Smoke box appearance detection method, system and equipment based on global shutter

Through global shutter technology and deep learning model, the problem of image blurring in traditional cigarette box appearance detection system in dark light environments is solved, the accuracy and traceability of cigarette box appearance detection are improved, and the accurate identification of cigarette box appearance defects is achieved. and efficient detection.

CN119064362APending Publication Date: 2024-12-03CHINA TOBACCO GUANGXI IND
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411301218.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-09-18
Publication Date
2024-12-03

AI Technical Summary

Technical Problem

When shooting fast moving objects, especially in dark light environments, traditional cigarette box appearance detection systems are prone to the jelly effect caused by partial exposure, slope patterns or shaking, resulting in image deformation and blurring, low detection accuracy, and the cigarette box cannot be accurately identified. Appearance defects.

Method used

The global shutter technology is used to obtain the outbound data set, appearance data set and standard data set of cigarette boxes. The tolerance compensation index is calculated through preprocessing and deep learning models, and the characteristic normal data set is determined, so as to judge the appearance status of cigarette boxes and generate a detection report. , reduce exposure and shaking phenomena, improve image clarity and detection accuracy.

Benefits of technology

It improves the accuracy and traceability of detection of cigarette box appearance defects, ensures the accuracy and efficiency of cigarette box appearance detection, and is suitable for capture with high synchronization requirements in dynamic scenarios.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119064362A_ABST
    Figure CN119064362A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of appearance detection, and discloses a smoke box appearance detection method, system and equipment based on a global shutter, and the method comprises the steps: obtaining a warehouse-out data set, an appearance data set and a standard data set of a smoke box; preprocessing the ex-warehouse data set and the appearance data set to obtain a smoke box image group, and determining a feature normal data group and a tolerance compensation index according to the smoke box image group and a standard data set; and judging the appearance state of the smoke box according to the characteristic normal data set and the tolerance compensation index, and generating a smoke box appearance detection report. According to the smoke box appearance detection method based on the global shutter, by obtaining the ex-warehouse data set, the appearance data set and the standard data set of the smoke box and determining the feature normal data set and the tolerance compensation index, the appearance state of the smoke box is accurately judged, the identification precision of the appearance defect of the smoke box is improved, and the detection efficiency is improved. And a detection report is generated according to the appearance state of the smoke box, so that the traceability of appearance detection of the smoke box is improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of appearance detection, and in particular to a method, system and device for detecting the appearance of a cigarette box based on a global shutter. Background Art

[0002] The cigarette box appearance inspection system is an automated visual inspection system specially designed to detect and ensure the quality of tobacco product packaging. It is crucial to accurately and quickly detect whether the cigarette box is intact, stain-free, and free of printing errors. The cigarette box appearance inspection system usually consists of a high-resolution camera, image processing software, light source control equipment, mechanical positioning devices, and automatic rejection or marking mechanisms. The cigarette box appearance inspection system uses machine vision technology to capture images of the cigarette box surface and identifies various defects through advanced image analysis algorithms. When actually applied to high-speed production lines, the cigarette box appearance inspection system based on computer vision inspection technology has higher production accuracy and efficiency, can quickly capture high-resolution images, and with advanced image processing algorithms, can achieve accurate detection of tiny defects. Highly automated control inspection reduces interference from human factors, improves the consistency and reliability of inspection, and the system can be adjusted according to different products and quality standards to adapt to diverse production needs. The system's data management capabilities help trace product quality and optimize the production process. With the continuous advancement of artificial intelligence and machine vision technology, the cigarette box appearance inspection system, as an efficient, accurate and reliable automated visual inspection solution, can not only improve the automation level of the cigarette box production line, but also ensure product quality, reduce labor costs, and bring significant economic benefits to enterprises. It will play an increasingly important role in the future manufacturing industry.

[0003] At present, traditional cigarette box appearance inspection systems usually use rolling shutter cameras to continuously shoot cigarette boxes passing through the production line. However, in actual use, when shooting fast-moving objects, especially in dark environments, the use of rolling shutters may cause partial exposure, slope graphics, or shaking, resulting in jello effects, etc., which can easily cause the collected images to be deformed and blurred, and cannot accurately and reliably identify defects in the appearance of cigarette boxes, resulting in low detection accuracy and poor traceability. Summary of the invention

[0004] In view of this, the present invention provides a method, system and device for detecting appearance of a cigarette box based on a global shutter, so as to solve the problem of low accuracy in identifying defects of the appearance of a cigarette box in the prior art.

