Monitoring method and system for playing card production line

Abstract

The invention provides a monitoring method and system for a playing card production line. The monitoring method for the playing card production line comprises the steps that an original video, collected by monitoring equipment, of the playing card production line is received; segmenting the original video into a plurality of original video frames, screening out at least two continuous original video frames as target images based on an image feature matching algorithm, and adjusting the size of each target image to be consistent with the size of a preset standard image to obtain a plurality of standard target images; splicing the plurality of standard target images into a standard target image sequence according to the sequence of the time nodes corresponding to each standard target image; and inputting the standard target image sequence into the pre-trained YOLO model for analysis to obtain the production condition of the playing card production line, and when the production condition is abnormal, outputting alarm information, thereby improving the monitoring efficiency, timeliness and accuracy of the playing card production line.

Description

Technical Field

[0001] The present invention relates to the technical field of monitoring of a playing card production line, and in particular to a monitoring method and system for a playing card production line. Background Art

[0002] The playing card production line is a highly automated production process that involves multiple steps, from the preparation of raw materials to the packaging and quality inspection of the final product. It requires multi-faceted monitoring to ensure the normal operation of the production process and its safety during production.

[0003] The traditional monitoring method of playing card production lines mainly relies on manual inspections and offline detection. For example, maintenance is only carried out when production equipment has abnormalities, making it difficult to provide effective early warnings. There are problems such as untimely, inaccurate, and inefficient monitoring.

[0004] In the technical solution with application number CN201310597085.1, although it generates light under set conditions, illuminates the playing cards, takes images of the playing cards and recognizes them, and matches the target features extracted from the playing card images with the template features stored in the database to obtain recognition results, so as to achieve intelligent detection of playing cards, this technical solution only detects the quality of playing cards, and does not monitor the entire production status of the playing card production line, resulting in poor monitoring effect.

[0005] In the technical solution with application number CN 201880068678.X, the suit and grade are printed on one side of the card base paper, and the back pattern is printed on the other side, and the different paper page IDs for one or more card base papers are printed on the card base paper, and the individual cards cut from the card base paper by a cutting machine are used to form one or more decks of playing cards, and the one or more decks of playing cards are shuffled by a shuffling machine to form a group of shuffled playing cards, and the different shuffled playing card IDs for each group are assigned to the shuffled playing cards as ID codes, and the paper page IDs constituting the shuffled playing cards and the shuffled playing card IDs are associated with each other in the database. Therefore, this technical solution only realizes the manufacture of playing cards, but does not realize the monitoring of playing cards during the production process. Summary of the invention

[0006] The invention provides a monitoring method and system for a playing card production line, so as to improve the monitoring efficiency, timeliness and accuracy of the playing card production line.

[0007] To solve the above problems, the present invention adopts the following technical solutions: The present invention provides a monitoring method for a playing card production line, comprising: Receiving raw video of the playing card production line collected by monitoring equipment; The original video is divided into a plurality of original video frames, and at least two consecutive original video frames are selected as target images based on an image feature matching algorithm, wherein the target image includes at least one original video frame matching a preset image feature; Respectively adjusting the size of each target image to be consistent with the size of a preset standard image to obtain a plurality of standard target images; splicing the plurality of standard target images according to the sequence of time nodes corresponding to each of the standard target images to generate a standard target image sequence; Inputting the standard target image sequence into a pre-trained YOLO model for analysis to obtain the production status of the playing card production line; When the production status of the playing card production line is detected to be abnormal, an alarm message is output.

[0008] Preferably, the standard target image sequence is input into a pre-trained YOLO model for analysis to obtain the production status of the playing card production line, including: Using a pre-trained YOLO model, each of the standard target images in the standard target image sequence is evenly divided into a plurality of image blocks; For each of the standard target images, respectively predicting the bounding box and probability value of each corresponding image block, the bounding box includes the center point coordinates and size information of the image block, and the probability value is the probability that the target to be monitored falls into the bounding box, and the target to be monitored includes playing cards and production equipment on the playing card production line; After removing the overlapping bounding boxes using the non-maximum suppression algorithm, the remaining bounding boxes whose probability values ​​are lower than the preset probability value are removed according to the threshold filtering method to obtain the target bounding box of each of the standard target images; The monitored targets corresponding to the target boundary boxes of each of the standard target images are analyzed respectively to obtain the production status of the playing card production line.

[0009] Preferably, predicting the bounding box of each corresponding image block respectively includes: According to an edge detection algorithm, identifying all edge segments of each image block corresponding to each of the standard target images; Starting from the preset starting point, traverse each edge segment of each image block in a counterclockwise direction and assign a chain code value to each edge segment; In the counterclockwise direction, the difference between the chain code values ​​of every two adjacent edge segments is calculated in sequence to obtain all the chain code differences corresponding to each image block; Calculate the mean square error of all chain code differences corresponding to each image block to obtain the edge smoothness of each image block, where the edge smoothness is used to characterize the smoothness of the edge formed by all edge segments of the corresponding image block; When there is an abnormal image block whose edge smoothness is less than a preset edge smoothness, edge smoothing processing is performed on the abnormal image block according to an edge smoothing algorithm until the edge smoothness of the abnormal image block is greater than or equal to the preset edge smoothness; When the edge smoothness of each image block is greater than or equal to a preset edge smoothness, all edge segments of each image block are connected to form a corresponding boundary box.

[0010] Preferably, when the production status of the playing card production line is monitored to be abnormal, an alarm message is output, including: When the production status of the playing card production line is detected to be abnormal, abnormal information and alarm information are generated, wherein the abnormal information includes the location and type of the abnormal point on the playing card production line; generating an adjustment strategy for the playing card production line based on the abnormal information; The adjustment strategy and the alarm information are compressed and then output.

[0011] Furthermore, after the standard target image sequence is input into a pre-trained YOLO model for analysis and the production status of the playing card production line is obtained, the method further includes: When it is monitored that there is no abnormality in the production status of the playing card production line, receiving a playing card simulation image obtained by photographing the playing cards produced by the playing card production line by a monitoring device; Performing analog-to-digital conversion on the simulated playing card image to generate a digital playing card image; Recognize the digital image of the playing cards and extract image features of the digital image of the playing cards; Matching the image features of the digital image of the playing card with the template image features pre-stored in the database to obtain a matching result of the playing card; The playing cards are classified and packaged according to the matching results.

