multiple branch paths;

[0155] A second acquisition submodule is used to acquire a feasibility of the inspection device to bypass the second obstacle according to the preset inspection route when the inspection device moves to a first main road and detects a second obstacle on the first main road, wherein the first main road is any main road path among the multiple main road paths;

[0156] A second control submodule is used to control the inspection device to bypass the second obstacle along a supplementary route when the feasibility degree is greater than or equal to a third preset threshold value, wherein the supplementary route is used to bypass the second obstacle;

[0157] The third control submodule is used to control the inspection device to return to the branch path of the last movement of the inspection device and move from the target entrance of the branch path of the last movement to the next branch path when the feasibility degree is less than the third preset threshold value. The next branch path is the branch path of the next movement of the inspection device according to the preset inspection route.

[0158] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present application, and are devices corresponding to the above-mentioned methods. All implementation methods in the above-mentioned method embodiment are applicable to the embodiments of the device. Its specific functions and technical effects brought about can be found in the method embodiment part, and will not be repeated here.

[0159] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0160] Fig. 9 A schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present application is shown.

[0161] The device may include a processor 901 and a memory 902 storing program instructions.

[0162] When the processor 901 executes the program, the steps in any of the above method embodiments are implemented.

[0163] Exemplarily, the program may be divided into one or more modules / units, one or more modules / units are stored in the memory 902 and executed by the processor 901 to complete the present application. One or more modules / units may be a series of program instruction segments capable of completing a specific function, and the instruction segments are used to describe the execution process of the program in the device.

[0164] Specifically, the processor 901 may include a central processing unit (CPU), or an application specific integrated circuit (ASIC), or may be configured to implement one or more integrated circuits of the embodiments of the present application.

[0165] The memory 902 may include a large capacity memory for data or instructions. By way of example and not limitation, the memory 902 may include a hard disk drive (HDD), a floppy disk drive, a flash memory, an optical disk, a magneto-optical disk, a magnetic tape, or a universal serial bus (USB) drive or a combination of two or more of these. In appropriate cases, the memory 902 may include a removable or non-removable (or fixed) medium. In appropriate cases, the memory 902 may be inside or outside the integrated gateway disaster recovery device. In a specific embodiment, the memory 902 is a non-volatile solid-state memory.

[0166] The memory may include read-only memory (ROM), random access memory (RAM), magnetic disk storage media devices, optical storage media devices, flash memory devices, electrical, optical or other physical / tangible memory storage devices. Thus, typically, the memory includes one or more tangible (non-transitory) readable storage media (e.g., memory devices) encoded with software including computer executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to an aspect of the present disclosure.

[0167] The processor 901 implements any one of the methods in the above embodiments by reading and executing program instructions stored in the memory 902 .

[0168] In one example, the electronic device may further include a communication interface 903 and a bus 910. The processor 901, the memory 902, and the communication interface 903 are connected via the bus 910 and communicate with each other.

[0169] The communication interface 903 is mainly used to implement communication between various modules, devices, units and / or equipment in the embodiments of the present application.

[0170] Bus 910 includes hardware, software or both, and the parts of online data flow billing equipment are coupled to each other. For example, but not limitation, bus may include accelerated graphics port (AGP) or other graphics bus, enhanced industrial standard architecture (EISA) bus, front-end bus (FSB), hypertransport (HT) interconnection, industrial standard architecture (ISA) bus, infinite bandwidth interconnection, low pin count (LPC) bus, memory bus, micro channel architecture (MCA) bus, peripheral component interconnection (PCI) bus, PCI-Express (PCI-X) bus, serial advanced technology attachment (SATA) bus, video electronics standard association local (VLB) bus or other suitable bus or two or more of these combinations. In appropriate cases, bus 910 may include one or more buses. Although the present application embodiment describes and shows a specific bus, the present application considers any suitable bus or interconnection.

[0171] In addition, in combination with the method in the above embodiment, the embodiment of the present application can provide a storage medium for implementation. The storage medium stores program instructions; when the program instructions are executed by a processor, any one of the methods in the above embodiment is implemented.

[0172] An embodiment of the present application further provides a chip, which includes a processor and a communication interface, wherein the communication interface and the processor are coupled, and the processor is used to run programs or instructions to implement the various processes of the above-mentioned method embodiment, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0173] It should be understood that the chip mentioned in the embodiments of the present application can also be called a system-level chip, a system chip, a chip system or a system-on-chip chip, etc.