[0005] In a first aspect, the present invention provides a method for detecting the appearance of a smoke box based on a global shutter, the method comprising: Obtain the outbound data set, appearance data set, and standard data set of cigarette boxes; Preprocess the outbound data set and the appearance data set to obtain a cigarette box image group, and determine the characteristic normal data group and the tolerance compensation index based on the cigarette box image group and the standard data set; The appearance status of the cigarette box is judged according to the characteristic normal data group and the tolerance compensation index, and a cigarette box appearance inspection report is generated.

[0006] The global shutter-based cigarette box appearance inspection method provided by the present invention obtains the outgoing data set, appearance data set and standard data set of the cigarette box, determines the characteristic normal data group and the tolerance compensation index, and then accurately judges the appearance status of the cigarette box, improves the recognition accuracy of the appearance defects of the cigarette box, generates an inspection report for the appearance status of the cigarette box, and improves the traceability of the appearance inspection of the cigarette box.

[0007] In an optional implementation, obtaining a cigarette box outbound data set, an appearance data set, and a standard data set includes: Acquire multiple outbound data, and number the outbound data of the cigarette boxes according to the link at which the cigarette boxes are outbound, to obtain an outbound data set; Using a global shutter camera to continuously shoot the cigarette box according to a preset shooting cycle to obtain a plurality of appearance data of the cigarette box, and numbering the plurality of appearance data according to a preset appearance numbering rule to form an appearance data set; According to multiple preset measurement standards for smoke box appearance inspection, multiple standard data are numbered to form a standard data set.

[0008] The global shutter-based cigarette box appearance detection method provided by the present invention utilizes the global shutter to capture and obtain appearance data, thereby reducing phenomena such as partial exposure, slope graphics or jello effect caused by shaking, and the captured appearance data is clearer, which is suitable for capturing dynamic scenes with high synchronization requirements, thereby improving the accuracy of appearance detection.

[0009] In an optional implementation, the outbound data set and the appearance data set are preprocessed to obtain a cigarette box image group, including: Classify the appearance data set according to the number of the outbound data set to obtain multiple appearance data corresponding to different outbound data; Each outgoing warehouse data is combined with the corresponding multiple appearance data to obtain multiple cigarette box image data as a cigarette box image group.

[0010] The global shutter-based cigarette box appearance detection method provided by the present invention classifies the appearance data of the cigarette boxes according to the cigarette box outbound link to obtain a cigarette box image group, thereby avoiding large errors in the detection results caused by different detection standards in different outbound links and improving the accuracy of cigarette box appearance detection.

[0011] In an optional implementation, determining a characteristic normal data group according to the smoke box image group and the standard data set includes: The smoke box image group is measured according to each standard data in the standard data set, and the smoke box image data that meets all the standard data is screened out as the feature normal data group.

[0012] The global shutter-based cigarette box appearance detection method provided by the present invention analyzes and measures each cigarette box image in each outbound link to determine the cigarette box image data that fully meets the standards. It can quickly screen out cigarette box image data with normal appearance, avoid missed detection and false detection in a single outbound link, and improve the accuracy of cigarette box appearance detection.

[0013] In an optional implementation, determining a tolerance compensation index according to the smoke box image group and the standard data set includes: Input multiple standard data in the standard data set into a preset deep learning model to obtain a fusion standard; The sample variance between the normal image data and the fusion standard in the feature normal data group is calculated as the tolerance compensation index.

[0014] The global shutter-based cigarette box appearance detection method provided by the present invention further analyzes the degree of deviation between the appearance images of a specified batch of cigarette boxes and the standard appearance images by calculating the tolerance compensation index, thereby avoiding a too high unqualified rate due to a too small measurement standard range, affecting the normal delivery of qualified cigarette boxes, and facilitating the dynamic management of the delivery of cigarette boxes.

[0015] In an optional implementation, judging the appearance state of the cigarette box according to the characteristic normal data group and the tolerance compensation index includes: If the ratio of the number of normal images in the characteristic normal data group to the total number of smoke box image data in the smoke box image group is less than a preset threshold, or the tolerance compensation index is not less than a preset index, the appearance state of the smoke box is determined to be unqualified.