[0012] Furthermore, after the standard target image sequence is input into a pre-trained YOLO model for analysis and the production status of the playing card production line is obtained, the method further includes: When it is monitored that there is no abnormality in the production status of the playing card production line, receiving an image of the playing card production line taken by a monitoring device; Identifying posture features of workers in the playing card production line image; Comparing the posture features of the staff member with abnormal posture features pre-stored in the database to obtain a comparison result; Determining whether the posture of the staff member is abnormal according to the comparison result; When it is determined that the posture of the staff member is abnormal, a prompt message of abnormal posture is sent to the staff member.

[0013] Preferably, identifying the posture features of the workers in the playing card production line image includes: Inputting the playing card production line image into a pre-built convolutional neural network model to obtain a target playing card production line image, wherein the convolutional neural network model includes a plurality of residual blocks, each of which is used to mark at least one local feature of the playing card production line image, and between the residual blocks, an output result of a previous residual block is input to an input of a next adjacent residual block through a jump connection; Extracting target local features from all marked local features of the target playing card production line image to obtain posture features of the staff; The posture features of the staff member are compared with the abnormal posture features pre-stored in the database to obtain a comparison result, including: Aligning the posture features of the worker with the abnormal posture features pre-stored in the database in terms of format and dimension; According to the Pr-VIPE algorithm, the posture key points of the posture feature and the posture key points of the abnormal posture feature are mapped to the same embedding space with perspective invariance; In the embedding space, calculating the Euclidean distance between the posture feature and the abnormal posture feature; The Euclidean distance is compared with a preset threshold value to generate a comparison result between the posture feature of the worker and the abnormal posture feature.

[0014] Furthermore, when the production status of the playing card production line is detected to be abnormal, after outputting the alarm information, the method further includes: Receiving the operation data of each production equipment in the playing card production line collected by the collection equipment; When it is determined according to the operating data that the target production equipment is abnormal, a control instruction is issued to a control device corresponding to the target production equipment, so that the control device corrects the operating parameters of the target production equipment according to the control instruction.

[0015] Furthermore, before the standard target image sequence is input into a pre-trained YOLO model for analysis to obtain the production status of the playing card production line, the method further includes: Acquire multiple video samples about a playing card production line, wherein the video samples include normal video samples and abnormal video samples; Segmenting each of the video samples respectively, and selecting a plurality of training images from each of the video samples based on an image feature matching algorithm; For each of the video samples, the size of each of the training images is adjusted to be consistent with the size of a preset standard image, so as to obtain a plurality of standard training images for each of the video samples; For each of the video samples, respectively, according to the sequence of time nodes corresponding to each of the standard training images, multiple standard training images are spliced ​​together to generate a standard training image sequence for each of the video samples; The YOLO algorithm is used to train the standard training image sequence of each video sample, and when the training result meets the requirements, a YOLO model is obtained.

[0016] The present invention also provides a monitoring system for a playing card production line, comprising: A receiving module, used for receiving the original video of the playing card production line collected by the monitoring equipment; A segmentation module, used to segment the original video into multiple original video frames, and select at least two consecutive original video frames as target images based on an image feature matching algorithm, wherein the target image includes at least one original video frame matching a preset image feature; An adjustment module, used for adjusting the size of each target image to be consistent with the size of a preset standard image, so as to obtain a plurality of standard target images; A splicing module, used for splicing a plurality of the standard target images according to the sequence of time nodes corresponding to each of the standard target images to generate a standard target image sequence; An analysis module, used for inputting the standard target image sequence into a pre-trained YOLO model for analysis to obtain the production status of the playing card production line; The output module is used to output alarm information when the production status of the playing card production line is detected to be abnormal.

[0017] Compared with the prior art, the technical solution of the present invention has at least the following advantages: The monitoring method and system for a playing card production line provided by the present invention reduce the workload of manual monitoring through automated original video frame segmentation and target image screening, and also reduce the image processing amount during subsequent computer analysis, thereby improving monitoring efficiency; by utilizing an image feature matching algorithm, frames matching preset image features can be more accurately screened out from continuous original video frames, thereby improving the accuracy of target detection; by adjusting the size of the target image to be consistent with the preset standard image, it is helpful to unify the data format, which is convenient for the processing and analysis of the YOLO model; by splicing a plurality of standard target images into a standard target image sequence according to the sequence of time nodes, and inputting the pre-trained YOLO model for analysis, the YOLO model can comprehensively consider the previous and next standard target images during the image analysis process, thereby improving the analysis accuracy of the production status, and outputting alarm information when an abnormality is detected, thereby improving the real-time performance of monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 A flowchart of an embodiment of a monitoring method for a playing card production line of the present invention; Figure 2 A flowchart of another embodiment of the monitoring method of the playing card production line of the present invention; Figure 3 A flowchart of another embodiment of a monitoring method for a playing card production line of the present invention; Figure 4 A flowchart of another embodiment of the monitoring method of the playing card production line of the present invention; Figure 5 The present invention is a structural block diagram of an embodiment of the monitoring system for a playing card production line. DETAILED DESCRIPTION

[0019] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention.

[0020] In some of the processes described in the specification and claims of the present invention and the above-mentioned figures, multiple operations that appear in a specific order are included, but it should be clearly understood that these operations may not be executed in the order in which they appear in this article or executed in parallel. The sequence numbers of the operations, such as S11, S12, etc., are only used to distinguish between different operations, and the sequence numbers themselves do not represent any execution order. In addition, these processes may include more or fewer operations, and these operations may be executed in sequence or in parallel. It should be noted that the descriptions of "first", "second", etc. in this article are used to distinguish different messages, devices, modules, etc., do not represent the order of precedence, and do not limit the "first" and "second" to be different types.

[0021] It will be understood by those of ordinary skill in the art that, unless expressly stated, the singular forms "a", "an", "said" and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present invention refers to the presence of the features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof. It should be understood that when we refer to an element as being "connected" or "coupled" to another element, it may be directly connected or coupled to the other element, or there may be intermediate elements. In addition, the "connection" or "coupling" used herein may include wireless connection or wireless coupling. The term "and / or" used herein includes all or any unit and all combinations of one or more associated listed items.

[0022] It will be understood by those skilled in the art that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as generally understood by those skilled in the art in the art to which the present invention belongs. It should also be understood that terms such as those defined in common dictionaries should be understood to have meanings consistent with the meanings in the context of the prior art, and will not be interpreted with idealized or overly formal meanings unless specifically defined as herein.

