Visual detection method and system for marked characters and storage medium

Through visual detection methods and intelligent defect detection algorithms, the problem of determining defects of marking characters in laser marking technology is solved, efficient and accurate automated detection is achieved, production efficiency and accuracy are improved, and costs are reduced.

CN119023675APending Publication Date: 2024-11-26ZHEJIANG ANSHENG TECH CO LTD +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202411115796.5
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Priority Date
2024-03-08
Filing Date
2024-08-14
Publication Date
2024-11-26

AI Technical Summary

Technical Problem

The existing laser marking technology is difficult to determine the defects of marking characters, resulting in low production efficiency, high cost, and inability to meet the needs of batch production.

Method used

By using visual detection method, the integrity and offset of marking characters are detected by obtaining the image of the steel bottom of the thermos cup, and edge similarity detection, black hat algorithm, ring mask algorithm, RANSAC algorithm and circular caliper algorithm are used, combining intelligence Defect detection algorithm realizes automatic detection of marking characters.

Benefits of technology

It improves inspection efficiency and accuracy, reduces labor costs, realizes rapid change of modeling and accurate inspection, meets the production needs of products of different specifications, and reduces time and financial costs.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119023675A_ABST
    Figure CN119023675A_ABST
Patent Text Reader

Abstract

The invention provides a visual detection method for marking characters. The visual detection method comprises the following steps: S1, marking the marking characters on a steel bottom of a vacuum cup; s2, acquiring an image of the steel bottom of the vacuum cup marked with the marking characters, and detecting whether the steel bottom of the vacuum cup has the marking characters or not; if the steel bottom of the vacuum cup is provided with the marking characters, entering the step S3, otherwise, ending visual detection of the marking characters on the steel bottom of the vacuum cup; s3, detecting whether the marked characters on the steel bottom of the vacuum cup are complete or not; if the marked characters on the steel bottom of the vacuum cup are complete, entering the step S4, otherwise, ending visual detection of the marked characters on the steel bottom of the vacuum cup; and S4, detecting whether the marked characters on the steel bottom of the vacuum cup deviate or not, and finishing visual detection of the marked characters on the steel bottom of the vacuum cup. Under the condition of the same time complexity, an intelligent defect detection algorithm is applied to defect detection of vacuum cup laser marking, the defect detection is carried out by being compatible with a mainstream industrial camera, and the actual precision of customers can be met.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of laser marking detection technology, and in particular to a method, system and storage medium for visually detecting marked characters. Background Art

[0002] The main purpose of laser marking technology is to leave a permanent mark on the surface of an object. At present, laser marking technology has become a precision machining method and is widely used in various industries. During the laser marking process, the vapor formed by grease on the workpiece is easily condensed on the protective laser filter, which hinders the penetration of the laser and makes the laser marking ineffective. Therefore, how to detect defects in laser marked products and improve the quality of laser marked products is crucial.

[0003] In the prior art, Zhuang Jianrong, Yang Xiongjie and others put products in batches on the assembly line, and then transfer them to the bottom of the galvanometer and the marking camera through the assembly line. After the marking camera takes a picture of the product, the program in the computer host analyzes and captures the marking area, and transmits the capture result to the laser, and the galvanometer is used to laser mark the designated area. This type of product can only be marked with characters, and it is difficult to effectively determine the defects of the marked characters, which requires a lot of manpower for re-inspection. In addition, this type of product is manually loaded and unloaded, and cannot be loaded and unloaded in batches, which also affects the industrial production efficiency and cannot meet the mass production requirements. In addition, the computing performance is not very ideal, and it is only used for the production of products that do not require very high CT.

[0004] Shu Shubao, Jiang Yiping et al. proposed a method for detecting contamination defects on the surface of an optical chip, which includes collecting a surface image of a chip to be detected, performing defect extraction on the preprocessed image, and calculating defect feature values ​​of the extracted chip surface image samples; processing the surface image of a chip to which a known defect belongs to obtain defect feature values ​​of the chip surface samples, and using the defect feature values ​​of the category to which the defect of the chip sample belongs to perform processing to obtain chip surface defect feature values ​​of the category to which the known defect belongs; and then using a digital image processing algorithm to obtain the defect feature values ​​of the chip to be detected, and obtain the category to which the defect of the chip to be detected belongs.

[0005] Some marking equipment on the market is nothing more than changing the head but not the surface. The galvanometer and laser are installed under the conveyor belt, and the PLC is used to control the camera hard trigger. The product is placed on the assembly line and conveyed to the bottom of the galvanometer and marking camera through the assembly line. The marking camera takes a picture of the product and the program in the computer host analyzes and captures the marking area. The captured result is transmitted to the laser, and the designated area is laser marked through the galvanometer. The assembly line continues to convey, that is, it can only mark pattern characters, and cannot detect defects in the characters.