[0016] In an optional implementation, judging the appearance state of the cigarette box according to the characteristic normal data group and the tolerance compensation index further includes: If the ratio of the number of normal images in the characteristic normal data group to the total number of smoke box image data in the smoke box image group is not less than a preset threshold, and the tolerance compensation index is less than a preset index, the appearance state of the smoke box is determined to be qualified.

[0017] The global shutter-based cigarette box appearance detection method provided by the present invention determines the appearance status of the cigarette box from two aspects: the characteristic normal data group and the tolerance compensation index. The determination result of whether the cigarette box appearance is qualified or unqualified is more accurate, thereby improving the accuracy of cigarette box appearance detection.

[0018] In a second aspect, the present invention provides a cigarette box appearance detection system, the system comprising: A multi-dimensional data acquisition module is used to acquire the outbound data set, appearance data set and standard data set of the cigarette box; The defect analysis module is used to pre-process the outbound data set and the appearance data set to obtain the cigarette box image group, and determine the characteristic normal data group and the tolerance compensation index based on the cigarette box image group and the standard data set; The data monitoring management module is used to judge the appearance status of the cigarette box according to the characteristic normal data group and the tolerance compensation index, and generate a cigarette box appearance inspection report.

[0019] In a third aspect, the present invention provides a computer device, comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the method of the first aspect or any corresponding embodiment thereof by executing the computer instructions.

[0020] In a fourth aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to cause a computer to execute the method of the first aspect or any corresponding embodiment thereof. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0022] Figure 1 is a flow chart of a method for detecting the appearance of a smoke box based on a global shutter according to an embodiment of the present invention; Figure 2 is a flow chart of another method for detecting the appearance of a smoke box based on a global shutter according to an embodiment of the present invention; Figure 3 is a structural block diagram of a smoke box appearance detection system based on a global shutter according to an embodiment of the present invention; Figure 4 It is a schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0023] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the 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] The embodiments of the present invention provide a method, system and device for cigarette box appearance detection based on a global shutter. By acquiring the outbound data set, appearance data set and standard data set of the cigarette box, the characteristic normal data group and the tolerance compensation index are determined, and then the appearance status of the cigarette box is accurately judged and a detection report is generated, so as to achieve the effect of improving the accuracy and traceability of cigarette box appearance detection.

[0025] According to an embodiment of the present invention, an embodiment of a smoke box appearance detection method based on a global shutter is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0026] In this embodiment, a smoke box appearance detection method based on a global shutter is provided, which can be used in the above-mentioned computer system. Figure 1 FIG. 1 is a flow chart of a method for detecting the appearance of a smoke box based on a global shutter according to an embodiment of the present invention. Figure 1 As shown, the process includes the following steps: Step S101, obtaining a cigarette box outbound data set, an appearance data set, and a standard data set.

[0027] Specifically, the central control system stores the outbound data set and the standard data set in advance, and can directly obtain the outbound data set and the standard data set from the database of the central control system through the network connection. The outbound data in the outbound data set represents the outbound link in the outbound process of the cigarette box, and the standard data in the standard data set represents multiple standard data of the appearance detection of the cigarette box in different outbound links. The appearance data in the appearance data set represents the appearance pictures of the cigarette box collected by the image acquisition device during the outbound process of the cigarette box.

[0028] Step S102 , preprocessing the outgoing data set and the appearance data set to obtain a cigarette box image group, and determining a characteristic normal data group and a tolerance compensation index based on the cigarette box image group and the standard data set.

[0029] Specifically, the outbound link is determined according to the difference of outbound data in the outbound data set, and the appearance data in the appearance data set is classified and processed according to the outbound link to obtain the appearance data under different outbound links as the cigarette box image group.

[0030] The cigarette box images in the cigarette box image group are classified according to the outbound link, and are used to characterize the actual collected cigarette box appearance data. The features of the cigarette box images in the cigarette box image group are judged respectively according to multiple standard data in the standard data set to determine whether all the cigarette box images meet the multiple standard data in the standard data set. If all meet the requirements, the corresponding cigarette box images are used as feature normal data. According to this step, multiple feature normal data in the cigarette box image group are screened out to form a feature normal data group.

[0031] The tolerance compensation index is calculated using the characteristic normal data set and the standard data set according to the preset tolerance compensation index calculation formula.

[0032] Step S103, judging the appearance state of the cigarette box according to the characteristic normal data group and the tolerance compensation index, and generating a cigarette box appearance inspection report.