[0023] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention, wherein the same or similar reference numerals throughout represent the same or similar elements or elements with the same or similar functions. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0024] Please refer to Figure 1 The present invention provides a monitoring method for a playing card production line, which may include the following steps: S11, receiving the original video of the playing card production line collected by the monitoring equipment; S12, dividing the original video into multiple original video frames, and selecting at least two consecutive original video frames as target images based on an image feature matching algorithm, wherein the target image includes at least one original video frame matching a preset image feature; S13, respectively adjusting the size of each target image to be consistent with the size of a preset standard image, to obtain a plurality of standard target images; S14, splicing the plurality of standard target images according to the sequence of time nodes corresponding to each of the standard target images to generate a standard target image sequence; S15, inputting the standard target image sequence into a pre-trained YOLO model for analysis to obtain the production status of the playing card production line; S16. When the production status of the playing card production line is detected to be abnormal, an alarm message is output.

[0025] First, video data is acquired from a surveillance camera installed on a playing card production line to obtain an original video. The original video is used for subsequent analysis to monitor the operating status of the production line. Videos are usually composed of continuous frames. In this embodiment, the original video can be decomposed into separate image frames to obtain multiple original video frames so that each frame can be analyzed separately.

[0026] Then, this embodiment can filter out the original video frames that match the preset features from the segmented multiple original video frames as the target image through feature matching based on an image feature matching algorithm, such as SIFT (Scale-Invariant Feature Transform). The preset features can be specific production line signs, signs of abnormal events, etc.

[0027] Among them, SIFT is a feature extraction and description algorithm that detects and describes local features in an image. This feature is invariant to image scaling, rotation, and brightness changes, and is also relatively stable to perspective changes and noise.

[0028] Specifically, this embodiment can simulate the changes of the original video frame at different scales by constructing a Gaussian difference pyramid (Difference of Gaussian, DoG), and then use the Hessian matrix at each scale to detect extreme points, that is, feature points of the original video frame. Then, a feature descriptor is generated for each feature point. The feature descriptor is a vector that contains the gradient direction and size information of the pixels in the neighborhood of the key point. The similarity between the feature descriptor of the original video frame and the corresponding vector of the preset image feature can be calculated, and the original video frame with a similarity greater than the preset similarity is used as the target image.

[0029] In addition, in order to ensure that the images input to the YOLO model have a consistent size, the selected target images need to be resized to help the model recognize and analyze the image content more accurately. Then, the resized standard target images are spliced ​​in the order in which they appear in the original video to form a series of continuous image sequences, which helps the YOLO model capture dynamic changes on the production line.

[0030] Finally, the pre-trained YOLO model can be used to analyze the stitched image sequence to identify and locate objects and events in the image, thereby evaluating the operating status of the playing card production line, such as each link of the production process, equipment status, product quality, production efficiency, etc. If the YOLO model identifies any abnormal situation, such as equipment jitter, operating errors or other unexpected events, the system will automatically trigger an alarm and notify relevant personnel to intervene, so that production line managers can quickly respond to safety violations and take measures to prevent potential safety accidents.

[0031] Among them, YOLO (You Only Look Once) is a target detection algorithm that is used to convert the target detection task into a regression problem, and simultaneously locate and classify the target through a single neural network to achieve real-time and efficient target detection, such as detecting the production status of a playing card production line.

[0032] The monitoring method for a playing card production line provided by the present invention reduces the workload of manual monitoring through automated original video frame segmentation and target image screening, and also reduces the image processing amount during subsequent computer analysis, thereby improving monitoring efficiency; by utilizing an image feature matching algorithm, frames matching preset image features can be screened out from continuous original video frames more accurately, thereby improving the accuracy of target detection; by adjusting the size of the target image to be consistent with the preset standard image, it is helpful to unify the data format, thereby facilitating the processing and analysis of the YOLO model; by splicing a plurality of standard target images into a standard target image sequence according to the sequence of time nodes, and inputting the pre-trained YOLO model for analysis, the YOLO model can comprehensively consider the previous and next standard target images during the image analysis process, thereby improving the analysis accuracy of the production status, and outputting alarm information when an abnormality is detected, thereby improving the real-time performance of monitoring.

[0033] Preferably, please refer to Figure 2 , inputting the standard target image sequence into the pre-trained YOLO model for analysis to obtain the production status of the playing card production line, which may specifically include: S151, using a pre-trained YOLO model, evenly dividing each of the standard target images in the standard target image sequence into a plurality of image blocks; S152, for each of the standard target images, respectively predicting a bounding box and a probability value of each corresponding image block, wherein the bounding box includes the center point coordinates and size information of the image block, and the probability value is a probability that the target to be monitored falls into the bounding box, and the target to be monitored includes playing cards and production equipment on the playing card production line; S153, after removing the overlapping bounding boxes using the non-maximum suppression algorithm, removing the remaining bounding boxes whose probability values ​​are lower than the preset probability value according to the threshold filtering method, to obtain the target bounding box of each of the standard target images; S154, analyzing the monitored targets corresponding to the target boundary boxes of each of the standard target images respectively to obtain the production status of the playing card production line.

[0034] In this embodiment, in order to improve the accuracy and efficiency of detection, each standard target image in the standard target image sequence is divided into multiple small blocks to obtain multiple image blocks, and all image blocks have the same size and dimension, so that separate target detection can be performed on each image block, which is helpful to identify targets in different areas in the image.

[0035] For each image block, the YOLO model predicts a bounding box that defines the center point coordinates and size information of the image block. At the same time, the YOLO model also assigns a probability value to each bounding box, indicating the probability that the monitored object exists within the bounding box.

[0036] The non-maximum suppression algorithm is used to remove overlapping bounding boxes and retain only the bounding boxes that are most likely to contain the target to be monitored. The threshold filtering method removes those bounding boxes whose probability values ​​are lower than a certain preset probability value because the probability of the bounding box containing the target to be monitored is low, so as to ensure the accuracy of the final result.

[0037] Then, the filtered target bounding box is analyzed to determine the location and status of the monitored target (playing cards and production equipment) in each standard target image, thereby evaluating the operating status of the production line.

[0038] This embodiment can improve the accuracy of target detection by dividing and detecting image blocks separately; at the same time, by utilizing the fast and accurate detection speed of the YOLO model, real-time monitoring of the playing card production line can be achieved; through the non-maximum suppression algorithm and threshold filtering method, the possibility of false detection is reduced, and the bounding box required for analysis is also reduced, further improving the processing efficiency of the YOLO model. In addition, this technical solution can also adapt to images of different sizes and shapes, improving the flexibility of the system.