[0006] At present, the methods for detecting surface defects of workpieces mainly focus on the grayscale characteristics of defects and the signal characteristics of images. Usually, the development of defect detection based on machine vision is targeted at specific products. Once the product to be inspected is replaced, or even when products of different specifications are replaced, reprogramming is required, which increases the time and financial costs and cannot meet the needs of actual production. In addition, the laser marking equipment has complex software configuration and simple hardware. It only uses the PLC controller to hard trigger the camera and take pictures of the product for laser marking. It is difficult to determine the defects of the marked characters, and additional personnel are required to check the marked characters. It is inconvenient to use and wastes a lot of manpower and material resources. Summary of the invention

[0007] The present invention provides a method, system and storage medium for visually detecting marked characters, so as to solve the technical problem that it is difficult to perform defect determination on marked characters in the existing laser marking.

[0008] The technical solution provided by the present invention is as follows:

[0009] An object of the present invention is to provide a method for visually detecting a marked character, the method comprising the following steps:

[0010] S1. Mark the steel bottom of the thermos cup;

[0011] S2, obtaining an image of the steel bottom of the thermos cup with marking characters, and detecting whether the steel bottom of the thermos cup has marking characters;

[0012] If the thermos cup steel bottom has marking characters, then proceed to step S3, otherwise the visual inspection of the marking characters on the thermos cup steel bottom ends;

[0013] S3. Check whether the marking characters on the steel bottom of the thermos cup are complete;

[0014] If the marking characters on the steel bottom of the thermos flask are complete, then the process proceeds to step S4, otherwise the visual inspection of the marking characters on the steel bottom of the thermos flask ends;

[0015] S4, detecting whether the marking characters on the steel bottom of the thermos flask are offset, and ending the visual inspection of the marking characters on the steel bottom of the thermos flask.

[0016] In a preferred embodiment, in step S2, whether the steel bottom of the thermos cup has marked characters is detected by the following method:

[0017] S21, performing edge similarity detection between the marking characters in the image of the steel bottom of the thermos cup with the marking characters and the marking characters of the standard product;

[0018] S22, according to the window search strategy, find the edge contour feature with the largest similarity score between the marking characters in the steel bottom image of the thermos cup with the marking characters and the marking characters of the standard product;

[0019] S23, determining the angle relationship between the marking characters in the image of the steel bottom of the thermos cup with the marking characters and the marking characters of the standard product by scaling and affine transformation;

[0020] Determine the angle value of the marking character in the image of the steel bottom of the thermos cup with marking characters according to the angle relationship between the marking character in the image of the steel bottom of the thermos cup with marking characters and the marking character of the standard product;

[0021] S24. When the angle value of the marking characters in the image of the thermos cup steel bottom with the marking characters engraved thereon does not exceed the preset threshold, it is determined that the thermos cup steel bottom has the marking characters, otherwise it is determined that the thermos cup steel bottom does not have the marking characters.

[0022] In a preferred embodiment, in step S3, the marking characters on the steel bottom of the thermos cup are detected to be complete by the following method:

[0023] S31, extracting each character outline of the marked characters from the image of the steel bottom of the thermos cup with the marked characters;

[0024] S32, removing noise points outside each character outline by using a black hat algorithm and a ring mask algorithm;

[0025] S33, performing feature matching on each character contour after removing noise points and each standard character contour in the standard product;

[0026] S34. Determine whether the marking characters on the steel bottom of the thermos cup are complete according to the maximum matching algorithm.

[0027] In a preferred embodiment, in step S4, the marking characters on the steel bottom of the thermos cup are detected to see if they are offset by the following method:

[0028] S41, extracting the gravure edge of the steel bottom of the thermos cup from the image of the steel bottom of the thermos cup with the marking characters engraved thereon;

[0029] S42, performing regional pixel enhancement on the concave printing edge of the steel bottom of the thermos cup;

[0030] S43, using the RANSAC algorithm to fit the gravure edge after pixel enhancement, and generate a gravure edge fitting circle;

[0031] S44, extracting the outline of the marking character from the image of the steel bottom of the thermos cup with the marking character, and using the angle discrimination method to filter out the outermost point set of the marking character outline;

[0032] S45, using a circular caliper algorithm tool to fit the outermost point set of the marking character contour to generate a marking character contour circle;

[0033] S46, comparing the distance between the center of the gravure edge fitting circle and the center of the marking character outline circle;

[0034] If the distance between the center of the gravure edge fitting circle and the center of the marking character outline circle is greater than a preset threshold, it is determined that the marking characters on the steel bottom of the thermos flask are offset; otherwise, the marking characters on the steel bottom of the thermos flask are not offset.

[0035] Another object of the present invention is to provide a visual detection system for marked characters, wherein the visual detection system is used to execute a visual detection method for marked characters provided by the present invention.