[0033] Specifically, the degree of deviation between the features of the specified batch of cigarette box appearance images and the standard data set is further analyzed based on the feature normal data group and the tolerance compensation index, and whether the appearance status of the cigarette boxes is qualified is judged based on the degree of deviation, and a test report is generated based on the test results of the appearance status of the cigarette boxes. The specific form of the test report is not limited here.

[0034] The global shutter-based cigarette box appearance inspection method provided in this embodiment obtains the outgoing data set, appearance data set and standard data set of the cigarette box, determines the characteristic normal data group and the tolerance compensation index, and then accurately judges the appearance status of the cigarette box, thereby improving the recognition accuracy of the appearance defects of the cigarette box, generating an inspection report for the appearance status of the cigarette box, and improving the traceability of the appearance inspection of the cigarette box.

[0035] In this embodiment, a smoke box appearance detection method based on a global shutter is provided, which can be used in the above-mentioned computer system. Figure 2 FIG. 1 is a flow chart of a method for detecting the appearance of a smoke box based on a global shutter according to an embodiment of the present invention. Figure 2 As shown, the process includes the following steps: Step S201, obtaining a cigarette box outbound data set, an appearance data set, and a standard data set.

[0036] Specifically, the above step S201 includes: Step S2011, obtaining a plurality of warehouse-out data, and numbering the warehouse-out data of the cigarette boxes according to the link at which the cigarette boxes are shipped out, to obtain a warehouse-out data set.

[0037] Specifically, the outbound links in the process of cigarette box outbound include: the outbound link of finished cigarette pallets, the outbound link of manual pallet removal, the return link of empty pallets and the temporary storage of tail pallets. The logistics transportation data of each outbound link when the cigarette box is outbound is numbered to obtain the outbound data. For example, the outbound link of finished cigarette pallets is numbered as , the manual unpacking and delivery link number is , the link number of empty pallet return to warehouse is , the number of the temporary storage processing link for the last trading is , the outbound data set includes: , is only used as an example, but is not limited to this.

[0038] Step S2012, using a global shutter camera to continuously shoot the cigarette box according to a preset shooting cycle to obtain a plurality of appearance data of the cigarette box, and numbering the plurality of appearance data according to a preset appearance numbering rule to form an appearance data set.

[0039] Specifically, a network global shutter camera is used to continuously shoot cigarette boxes in the process of leaving the warehouse according to a preset shooting cycle to obtain multiple continuous images as appearance data. The multiple continuous images are numbered according to the order of shooting to obtain the appearance data set. . The appearance data in the appearance dataset is obtained by continuously shooting the cigarette box with a global shutter camera according to a preset shooting cycle. The shooting frequency, i.e. the preset shooting cycle, determines the time interval for data collection. The captured appearance data are assigned different numbers, which are usually based on the order of shooting rather than a specific time point. Therefore, the numbers of different appearance data represent different shooting times. In the process of cigarette box outbound delivery, each outbound delivery link is equipped with a global shutter camera. When the cigarette box passes through each outbound delivery link, the corresponding global shutter camera will shoot the cigarette box to record the status of the cigarette box in that link. For the whole pallet delivery of finished cigarettes, the image of the cigarette box appearance is the whole pallet, and the appearance of the cigarette box detected includes the pallet and the cigarette box on the pallet; for the manual disassembly of the pallet delivery, the image of the cigarette box appearance is the non-full pallet, and the appearance of the cigarette box detected includes the pallet and the cigarette box on the pallet; for the empty pallet return to the warehouse, the image of the cigarette box appearance is the pallet, and the appearance of the cigarette box detected includes the pallet; for the temporary storage of the last pallet, the image of the cigarette box appearance is the last pallet, and the appearance of the cigarette box detected includes the pallet and the cigarette box on the pallet. Global shutter technology allows all pixels on the image sensor to start and end exposure at the same time, which means that the entire picture is captured at the same time, thereby avoiding the distortion or blur that may occur when shooting fast-moving objects. Global shutter technology clearly captures the appearance of fast-moving cigarette boxes by synchronizing the exposure time of all pixels, ensuring the validity of the data.

[0040] Step S2013: numbering the plurality of standard data according to the plurality of preset measurement standards for the cigarette box appearance inspection to form a standard data set.