[0039] For example, suppose that on a playing card production line, the quality of playing cards and the operating status of the machine need to be monitored. First, the video stream is obtained from the surveillance camera and divided into frames, and then adjusted to the standard size of the YOLO model input, and each frame image is divided into several small blocks. For example, if the image resolution is high, it can be divided into 16 small blocks. Then the YOLO model is used to detect each small block, predict the bounding box and probability value in each small block, and remove the overlapping bounding boxes using the non-maximum suppression algorithm. After removing the remaining bounding boxes with probability values ​​lower than the preset probability value according to the threshold filtering method, the remaining bounding boxes are analyzed to determine the location and status of the playing cards and the machine. For example, if a part of the machine is detected to be abnormal (such as abnormal jitter), an alarm is triggered and the operator is immediately notified to intervene to effectively monitor the operating status of the playing card production line, timely discover and deal with potential problems, and improve production efficiency and product quality.

[0040] In one embodiment, please refer to Figure 3 , respectively predicting the bounding box of each corresponding image block, which may specifically include: S1521, identifying all edge segments of each image block corresponding to each of the standard target images according to an edge detection algorithm; S1522, starting from a preset starting point, traverse each edge segment of each image block in a counterclockwise direction, and assign a chain code value to each edge segment; S1523, calculating the difference between the chain code values ​​of every two adjacent edge segments in sequence in a counterclockwise direction, to obtain all chain code difference values ​​corresponding to each image block; S1524, calculating the mean square error of all chain code differences corresponding to each image block to obtain edge smoothness of each image block, where the edge smoothness is used to characterize the smoothness of the edge formed by all edge segments of the corresponding image block; S1525, when there is an abnormal image block whose edge smoothness is less than the preset edge smoothness, performing edge smoothing processing on the abnormal image block according to the edge smoothing algorithm until the edge smoothness of the abnormal image block is greater than or equal to the preset edge smoothness; S1526: When the edge smoothness of each image block is greater than or equal to a preset edge smoothness, all edge segments of each image block are connected respectively to form a corresponding boundary box.

[0041] In this embodiment, the system can use an edge detection algorithm (such as the Canny edge detection method) to process each image block corresponding to each standard target image and identify all edge segments in the image block. The edge detection algorithm can identify areas in the image block where the brightness changes dramatically, which usually correspond to the outline or boundary of the object.

[0042] Then, a starting point is determined (usually the first edge point identified by the edge detection algorithm), and each edge segment in each image block is traversed in a counterclockwise direction to ensure that the edges are traversed in a consistent order for subsequent processing. Chain code is a method of representing edge directions using a sequence of numbers. For each edge segment, a chain code value is assigned based on its direction relative to the previous edge segment. This process involves quantizing the edge direction into a finite number of discrete values, such as 0-7 or 0-15. In order to capture local changes in the edge, the chain code difference between adjacent edge segments is calculated in a counterclockwise direction. The chain code difference reflects the turning and curvature changes of the edge in the local area.

[0043] The mean square error is a statistic that measures the degree of discreteness of the chain code difference distribution. In this embodiment, the mean square error of all corresponding chain code differences in each image block can be calculated as an indicator for measuring edge smoothness. At the same time, a preset edge smoothness threshold is set. If the edge smoothness of an image block is lower than this threshold, it indicates that there may be noise or abnormality at the edge of the image block, and edge smoothing processing is required. At this time, an edge smoothing algorithm (such as Gaussian filtering, median filtering, etc.) can be used to process the image block with edge smoothness lower than the threshold until its edge smoothness reaches or exceeds the preset threshold.

[0044] When the edge smoothness of all image blocks reaches a preset threshold, all edge segments of each image block are connected to form a bounding box of the image block. The bounding box can be used for subsequent target detection and analysis.

[0045] This embodiment can more accurately identify and process the edges in the image by calculating the chain code difference and edge smoothness. At the same time, edge smoothing helps to reduce the impact of noise on edge detection results and improve the robustness of target detection. In addition, accurate edge detection and smoothing can also improve the accuracy of subsequent target detection, especially in the complex background of a playing card production line.

[0046] In one embodiment, please refer to Figure 4 When the production status of the playing card production line is detected to be abnormal, an alarm message is output, which may specifically include: S161, when the production status of the playing card production line is monitored to be abnormal, generating abnormal information and alarm information, wherein the abnormal information includes the location and type of the abnormal point on the playing card production line; S162, generating an adjustment strategy for the playing card production line based on the abnormal information; S163: compress the adjustment strategy and the alarm information and then output them.

[0047] This embodiment can use the pre-trained YOLO model to analyze the standard target image sequence to identify abnormal conditions on the playing card production line, such as equipment failure, operational errors, or other unexpected events. Once an abnormality is detected, the system will record the location (such as a specific device or a specific area on the production line) and type (such as excessive jitter amplitude, abnormal speed, etc.) of the abnormal point.

[0048] The abnormal information is a detailed description of the abnormal situation on the production line, including the specific location and type of the abnormal point. The abnormal information is crucial for subsequent fault diagnosis and repair. Based on the abnormal information, the system will generate corresponding adjustment strategies. For example, if the vibration amplitude of a machine is detected to be too large, the adjustment strategy may include reducing the machine's operating speed.

[0049] In addition, in order to reduce the amount of data transmission and improve the response speed, the adjustment strategy and alarm information are compressed and then output to the control center or sent directly to the equipment of relevant personnel.

[0050] The system of this embodiment can quickly identify abnormal points and generate alarm information, which helps to respond to problems on the production line in a timely manner. At the same time, the abnormal information provides the specific location and type of the abnormal point, which helps to quickly locate the source of the problem. In addition, the adjustment strategy automatically generated based on the abnormal information can reduce human intervention and improve the automation level of the production line. By compressing the adjustment strategy and alarm information, the burden of data transmission is also reduced, and the response speed and efficiency of the system are improved.