[0036] Another object of the present invention is to provide a computer storage medium, characterized in that the computer storage medium is used to store computer execution instructions; the computer execution instructions are used to execute a visual detection method for marked characters provided by the present invention.

[0037] The above technical solution of the present invention has at least the following beneficial effects compared with the prior art:

[0038] The present invention provides a method, system and storage medium for visual detection of marked characters, which requires setting algorithm parameters, has low cost, high detection efficiency and is easy to use. Compared with manual marking and manual quality inspection, the performance of the present invention is improved by 1 / 3, and the changeover time is reduced from 20 minutes to 1 minute, which truly achieves cost reduction and efficiency improvement.

[0039] The present invention provides a method, system and storage medium for visual inspection of marked characters. Under the condition of the same time complexity, an intelligent defect detection algorithm is applied to defect detection of laser marking on thermos cups. The system is compatible with mainstream industrial cameras for defect detection and can meet the actual accuracy of customers.

[0040] The present invention provides a method, system and storage medium for visually detecting marked characters, which are applicable to products of different specifications. When products of different specifications are replaced, reprogramming is not required, thereby reducing time cost and financial cost.

[0041] The present invention provides a visual detection method, system and storage medium for marking characters, an adaptive edge enhancement and feature extraction algorithm, which can learn the gravure circle features of the steel bottom of a thermos cup according to products of different specifications, and can accurately locate the center of the circle of the steel bottom of the thermos cup even if the gravure circle features are not obvious. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0043] Figure 1 The present invention is a flowchart of a method for visually detecting marked characters.

[0044] Figure 2 Schematic diagram of the arrangement of visual inspection cameras in one embodiment of the present invention.

[0045] Figure 3 It is a schematic diagram of marking characters on the steel bottom of a thermos cup in one embodiment of the present invention.

[0046] Figure 4 It is a schematic diagram of detecting that the steel bottom of a thermos cup has marked characters in one embodiment of the present invention.

[0047] Figure 5 It is a schematic diagram of detecting that the marking characters on the steel bottom of a thermos cup are offset in one embodiment of the present invention.

[0048] Figure 6 It is a schematic diagram showing that the marking characters on the steel bottom of a thermos cup detected in one embodiment of the present invention conform to the standard. DETAILED DESCRIPTION

[0049] In order to make the purpose, technical solution and advantages of the embodiment of the present invention clearer, the technical solution of the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings of the embodiment of the present invention. Obviously, the described embodiment is a part of the embodiment of the present invention, not all of the embodiments. Based on the described embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0050] Unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. Similarly, words such as "one", "one" or "the" do not indicate a quantitative limitation, but indicate the existence of at least one. Words such as "include" or "comprise" mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. Words such as "connect" or "connected" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect.

[0051] It should be noted that the terms "up", "down", "left", "right", "front", "back", etc. used in the present invention are only used to indicate relative position relationships. When the absolute position of the object being described changes, the relative position relationship may also change accordingly.

[0052] Combination Figure 1 According to an embodiment of the present invention, a method for visually detecting a marked character is provided, comprising the following method steps:

[0053] Step S1, marking characters on the steel bottom of the thermos cup.

[0054] The present invention uses a laser to mark the steel bottom of the thermos cup, thereby marking characters. Figure 2 As shown, a visual inspection camera 1 is arranged above the conveyor belt 4 of the assembly line. The visual inspection camera 1 adopts a 5-megapixel black-and-white industrial camera as an imaging system. The visual inspection camera 1 is equipped with a high-definition industrial lens 2.

[0055] An open parallel light source 3 is arranged under the high-definition industrial lens 2 to reduce the texture interference and reflection interference of the G steel bottom of the thermos cup, reduce the noise in the imaging, and facilitate the extraction of marking character features.

[0056] The steel bottom G of the thermos cup on the conveyor belt 4 of the assembly line is marked by a laser (not shown in the figure), so that marking characters are engraved on the steel bottom G of the thermos cup.

[0057] In one embodiment, the laser uses a galvanometer laser to mark the thermos cup steel bottom G. In another embodiment, the photometric stereo method is used to mark the thermos cup steel bottom G, eliminating the reflection of the thermos cup steel bottom G and reducing the influence of the reflective area in the imaging on the feature extraction.

[0058] like Figure 3As shown, in this embodiment, the marking characters engraved on the steel bottom G of the thermos cup are "YETI". The steel bottom G of the thermos cup has a steel bottom concave circle B. When marking, the center O of the steel bottom concave circle B is used as the marking center point, and the marking characters engraved on the steel bottom G of the thermos cup are "YETI".

[0059] Step S2, obtaining an image of the thermos cup steel bottom with marking characters, and detecting whether the thermos cup steel bottom G has marking characters.

[0060] If the thermos flask steel bottom G has marking characters, then the process proceeds to step S3, otherwise the visual inspection of the marking characters on the thermos flask steel bottom is terminated.