[0041] Specifically, for the appearance inspection of cigarette boxes, it is necessary to analyze and judge multiple features of the acquired appearance image data. Each feature has a corresponding measurement standard as standard data. The measurement standards of multiple features are numbered to form a standard data set. .

[0042] The global shutter-based smoke box appearance detection method provided in this embodiment uses a global shutter to capture appearance data, reducing phenomena such as partial exposure, slope graphics or jello effect caused by shaking. The captured appearance data is clearer and is suitable for capturing dynamic scenes with high synchronization requirements, thereby improving the accuracy of appearance detection.

[0043] Step S202 , preprocessing the outgoing data set and the appearance data set to obtain a cigarette box image group, and determining a characteristic normal data group and a tolerance compensation index based on the cigarette box image group and the standard data set.

[0044] Specifically, the above step S202 includes: Step S2021 , classifying the appearance data set according to the serial number of the outbound data set to obtain a plurality of appearance data corresponding to different outbound data.

[0045] Specifically, the appearance data in the appearance data set are classified according to the number of the outbound data in the outbound data set, and all the appearance data are divided into appearance data of different outbound links. For example, for the outbound data numbered The appearance data of the finished product cigarette pallet out of the warehouse can be represented by its number ; For outbound data number The appearance data in the manual unpacking and delivery stage can be represented by its number: ; For outbound data number The appearance data of the empty pallet in the return link can be represented by its number ; For outbound data number The appearance data in the temporary storage processing stage of the last batch can be represented by its number .

[0046] Step S2022 , combining each shipment data with the corresponding multiple appearance data respectively to obtain multiple cigarette box image data as a cigarette box image group.

[0047] Specifically, the intersection of the library data and the corresponding multiple appearance data is calculated using the following formula to obtain the cigarette box image data:

[0048] in, represents the smoke box image group, The logistics and transportation data of the finished product pallet out of the warehouse match the image data captured by the global shutter camera. It indicates that the logistics and transportation data of the manual unpacking and unloading of cigarette boxes matches the image data captured by the global shutter camera. It means matching the logistics transportation data of the empty pallet returning to the warehouse when the cigarette box is out of the warehouse with the image data captured by the global shutter camera. It means matching the logistics and transportation data of the temporary storage processing link of the last batch when the cigarette box is shipped out of the warehouse with the image data captured by the global shutter camera.

[0049] For the smoke box images in the smoke box image group, renumber them and express them as , where i represents the outbound link of the cigarette box, that is, the outbound data; n represents the nth appearance image taken corresponding to the outbound link. The cigarette box image group can be expressed as:

[0050] By using the cigarette box image group Yxtx, it is possible to quickly classify the high-definition images of each cigarette box in different warehouse delivery links, so as to facilitate the subsequent targeted evaluation of the appearance integrity, consistency and aesthetics of the cigarette box from aspects such as appearance edges, colors or textures. For example, edge detection can be used to identify whether there are damages, cracks or uneven areas; color detection can verify whether there are color differences, code segment accuracy or printing errors; texture analysis can be used to check whether there are abnormal bumps, foreign objects or stains, as an example only, but not limited to this.

[0051] The global shutter-based cigarette box appearance detection method provided in this embodiment classifies the appearance data of the cigarette boxes according to the cigarette box outbound link to obtain a cigarette box image group, thereby avoiding large errors in the detection results caused by different detection standards in different outbound links and improving the accuracy of cigarette box appearance detection.

[0052] Step S2023 , the smoke box image group is measured according to each standard data in the standard data set, and the smoke box images that meet all the standard data are screened out as the characteristic normal data group.

[0053] Specifically, each cigarette box image in the cigarette box image group is measured according to each standard data in the standard data set, and the cigarette box images that meet all the standard data are screened out as the characteristic normal data group.

[0054]

[0055] in, represents a characteristic normal data set, It represents multiple standard data in the outbound link i for the appearance inspection of cigarette boxes. It means that the image data of 1-n cigarette boxes in the outbound link i are measured by feature analysis using multiple standard data, and multiple cigarette box image data of each outbound link that meets all the standard data are screened out.

[0056] The measurement standards in different outbound links may be different, because the status of the cigarette boxes and the focus of attention in each link may change, so the standard data in the standard data set of different outbound links may be different. For example, the outbound link of the finished cigarette pallet may pay more attention to the overall condition of the pallet and the cigarette box on it; the manual disassembly of the pallet outbound link will pay more attention to the appearance of a single cigarette box; the empty pallet return link will pay more attention to the integrity of the pallet itself, whether the QR code on the pallet is damaged and whether the chip in the QR code is intact; the tail pallet storage processing link needs to check whether the tail pallet cigarette box meets the quality standards before storage.