[0051] For example, suppose that on a playing card production line, an important inspection step is to ensure the printing quality of each playing card. At this time, the YOLO model monitors the production line and identifies the abnormal situation of blurred printing on a playing card. The system records the abnormal location (such as a specific area of ​​the playing card) and type (blurred printing). Based on the abnormal information, the system determines that the ink volume of the printing machine may be improperly set or the printing pressure is insufficient, and then generates an adjustment strategy, such as adjusting the ink volume or increasing the printing pressure. The system compresses the adjustment strategy (adjusting the ink volume) and the alarm information (blurred printing alarm) and sends them to the production line control center. After receiving the compressed information, the control center decompresses and implements the adjustment strategy to adjust the settings of the printing machine. After the adjustment, the system continues to monitor the production line to ensure that the printing quality returns to normal, and records the adjustment effect to effectively monitor and manage the operation of the playing card production line, promptly detect and handle abnormal situations, reduce production interruptions and product quality problems, and improve production efficiency and product quality.

[0052] In yet another embodiment, after the standard target image sequence is input into a pre-trained YOLO model for analysis and the production status of the playing card production line is obtained, the following steps may also be included: When it is monitored that there is no abnormality in the production status of the playing card production line, receiving a playing card simulation image obtained by photographing the playing cards produced by the playing card production line by a monitoring device; Performing analog-to-digital conversion on the simulated playing card image to generate a digital playing card image; Recognize the digital image of the playing cards and extract image features of the digital image of the playing cards; Matching the image features of the digital image of the playing card with the template image features pre-stored in the database to obtain a matching result of the playing card; The playing cards are classified and packaged according to the matching results.

[0053] In this embodiment, when the production line is running normally and no abnormality is detected, the monitoring device (such as a high-definition camera) will take pictures of the playing cards on the production line to obtain images of the playing cards. The images are in the form of analog signals and need to be further processed.

[0054] Analog-to-digital conversion (ADC) is the process of converting an analog image into a digital image. Through this process, the analog image is converted into a digital signal that can be processed and analyzed by a computer system. Image features can be extracted from digital images using image processing and computer vision techniques such as edge detection, color recognition, texture analysis, etc. The image features may include the suit, number, pattern, etc. of a playing card.

[0055] Then, the extracted image features are compared with the standard template image features pre-stored in the database using a template matching algorithm or a feature point matching algorithm. Based on the matching results, the system can determine the exact identity of each playing card (such as Ace of Hearts, King of Spades, etc.) and guide the automated packaging line to sort the playing cards into the correct packaging box.

[0056] This embodiment can reduce the need for manual operation through automated image recognition and classification packaging, improve production efficiency, reduce human errors, and improve packaging accuracy. Fast image processing and matching algorithms can process a large number of playing cards in real time, improving the throughput of the production line. In addition, through accurate image recognition, it can ensure that each playing card is correctly classified, improving product quality.

[0057] For example, suppose a playing card production line needs to automatically sort and package the produced playing cards when it is operating normally. First, the camera on the production line takes pictures of the playing cards being packaged, obtains the image of the playing cards, and converts the analog image taken by the camera into a digital image through analog-to-digital conversion for computer processing. Then use image processing software to recognize the digital image of the playing cards and extract key features, such as suits, numbers, etc., and match the extracted features with the template features in the database to determine the specific type of each playing card. According to the matching results, the automated robotic arm sorts the playing cards into the corresponding packaging boxes, such as putting the Ace of Hearts into the packaging box of the Ace of Hearts. Throughout the process, the system also monitors the quality of the playing cards, such as printing quality, whether the cutting is neat, etc., to ensure that the final product meets the quality standards, thereby realizing automated quality control and sorting packaging, improving production efficiency and product quality, and reducing labor costs and errors.

[0058] In one embodiment, after the standard target image sequence is input into a pre-trained YOLO model for analysis and the production status of the playing card production line is obtained, the following steps may be further included: When it is monitored that there is no abnormality in the production status of the playing card production line, receiving an image of the playing card production line taken by a monitoring device; Identifying posture features of workers in the playing card production line image; Comparing the posture features of the staff member with abnormal posture features pre-stored in the database to obtain a comparison result; Determining whether the posture of the staff member is abnormal according to the comparison result; When it is determined that the posture of the staff member is abnormal, a prompt message of abnormal posture is sent to the staff member.

[0059] When the production line is operating normally, monitoring equipment (such as cameras) continuously captures real-time images of the production line to obtain a poker production line image. The image may include the staff, machine operation, product flow and other aspects.

[0060] Then, computer vision technology, especially posture estimation algorithms (such as OpenPose, AlphaPose, etc.), is used to extract the key points of the human body of the staff from the image of the playing card production line, including the position information of the head, hands, feet and other parts, so as to identify the posture characteristics of the staff. At the same time, the extracted posture features are matched and compared with the abnormal posture features pre-defined in the database to obtain the comparison results. The abnormal posture features can be obtained based on safety specifications, operating standards or historical data analysis, and the comparison results include the matching rate of the posture features and each abnormal posture feature.

[0061] By comparing the results, the system can determine whether the staff's posture meets the preset safety and operation standards. If the staff's posture matches the abnormal posture characteristics, the system will determine it as abnormal. Once the system determines that the staff's posture is abnormal, it can send abnormal posture prompt information to the staff through various means (such as display prompts, sound alarms, vibration devices, etc.), prompting them to immediately correct the improper posture to avoid possible safety accidents or operational errors.

[0062] This embodiment can timely detect and correct unsafe behaviors through real-time monitoring and analysis of the posture of workers, reduce safety accidents in the workplace, and ensure that workers comply with operating procedures to improve production efficiency and product quality. At the same time, through real-time feedback, employees' awareness of safety and operating regulations is enhanced, and incorrect operations caused by improper posture are reduced, reducing product defect rates.

[0063] For example, suppose that on a playing card production line, workers need to operate the machine in a specific posture to ensure safety and efficiency. First, the camera on the production line captures the worker's operation image in real time, uses the posture estimation algorithm to extract the worker's key point information from the image, identifies its posture features, and compares the extracted posture features with the normal and abnormal posture features pre-stored in the database. The system analyzes the comparison results and finds that the worker is bending over too much, which matches the abnormal posture features in the database. The system immediately sends a prompt message of abnormal posture to the worker through the workshop's alarm system. After receiving the prompt, the worker immediately adjusts his posture, so that the worker's operating posture can be monitored in real time, and abnormal posture can be discovered and corrected in time, thereby improving the safety and efficiency of the production line.