[0174] An embodiment of the present application provides a computer program product, which is stored in a storage medium. The program product is executed by at least one processor to implement the various processes of the above-mentioned method embodiment and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0175] It should be clear that the present application is not limited to the specific configuration and processing described above and shown in the figures. For the sake of simplicity, a detailed description of the known method is omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of the present application is not limited to the specific steps described and shown, and those skilled in the art can make various changes, modifications and additions, or change the order between the steps after understanding the spirit of the present application.

[0176] The functional modules shown in the above block diagram can be implemented as hardware, software, firmware or their combination. When implemented in hardware, it can be, for example, an electronic circuit, an application specific integrated circuit (ASIC), suitable firmware, a plug-in, a function card, etc. When implemented in software, the elements of the present application are programs or code segments that are used to perform the required tasks. Programs or code segments can be stored in machine-readable media, or transmitted on a transmission medium or a communication link by a data signal carried in a carrier wave. "Machine-readable media" can include any medium that can store or transmit information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, optical fiber media, radio frequency (RF) links, etc. Code segments can be downloaded via computer grids such as the Internet, intranets, etc.

[0177] It should also be noted that the exemplary embodiments mentioned in this application describe some methods or systems based on a series of steps or devices. However, this application is not limited to the order of the above steps, that is, the steps can be performed in the order mentioned in the embodiment, or in a different order from the embodiment, or several steps can be performed simultaneously.

[0178] Aspects of the present disclosure are described above with reference to the flowchart and / or block diagram of the method, device (system) and program product according to the embodiment of the present disclosure. It should be understood that each box in the flowchart and / or block diagram and the combination of each box in the flowchart and / or block diagram can be implemented by computer program instructions. These program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer or other programmable data processing device to produce a machine so that these instructions executed by the processor of the computer or other programmable data processing device enable the implementation of the function / action specified in one or more boxes of the flowchart and / or block diagram. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor or a field programmable logic circuit. It can also be understood that each box in the block diagram and / or flowchart and the combination of boxes in the block diagram and / or flowchart can also be implemented by dedicated hardware that performs a specified function or action, or can be implemented by a combination of dedicated hardware and computer instructions.

[0179] The above are only specific implementation methods of the present application. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, modules and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. It should be understood that the protection scope of the present application is not limited to this. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed in this application, and these modifications or replacements should be included in the protection scope of this application.

Claims

1. A low-power yarn break detection method, characterized in that: Applied to an inspection device, the inspection device is used for inspecting a spinning device, the spinning device includes a plurality of spinning stations, and the method includes: When the inspection device inspects the first spinning station according to the preset inspection route and detects the target component on the first spinning station, the acquisition component of the inspection device is triggered to collect the target image corresponding to the first spinning station, the first spinning station is any spinning station among the multiple spinning stations, and the preset inspection route is an S-shaped route; Based on the yarn feature information, edge detection is performed on the target image to obtain an initial region of interest, wherein the yarn feature information is used to characterize the yarn physical characteristics and yarn image characteristics of the yarn processed by the spinning device; Based on the yarn characteristic information, filtering the non-yarn area in the initial region of interest to obtain a target area; According to the texture features of the target area and the yarn feature information, yarn recognition is performed on the target area to obtain a recognition result; When the identification result indicates that no yarn exists in the target area, it is determined that a yarn breakage occurs in the first spinning station.

2. The low-power yarn breakage detection method according to claim 1, characterized in that: The step of performing edge detection on the target image based on the yarn feature information to obtain an initial region of interest includes: Obtaining the gradient value corresponding to each pixel in the target image; Determine a pixel point corresponding to a first value as an edge point, wherein the first value is a gradient value greater than a preset gradient threshold value among the gradient values ​​corresponding to the pixel points, and the preset gradient threshold value is determined according to the yarn characteristic information; When the distance between two target points is less than a first preset threshold, the two target points are determined as connected edge points, and the two target points are two edge points located in the same connected neighborhood, and the first preset threshold is determined according to the yarn feature information; Based on the contour shape of the processed yarn, contour fitting is performed on the connected edge points to form the initial region of interest.

3. The low-power yarn breakage detection method according to claim 1, characterized in that: The filtering of the non-yarn area in the initial region of interest based on the yarn characteristic information to obtain the target area includes: Performing connected pixel region identification on the initial region of interest to obtain multiple identification regions; When the boundary curvature of the first region is greater than a preset curvature threshold, or the area of ​​the first region is greater than a preset area threshold, or the aspect ratio of the first region is less than a second preset threshold, the first region is filtered from the initial region of interest to obtain the target region, the first region is any one of the multiple identification regions, and the preset curvature threshold, the preset area threshold and the second preset threshold are determined based on the yarn feature information.