[0057] The multiple smoke box image data in the feature normal data group need to be renumbered, which can be expressed as , where k represents the outbound link of the cigarette box, that is, the outbound data; n represents the nth cigarette box image corresponding to the outbound link. The characteristic normal data group can be expressed as: .

[0058] The global shutter-based cigarette box appearance detection method provided in this embodiment analyzes and measures each cigarette box image in each outbound link to determine the cigarette box image data that fully meets the standards. This can quickly screen out cigarette box image data with normal appearance, avoid missed detection and false detection in a single outbound link, and improve the accuracy of cigarette box appearance detection.

[0059] Step S2024, input multiple standard data in the standard data set into a preset deep learning model to obtain a fusion standard.

[0060] Specifically, the formula of the preset deep learning model is expressed as:

[0061] Among them, RH represents the fusion standard.

[0062] Step S2025, calculating the sample variance between the normal image data in the characteristic normal data group and the fusion standard as a tolerance compensation index.

[0063] Specifically, the formula for calculating the tolerance compensation index is:

[0064] in, represents the tolerance compensation index, Represents the difference between normal image data and fusion standard, Represents the sample variance of normal image data and fusion standard.

[0065] The global shutter-based cigarette box appearance detection method provided in this embodiment further analyzes the degree of deviation between the appearance image of a specified batch of cigarette boxes and the standard appearance image by calculating the tolerance compensation index, so as to avoid the unqualified rate being too high due to the small range of the measurement standard, which affects the normal delivery of qualified cigarette boxes, and facilitates the dynamic management of the delivery of cigarette boxes. For different batches of cigarette boxes, the measurement standards are unified in principle, unless there are special appearance quality requirements for the production feeder. For example, under normal circumstances, three strapping tapes are required, but the quality of a batch of cigarette boxes from a certain origin is poor and four strapping tapes are required, so the measurement standards may need to be adjusted accordingly.

[0066] Step S203 , judging the appearance state of the cigarette box according to the characteristic normal data group and the tolerance compensation index, and generating a cigarette box appearance inspection report.

[0067] Specifically, the above step S203 includes: Step S2031: if the ratio of the number of normal images in the characteristic normal data group to the total number of cigarette box image data in the cigarette box image group is less than a preset threshold, or the tolerance compensation index is not less than a preset index, the appearance state of the cigarette box is determined to be unqualified.

[0068] Specifically, in the process of judging the appearance of the cigarette box, the normal data group and the tolerance compensation index are combined to make a judgment and generate a corresponding inspection report. Due to the influence of the shooting environment or other objective factors, the images in the cigarette box image group may have large errors, so a preset threshold value of the ratio of the number of normal images to the total number of images is set, that is, if the ratio of the number of normal images in the normal data group to the total number of cigarette box image data in the cigarette box image group is less than the preset threshold value, the appearance of the cigarette box is judged to be unqualified. For example, if the preset threshold value is 90%, the number of normal images in the normal data group is 88, the total number of cigarette box image data is 100, and the ratio of the number of normal images to the total number of images is 88%, which is less than 90%, then the appearance of the cigarette box is judged to be unqualified.

[0069] The larger the tolerance compensation index is, the greater the deviation between the number of normal images in the feature normal data group and the fusion standard is. Therefore, when the tolerance compensation index is not less than the preset index, the appearance of the cigarette box is judged to be unqualified.

[0070] The unqualified appearance of the cigarette box means that there are damages or stains on the appearance of the cigarette box. The unqualified cigarette boxes need to be intercepted and an unqualified inspection report needs to be generated.

[0071] Step S2032: if the ratio of the number of normal images in the characteristic normal data group to the total number of cigarette box image data in the cigarette box image group is not less than a preset threshold, and the tolerance compensation index is less than a preset index, the appearance state of the cigarette box is determined to be qualified.

[0072] Specifically, when the ratio of the number of normal images in the feature normal data group to the total number of cigarette box image data in the cigarette box image group is not less than a preset threshold, and the tolerance compensation index is less than a preset index, when both conditions are met, the appearance state of the cigarette box is judged to be qualified. At this time, the cigarette box is shipped normally without interception, and a qualified inspection report is generated.