[0064] Preferably, identifying the posture features of the workers in the playing card production line image may specifically include: Inputting the playing card production line image into a pre-built convolutional neural network model to obtain a target playing card production line image, wherein the convolutional neural network model includes a plurality of residual blocks, each of which is used to mark at least one local feature of the playing card production line image, and between the residual blocks, an output result of a previous residual block is input to an input of a next adjacent residual block through a jump connection; Target local features are extracted from all marked local features of the target playing card production line image to obtain posture features of the worker.

[0065] This embodiment can use a pre-trained convolutional neural network (CNN) model to process the poker production line images obtained from the monitoring device. The CNN model is particularly suitable for image recognition tasks because it can automatically learn the features of the image.

[0066] Residual Block is a structure in CNN that solves the gradient vanishing problem in deep networks by introducing Skip Connection. Each residual block contains multiple convolutional layers to extract local features of the image.

[0067] In CNN, each residual block is responsible for capturing specific local features in the playing card production line image, such as edges, textures, or shapes. This local feature is crucial for understanding the image content.

[0068] Among them, the skip connection allows the network to directly pass the output of one residual block to the next adjacent residual block to retain more information and facilitate the propagation of gradients in the network, making the training of deep networks more effective.

[0069] After all local features of the target playing card production line image are marked, specific local features related to the task are extracted from all marked local features to obtain target local features. In this embodiment, the target local features are posture features of the staff, which are crucial for monitoring the behavior of personnel on the production line.

[0070] Ultimately, the system will identify and extract the worker's posture features, which can be used to analyze whether the worker's behavior complies with safety regulations or operating standards.

[0071] This embodiment can enable the deep network to learn complex features more effectively through the combination of residual blocks and jump connections; at the same time, jump connections help solve the gradient vanishing problem, making the training of deep networks easier. In addition, by learning local features, the model can better understand and generalize new scenes. The rapidly extracted posture features also enable the system to monitor the behavior of workers in real time and provide immediate feedback when anomalies are detected later.

[0072] Preferably, the posture feature of the staff member is compared with the abnormal posture feature pre-stored in the database to obtain a comparison result, which may specifically include: Aligning the posture features of the worker with the abnormal posture features pre-stored in the database in terms of format and dimension; According to the Pr-VIPE algorithm, the posture key points of the posture feature and the posture key points of the abnormal posture feature are mapped to the same embedding space with perspective invariance; In the embedding space, calculating the Euclidean distance between the posture feature and the abnormal posture feature; The Euclidean distance is compared with a preset threshold value to generate a comparison result between the posture feature of the worker and the abnormal posture feature.

[0073] In this embodiment, since the posture features extracted from the monitoring device and the abnormal posture features stored in the database may have different data formats and dimensions, they need to be converted to the same format and dimension for comparison. For example, they can be converted to the same format and dimension by normalization, scaling, or adjusting the data structure.

[0074] The Pr-VIPE (Probabilistic Viewpoint Invariant Human Pose Embedding) algorithm is a method for mapping the pose key points of pose features into a low-dimensional space that is view-invariant, meaning that the same human pose is represented similarly in the embedding space regardless of the viewpoint from which it is viewed.

[0075] In the embedding space, the system can measure the similarity between two points by calculating the Euclidean distance between them. The Euclidean distance is the actual distance between two points in multidimensional space, calculated as the square root of the sum of the squares of the differences in the coordinates of the two points.

[0076] In addition, by comparing the calculated Euclidean distance with a preset threshold, it is possible to determine whether the posture of the worker is similar to the abnormal posture feature. If the Euclidean distance is less than the preset threshold, the comparison result of the posture feature and the abnormal posture feature is considered to match; if the Euclidean distance is greater than the preset threshold, it is considered to be a mismatched comparison result.

[0077] This embodiment can improve the accuracy of posture recognition by using the Pr-VIPE algorithm, especially under different viewing angles. The viewing angle invariance enables the system to stably identify abnormal postures under different observation conditions. At the same time, fast posture comparison can provide real-time feedback for workers to correct unsafe postures in a timely manner. In addition, safety accidents in the workplace can be reduced by timely detecting and correcting abnormal postures.

[0078] In one embodiment, when the production status of the playing card production line is detected to be abnormal, after outputting the alarm information, the following steps may be further performed: Receiving the operation data of each production equipment in the playing card production line collected by the collection equipment; When it is determined according to the operating data that the target production equipment is abnormal, a control instruction is issued to a control device corresponding to the target production equipment, so that the control device corrects the operating parameters of the target production equipment according to the control instruction.

[0079] In this embodiment, sensors and data acquisition systems (such as PLC, SCADA system, etc.) can be used to collect real-time operation data of each production equipment on the playing card production line. The production equipment may include a printing machine, a die-cutting machine, a slitting machine, a packaging line, an embossing machine, etc. The operation data may include key parameters such as temperature, speed, pressure, current, and voltage.

[0080] The printing press includes a four-color printing press for printing the patterns and texts of playing cards; a two-color offset printing press for double-sided printing of playing cards. The die-cutting machine is used to die-cut the printed large sheets of paper into individual playing cards. The striping and slicing machine is used to strip and slice the die-cut playing cards. The packaging line is used for automatic packaging of playing cards. The embossing machine may include a fully automatic embossing machine for embossing the playing cards, so as to realize the simultaneous automatic detection of product quality and embossing.

[0081] Then, by analyzing the collected operation data in real time, it is possible to monitor whether the production equipment is operating as expected. If a parameter exceeds the preset normal range, or there is an abnormal correlation between multiple parameters, the system can determine that there is an abnormality in the target production equipment, and the target production equipment is the production equipment with abnormal operation data.

[0082] Once an abnormality is detected, the system automatically sends a control instruction to the control system of the target production equipment (such as a PLC or industrial PC). The control instruction is intended to adjust the operating parameters of the target production equipment to restore the normal operation of the equipment.

[0083] After receiving the control instruction, the control device will adjust the operating parameters of the target production equipment according to the control instruction, such as lowering the temperature, adjusting the speed, redistributing the load, etc., which will help the production equipment resume normal operation and avoid production interruptions or product quality problems.

[0084] This embodiment can reduce the need for manual intervention and improve the operating efficiency of the production line through automated abnormality detection and response; at the same time, it helps to reduce the downtime of production equipment and maintain production continuity by quickly identifying and responding to abnormal situations. In addition, by timely adjusting operating parameters, product quality problems caused by abnormal production equipment can be avoided. Preventive maintenance and timely troubleshooting also help reduce long-term maintenance costs and prevent potential safety accidents.