4. The low-power yarn breakage detection method according to claim 1, characterized in that: The step of performing yarn recognition on the target area according to the texture feature of the target area and the yarn feature information to obtain a recognition result includes: Traversing each pixel in the target area at a preset distance and a preset direction to obtain a plurality of pixel pairs, wherein the distance between two pixels in a pair of pixel pairs is the preset distance, and the direction between two pixels in a pair of pixel pairs is the preset direction; Determining a gray level co-occurrence matrix of the target area according to the gray values ​​of the plurality of pixel pairs; In the case where the texture features of the gray-level co-occurrence matrix meet the target preset conditions, the recognition result is determined as the existence of yarn in the target area, the texture features of the gray-level co-occurrence matrix include energy value, contrast, correlation value, and entropy value, and the target preset conditions are determined according to the yarn feature information; When the texture feature of the gray level co-occurrence matrix does not satisfy the target preset condition, the recognition result is determined as no yarn exists in the target area.

5. The low-power yarn breakage detection method according to claim 1, characterized in that: When the inspection device inspects the first spinning station according to the preset inspection route and detects the target component on the first spinning station, before triggering the acquisition component of the inspection device to acquire the target image corresponding to the first spinning station, the method further includes: The inspection device is controlled to move and inspect on a preset path according to the preset inspection route. The preset path includes multiple main paths and multiple branch paths. The multiple main paths are connected to the multiple branch paths. The main path is composed of channels between the spinning equipment, and the branch path is composed of aisles at the first end and the second end of the spinning equipment along a preset direction. The inspection device moves along the main path to any branch path.

6. The low-power yarn breakage detection method according to claim 5, characterized in that: The controlling the inspection device to move and inspect on a preset path according to the preset inspection route comprises: When the inspection device moves to a first branch road and a first obstacle is detected on the first branch road, the inspection device is controlled to exit the first branch road in a direction opposite to the current driving direction and move to a next branch road of the first branch road in the preset inspection route, wherein the first branch road is any branch road among the multiple branch road paths; When the inspection device moves to a first main road and a second obstacle is detected on the first main road, obtaining a degree of feasibility of the inspection device to bypass the second obstacle according to the preset inspection route, wherein the first main road is any one of the multiple main road paths; When the feasibility is greater than or equal to a third preset threshold, controlling the inspection device to bypass the second obstacle along a supplementary route, where the supplementary route is used to bypass the second obstacle; When the feasibility degree is less than the third preset threshold, the inspection device is controlled to return to the branch path of the last movement of the inspection device, and move from the target entrance of the branch path of the last movement to the next branch path, where the next branch path is the branch path of the next movement of the inspection device according to the preset inspection route.

7. A low power consumption yarn breakage detection device, characterized in that: Applied to an inspection device, the inspection device is used for inspecting a spinning device, the spinning device includes a plurality of spinning stations, and the device includes: a collection module, configured to trigger a collection component of the inspection device to collect a target image corresponding to the first spinning station when the inspection device inspects the first spinning station according to a preset inspection route and detects a target component on the first spinning station, wherein the first spinning station is any one of the multiple spinning stations and the preset inspection route is an S-shaped route; A detection module, configured to perform edge detection on the target image based on yarn feature information to obtain an initial region of interest, wherein the yarn feature information is used to characterize yarn physical features and yarn image features of the yarn processed by the spinning device; A filtering module, used for filtering the non-yarn area in the initial region of interest based on the yarn characteristic information to obtain a target area; An identification module, used for performing yarn identification on the target area according to the texture features of the target area and the yarn feature information to obtain an identification result; A determination module is used to determine whether the first spinning station has a yarn breakage condition when the identification result indicates that there is no yarn in the target area.

8. An electronic device, characterized in that: The device comprises: a processor and a memory storing computer program instructions; When the processor executes the computer program instructions, the low-power yarn breakage detection method according to any one of claims 1 to 6 is implemented.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer program instructions, and when the computer program instructions are executed by the processor, the low-power yarn breakage detection method according to any one of claims 1 to 6 is implemented.

10. A computer program product, characterized in that When the instructions in the computer program product are executed by a processor of an electronic device, the electronic device executes the low-power yarn break detection method according to any one of claims 1 to 6.

Images