[0073] Regardless of whether the inspection result of the appearance of the smoke box is qualified or unqualified, a test report will be generated and stored for backup, improving the traceability of the smoke box appearance inspection process.

[0074] The global shutter-based cigarette box appearance inspection method provided in this embodiment determines the appearance status of the cigarette box from two aspects: the characteristic normal data group and the tolerance compensation index. The determination result of whether the cigarette box appearance is qualified or unqualified is more accurate, thereby improving the accuracy of cigarette box appearance inspection, generating an inspection report for the cigarette box appearance status, and improving the traceability of cigarette box appearance inspection.

[0075] In this embodiment, a smoke box appearance detection system based on a global shutter is also provided, which is used to implement the above-mentioned embodiments and preferred implementation modes, and the descriptions that have been made will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware of a predetermined function. Although the system described in the following embodiments is preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0076] This embodiment provides a smoke box appearance detection system based on a global shutter, such as Figure 3 As shown, including: The multi-dimensional data acquisition module 301 is used to acquire the outbound data set, appearance data set and standard data set of the cigarette box.

[0077] The defect analysis module 302 is used to pre-process the outgoing data set and the appearance data set to obtain a cigarette box image group, and determine a characteristic normal data group and a tolerance compensation index based on the cigarette box image group and the standard data set.

[0078] The data monitoring management module 303 is used to determine the appearance status of the cigarette box according to the characteristic normal data group and the tolerance compensation index, and generate a cigarette box appearance detection report.

[0079] In some optional implementations, the multidimensional data acquisition module 301 includes: The outbound data numbering unit is used to obtain multiple outbound data and number the outbound data of the cigarette boxes according to the link at which the cigarette boxes are outbound, so as to obtain an outbound data set.

[0080] The appearance data numbering unit is used to continuously shoot the cigarette box with a global shutter camera according to a preset shooting cycle to obtain multiple appearance data of the cigarette box, and to form an appearance data set after numbering the multiple appearance data according to a preset appearance numbering rule.

[0081] The standard data numbering unit is used to number a plurality of standard data according to a plurality of preset measurement standards for smoke box appearance inspection to form a standard data set.

[0082] In some optional implementations, the defect analysis module 302 includes: The appearance data preprocessing unit is used to classify the appearance data set according to the serial number of the outbound data set to obtain multiple appearance data corresponding to different outbound data.

[0083] The cigarette box image generating unit is used to respectively combine each outgoing data with the corresponding plurality of appearance data to obtain a plurality of cigarette box image data as a cigarette box image group.

[0084] The image weighing unit is used to weigh the smoke box image group according to each standard data in the standard data set, and screen out the smoke box image data that meets all the standard data as the characteristic normal data group.

[0085] The standard fusion unit is used to input multiple standard data in the standard data set into a preset deep learning model to obtain a fusion standard.

[0086] The variance calculation unit is used to calculate the sample variance of the normal image data in the feature normal data group and the fusion standard as a tolerance compensation index.

[0087] In some optional implementations, the data monitoring management module 303 includes: The first state determination unit is used to determine that the appearance state of the cigarette box is unqualified if the ratio of the number of normal images in the characteristic normal data group to the total number of cigarette box image data in the cigarette box image group is less than a preset threshold, or the tolerance compensation index is not less than a preset index.

[0088] The second state determination unit is used to determine that the appearance state of the cigarette box is qualified if the ratio of the number of normal images in the characteristic normal data group to the total number of cigarette box image data in the cigarette box image group is not less than a preset threshold, and the tolerance compensation index is less than a preset index.

[0089] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.

[0090] The global shutter-based smoke box appearance inspection system in this embodiment is presented in the form of a functional unit, where the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and memory that executes one or more software or fixed programs, and / or other devices that can provide the above functions.

[0091] The embodiment of the present invention also provides a computer device having the above Figure 3 The smoke box appearance inspection system based on global shutter is shown.

[0092] See also Figure 4 , Figure 4 is a schematic diagram of the structure of a computer device provided by an optional embodiment of the present invention, such as Figure 4 As shown, the computer device includes: one or more processors 10, a memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components are connected to each other using different buses for communication, and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Similarly, multiple computer devices can be connected, and each device provides some necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 4 A processor 10 is taken as an example.