[0085] For example, suppose there is a device responsible for printing on a playing card production line, and its temperature and speed are key operating parameters. The sensor monitors the temperature and speed of the printing device in real time and sends the data to the central control system. The central control system analyzes the data and finds that the temperature is 10 degrees higher than the normal value. It determines that there is an abnormality in the equipment. The system automatically sends a control instruction to the PLC of the printing equipment, asking it to lower the temperature. After receiving the instruction, the PLC adjusts the operation of the cooling system to reduce the temperature to the normal range. The central control system continues to monitor the operating parameters until the temperature stabilizes within the normal range to ensure that the equipment resumes normal operation, so that it can monitor the equipment status in real time, detect and handle abnormal situations in time, thereby improving production efficiency and product quality, and reducing maintenance costs and safety risks.

[0086] In yet another embodiment, before the standard target image sequence is input into a pre-trained YOLO model for analysis to obtain the production status of the playing card production line, the method may further include: Acquire a plurality of video samples about a playing card production line, wherein the video samples include normal video samples and abnormal video samples; Segmenting each of the video samples respectively, and selecting a plurality of training images from each of the video samples based on an image feature matching algorithm; For each of the video samples, the size of each of the training images is adjusted to be consistent with the size of a preset standard image, so as to obtain a plurality of standard training images for each of the video samples; For each of the video samples, respectively, according to the sequence of time nodes corresponding to each of the standard training images, multiple standard training images are spliced ​​together to generate a standard training image sequence for each of the video samples; The YOLO algorithm is used to train the standard training image sequence of each video sample, and when the training result meets the requirements, a YOLO model is obtained.

[0087] In this embodiment, a video of a poker production line containing normal operation and abnormal conditions can be collected. The video will be used to train a YOLO model that can identify abnormalities on the production line. Each video sample is split into individual frames as potential training images. An image feature matching algorithm (such as SIFT, SURF, or ORB) is used to extract key frames from each video sample as training images, which contain feature information used to train the model.

[0088] To ensure the consistency of training data, the selected training images are adjusted to the same size as the preset standard images, which helps the YOLO model to learn and recognize better. The resized standard training images are spliced ​​into a sequence in chronological order to obtain a standard training image sequence for each video sample. This standard training image sequence will be used to simulate the video stream so that the YOLO model can learn the characteristics of the video that change over time.

[0089] Then, the stitched standard training image sequence is trained using the YOLO (You Only Look Once) algorithm. YOLO is a real-time object detection method that can identify objects and anomalies in images. Training continues until the model reaches a satisfactory accuracy and recall rate. Once this requirement is met, the model can be used to monitor the playing card production line in real time and detect anomalies.

[0090] In this embodiment, the model can be trained with normal and abnormal samples, and can more accurately identify anomalies. At the same time, the trained YOLO model can analyze image sequences in real time and quickly respond to anomalies on the production line, which also helps prevent accidents and improve product quality. In addition, automated anomaly detection reduces the need for manual monitoring, reduces costs and improves efficiency.

[0091] Please refer to Figure 5 , an embodiment of the present invention further provides a monitoring system for a playing card production line, which may include: A receiving module 21 is used to receive the original video of the playing card production line collected by the monitoring device; A segmentation module 22, configured to segment the original video into a plurality of original video frames, and select at least two consecutive original video frames as target images based on an image feature matching algorithm, wherein the target image includes at least one original video frame matching a preset image feature; An adjustment module 23, used for adjusting the size of each target image to be consistent with the size of a preset standard image, so as to obtain a plurality of standard target images; A stitching module 24 is used to stitch the plurality of standard target images according to the sequence of time nodes corresponding to each of the standard target images to generate a standard target image sequence; An analysis module 25 is used to input the standard target image sequence into a pre-trained YOLO model for analysis to obtain the production status of the playing card production line; The output module 26 is used to output an alarm message when it is detected that the production status of the playing card production line is abnormal.

[0092] The monitoring system for a playing card production line provided by the present invention reduces the workload of manual monitoring through automated original video frame segmentation and target image screening, and also reduces the image processing amount during subsequent computer analysis, thereby improving monitoring efficiency; by utilizing an image feature matching algorithm, frames matching preset image features can be screened out from continuous original video frames more accurately, thereby improving the accuracy of target detection; by adjusting the size of the target image to be consistent with the preset standard image, it is helpful to unify the data format, thereby facilitating the processing and analysis of the YOLO model; by splicing a plurality of standard target images into a standard target image sequence according to the chronological order of time nodes, and inputting the pre-trained YOLO model for analysis, the YOLO model can comprehensively consider the previous and next standard target images during the image analysis process, thereby improving the analysis accuracy of the production status, and outputting alarm information when an abnormality is detected, thereby improving the real-time performance of monitoring.

[0093] Regarding the system in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated here.

[0094] In one embodiment, the present invention further proposes a storage medium storing computer-readable instructions, which, when executed by one or more processors, enables one or more processors to execute the above-mentioned method for monitoring the playing card production line. The storage medium may be a ROM, a random access memory (RAM), a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, etc.

[0095] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a storage medium. When the program is executed, it can include the processes of the embodiments of the above-mentioned methods. The aforementioned storage medium can be a non-volatile storage medium such as a disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0096] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0097] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A method for monitoring a playing card production line, characterized in that: include: Receiving raw video of the playing card production line collected by monitoring equipment; The original video is divided into a plurality of original video frames, and at least two consecutive original video frames are selected as target images based on an image feature matching algorithm, wherein the target image includes at least one original video frame matching a preset image feature; Respectively adjusting the size of each target image to be consistent with the size of a preset standard image to obtain a plurality of standard target images; splicing the plurality of standard target images according to the sequence of time nodes corresponding to each of the standard target images to generate a standard target image sequence; Inputting the standard target image sequence into a pre-trained YOLO model for analysis to obtain the production status of the playing card production line; When the production status of the playing card production line is detected to be abnormal, an alarm message is output.

2. The monitoring method of a playing card production line according to claim 1, characterized in that: The standard target image sequence is input into the pre-trained YOLO model for analysis to obtain the production status of the playing card production line, including: Using a pre-trained YOLO model, each of the standard target images in the standard target image sequence is evenly divided into a plurality of image blocks; For each of the standard target images, respectively predicting the bounding box and probability value of each corresponding image block, the bounding box includes the center point coordinates and size information of the image block, and the probability value is the probability that the target to be monitored falls into the bounding box, and the target to be monitored includes playing cards and production equipment on the playing card production line; After removing the overlapping bounding boxes using the non-maximum suppression algorithm, the remaining bounding boxes whose probability values ​​are lower than the preset probability value are removed according to the threshold filtering method to obtain the target bounding box of each of the standard target images; The monitored targets corresponding to the target boundary boxes of each of the standard target images are analyzed respectively to obtain the production status of the playing card production line.