[0093] The processor 10 may be a central processing unit, a network processor or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be a dedicated integrated circuit, a programmable logic device or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic or any combination thereof.

[0094] The memory 20 stores instructions executable by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiment.

[0095] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created according to the use of the computer device, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely arranged relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0096] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid state drive; the memory 20 may also include a combination of the above types of memory.

[0097] The computer device further comprises a communication interface 30 for the computer device to communicate with other devices or a communication network.

[0098] The embodiment of the present invention also provides a computer-readable storage medium. The method according to the embodiment of the present invention can be implemented in hardware, firmware, or can be implemented as a computer code that can be recorded in a storage medium, or can be implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and will be stored in a local storage medium through a network download, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state hard disk, etc.; further, the storage medium can also include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor, or hardware, the method shown in the above embodiment is implemented.

[0099] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A smoke box appearance inspection method based on global shutter, It is characterized in that The method comprises: Obtain the outbound data set, appearance data set, and standard data set of cigarette boxes; Preprocessing the outgoing data set and the appearance data set to obtain a cigarette box image group, and determining a feature normal data group and a tolerance compensation index based on the cigarette box image group and the standard data set; The appearance status of the cigarette box is judged according to the characteristic normal data group and the tolerance compensation index, and a cigarette box appearance inspection report is generated.

2. The method according to claim 1, It is characterized in that The step of obtaining the outbound data set, appearance data set and standard data set of the cigarette boxes includes: Acquire multiple outbound data, and number the outbound data of the cigarette boxes according to the link at which the cigarette boxes are outbound, to obtain an outbound data set; Continuously photographing the cigarette box according to a preset shooting cycle using a global shutter camera to obtain a plurality of appearance data of the cigarette box, and numbering the plurality of appearance data according to a preset appearance numbering rule to form an appearance data set; According to multiple preset measurement standards for smoke box appearance inspection, multiple standard data are numbered to form a standard data set.

3. The method according to claim 2, It is characterized in that The outbound data set and the appearance data set are preprocessed to obtain a cigarette box image group, including: Classifying the appearance data set according to the serial number of the outbound data set to obtain a plurality of appearance data corresponding to different outbound data; Each outgoing warehouse data is combined with the corresponding multiple appearance data to obtain multiple cigarette box image data as a cigarette box image group.

4. The method according to claim 1, It is characterized in that Determining a characteristic normal data group according to the smoke box image group and the standard data set includes: The smoke box image group is measured according to each standard data in the standard data set, and the smoke box image data that meets all the standard data is screened out as the characteristic normal data group.

5. The method according to claim 1, It is characterized in that Determining a tolerance compensation index according to the smoke box image group and the standard data set includes: Inputting multiple standard data in the standard data set into a preset deep learning model to obtain a fusion standard; The sample variance between the normal image data and the fusion standard in the feature normal data group is calculated as the tolerance compensation index.

6. The method according to claim 1, It is characterized in that Judging the appearance state of the cigarette box according to the characteristic normal data group and the tolerance compensation index includes: If the ratio of the number of normal images in the characteristic normal data group to the total number of cigarette box image data in the cigarette box image group is less than a preset threshold, or the tolerance compensation index is not less than a preset index, the appearance state of the cigarette box is determined to be unqualified.

7. The method according to claim 1, It is characterized in that Judging the appearance state of the cigarette box according to the characteristic normal data group and the tolerance compensation index also includes: If the ratio of the number of normal images in the characteristic normal data group to the total number of cigarette box image data in the cigarette box image group is not less than a preset threshold, and the tolerance compensation index is less than a preset index, the appearance state of the cigarette box is determined to be qualified.

8. A smoke box appearance inspection system based on global shutter, It is characterized in that The system comprises: A multi-dimensional data acquisition module is used to acquire the outbound data set, appearance data set and standard data set of the cigarette box; A defect analysis module is used to pre-process the outgoing data set and the appearance data set to obtain a cigarette box image group, and determine a feature normal data group and a tolerance compensation index based on the cigarette box image group and the standard data set; The data monitoring management module is used to judge the appearance status of the cigarette box according to the characteristic normal data group and the tolerance compensation index, and generate a cigarette box appearance detection report.

9. A computer device, It is characterized in that include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the method according to any one of claims 1 to 7 by executing the computer instructions.

10. A computer-readable storage medium, It is characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the method according to any one of claims 1 to 7.