3. The monitoring method of a playing card production line according to claim 2, characterized in that: Predict the bounding box of each corresponding image block separately, including: According to the edge detection algorithm, identifying all edge segments of each image block corresponding to each of the standard target images; Starting from the preset starting point, traverse each edge segment of each image block in a counterclockwise direction and assign a chain code value to each edge segment; In the counterclockwise direction, the difference between the chain code values ​​of every two adjacent edge segments is calculated in sequence to obtain all the chain code differences corresponding to each image block; Calculate the mean square error of all chain code differences corresponding to each image block to obtain the edge smoothness of each image block, where the edge smoothness is used to characterize the smoothness of the edge formed by all edge segments of the corresponding image block; When there is an abnormal image block whose edge smoothness is less than a preset edge smoothness, edge smoothing processing is performed on the abnormal image block according to an edge smoothing algorithm until the edge smoothness of the abnormal image block is greater than or equal to the preset edge smoothness; When the edge smoothness of each image block is greater than or equal to a preset edge smoothness, all edge segments of each image block are connected to form a corresponding boundary box.

4. The monitoring method for a playing card production line according to claim 1, characterized in that: When the production status of the playing card production line is detected to be abnormal, an alarm message is output, including: When the production status of the playing card production line is monitored to be abnormal, abnormal information and alarm information are generated, wherein the abnormal information includes the location and type of the abnormal point on the playing card production line; generating an adjustment strategy for the playing card production line based on the abnormal information; The adjustment strategy and the alarm information are compressed and then output.

5. The method for monitoring a playing card production line according to claim 1, characterized in that: After inputting the standard target image sequence into the pre-trained YOLO model for analysis to obtain the production status of the playing card production line, the method further includes: When it is monitored that there is no abnormality in the production status of the playing card production line, receiving a playing card simulation image obtained by photographing the playing cards produced by the playing card production line by a monitoring device; Performing analog-to-digital conversion on the simulated playing card image to generate a digital playing card image; Recognize the digital image of the playing cards and extract image features of the digital image of the playing cards; Matching the image features of the digital image of the playing card with the template image features pre-stored in the database to obtain a matching result of the playing card; The playing cards are classified and packaged according to the matching results.

6. The method for monitoring a playing card production line according to claim 1, characterized in that: After inputting the standard target image sequence into the pre-trained YOLO model for analysis to obtain the production status of the playing card production line, the method further includes: When it is monitored that there is no abnormality in the production status of the playing card production line, receiving an image of the playing card production line taken by a monitoring device; Identifying posture features of workers in the playing card production line image; Comparing the posture features of the staff member with abnormal posture features pre-stored in the database to obtain a comparison result; Determining whether the posture of the staff member is abnormal according to the comparison result; When it is determined that the posture of the staff member is abnormal, a prompt message of abnormal posture is sent to the staff member.

7. The method for monitoring a playing card production line according to claim 6, characterized in that: Identifying the posture features of the workers in the playing card production line image includes: Inputting the playing card production line image into a pre-built convolutional neural network model to obtain a target playing card production line image, wherein the convolutional neural network model includes a plurality of residual blocks, each of which is used to mark at least one local feature of the playing card production line image, and between the residual blocks, an output result of a previous residual block is input to an input of a next adjacent residual block through a jump connection; Extracting target local features from all marked local features of the target playing card production line image to obtain posture features of the staff; The posture features of the staff member are compared with the abnormal posture features pre-stored in the database to obtain a comparison result, including: Aligning the posture features of the worker with the abnormal posture features pre-stored in the database in terms of format and dimension; According to the Pr-VIPE algorithm, the posture key points of the posture feature and the posture key points of the abnormal posture feature are mapped to the same embedding space with perspective invariance; In the embedding space, calculating the Euclidean distance between the posture feature and the abnormal posture feature; The Euclidean distance is compared with a preset threshold value to generate a comparison result between the posture feature of the worker and the abnormal posture feature.

8. The method for monitoring a playing card production line according to claim 1, characterized in that: When the production status of the playing card production line is detected to be abnormal, after outputting the alarm information, the method further includes: Receiving the operation data of each production equipment in the playing card production line collected by the collection equipment; When it is determined according to the operating data that the target production equipment is abnormal, a control instruction is issued to a control device corresponding to the target production equipment, so that the control device corrects the operating parameters of the target production equipment according to the control instruction.

9. The method for monitoring a playing card production line according to claim 1, characterized in that: Before inputting the standard target image sequence into the pre-trained YOLO model for analysis to obtain the production status of the playing card production line, the method further includes: Acquire multiple video samples about a playing card production line, wherein the video samples include normal video samples and abnormal video samples; Segmenting each of the video samples respectively, and selecting a plurality of training images from each of the video samples based on an image feature matching algorithm; For each of the video samples, the size of each of the training images is adjusted to be consistent with the size of a preset standard image, so as to obtain a plurality of standard training images for each of the video samples; For each of the video samples, respectively, according to the sequence of time nodes corresponding to each of the standard training images, multiple standard training images are spliced ​​together to generate a standard training image sequence for each of the video samples; The YOLO algorithm is used to train the standard training image sequence of each video sample, and when the training result meets the requirements, a YOLO model is obtained.

10. A monitoring system for a playing card production line, characterized in that: include: A receiving module, used for receiving the original video of the playing card production line collected by the monitoring equipment; A segmentation module, used to segment the original video into multiple original video frames, and select at least two consecutive original video frames as target images based on an image feature matching algorithm, wherein the target image includes at least one original video frame matching a preset image feature; An adjustment module, used for adjusting the size of each target image to be consistent with the size of a preset standard image, so as to obtain a plurality of standard target images; A splicing module, used for splicing a plurality of the standard target images according to the sequence of time nodes corresponding to each of the standard target images to generate a standard target image sequence; An analysis module, used for inputting the standard target image sequence into a pre-trained YOLO model for analysis to obtain the production status of the playing card production line; The output module is used to output alarm information when the production status of the playing card production line is detected to be abnormal.

Images