[0061] According to an embodiment of the present invention, after marking characters on the steel bottom G of the thermos cup, the high-definition industrial lens 2 configured by the visual inspection camera 1 is used to photograph the steel bottom G of the thermos cup.

[0062] The visual inspection camera 1 obtains the image of the thermos cup steel bottom with the marking characters, and detects whether the thermos cup steel bottom G has the marking characters "YETI".

[0063] According to an embodiment of the present invention, whether the steel bottom G of the thermos cup has the marking character "YETI" is detected by the following method:

[0064] Step S21, performing edge similarity detection between the marked characters "YETI" in the image of the steel bottom of the thermos cup with the marked characters and the marked characters "YETI" of the standard product.

[0065] Standard products refer to thermos cup steel bottom G with the marking characters "YETI" engraved on it that meets the production requirements.

[0066] Step S22, according to the window search strategy, find the edge contour feature with the maximum similarity score between the marked characters "YETI" in the image of the steel bottom of the thermos cup with the marked characters and the marked characters "YETI" of the standard product.

[0067] That is, through the window search strategy, the edge contour of the marking character "YETI" in the captured image of the steel bottom of the thermos cup is found to be closest to the edge contour of the marking character "YETI" of the standard product.

[0068] Step S23, by scaling and affine transformation, determine the angle relationship between the marking characters "YETI" in the image of the steel bottom of the thermos cup with the marking characters and the marking characters "YETI" of the standard product.

[0069] According to the angle relationship between the marking characters "YETI" in the image of the steel bottom of the thermos cup with marking characters and the marking characters "YETI" of the standard product, the angle value of the marking characters "YETI" in the image of the steel bottom of the thermos cup with marking characters is determined.

[0070] Step S24, when the angle value of the marking character "YETI" in the image of the thermos cup steel bottom with the marking character does not exceed the preset threshold, it is determined that the thermos cup steel bottom G has the marking character "YETI", otherwise it is determined that the thermos cup steel bottom G does not have the marking character "YETI".

[0071] If the thermos cup steel bottom G has the marking character "YETI", then proceed to step S3, otherwise the visual inspection of the marking character on the thermos cup steel bottom G ends. Figure 4 As shown, in one embodiment, it is detected that the steel bottom G of the thermos cup has the marked characters "YETI".

[0072] Step S3, check whether the marking characters "YETI" on the steel bottom G of the thermos flask are complete. If the marking characters "YETI" on the steel bottom G of the thermos flask are complete, proceed to step S4, otherwise the visual inspection of the marking characters on the steel bottom of the thermos flask is terminated.

[0073] According to an embodiment of the present invention, the following method is used to detect whether the marking characters on the steel bottom G of the thermos cup are complete:

[0074] Step S31, extracting each character outline of the marked characters "YETI" from the image of the steel bottom of the thermos cup with the marked characters.

[0075] For example, extract the contours of each character "Y", "E", "T", and "I" of the marked characters "YETI" in the steel bottom image of the thermos cup.

[0076] Step S32: Remove noise points outside each character outline through a black hat algorithm and a ring mask algorithm, and set the pixel value of the area within each character outline to 0.

[0077] Step S33: feature matching each character contour after noise points are removed with each standard character contour in the standard product.

[0078] That is, feature matching is performed between each character contour (the contours of "Y", "E", "T", and "I") of the marked characters "YETI" in the steel bottom image of the thermos cup after removing noise points and each character contour (standard character contour, i.e., the contours of "Y", "E", "T", and "I" of the marked characters "YETI" in the standard product).

[0079] Step S34: determine whether the marking characters on the steel bottom of the thermos cup are complete according to the maximum matching algorithm.

[0080] For example, according to the maximum matching algorithm, when the noise points are removed, each character contour of the marked characters "YETI" in the steel bottom image of the thermos bottle completely matches each character contour of the marked characters "YETI" in the standard product, then the marked characters "YETI" on the steel bottom G of the thermos bottle are complete.

[0081] If the marking characters on the steel bottom G of the thermos flask are complete, the process proceeds to step S4, otherwise the visual inspection of the marking characters on the steel bottom of the thermos flask is terminated.

[0082] Step S4, detecting whether the marking characters on the steel bottom G of the thermos flask are offset, and ending the visual inspection of the marking characters on the steel bottom of the thermos flask.

[0083] According to an embodiment of the present invention, the following method is used to detect whether the marking characters on the steel bottom G of the thermos cup are offset:

[0084] Step S41, extracting the concave edge of the steel bottom concave circle B of the steel bottom G of the thermos cup from the image of the steel bottom of the thermos cup with the marking characters.

[0085] Step S42: performing regional pixel enhancement on the concave edge of the steel bottom concave circle B of the steel bottom G of the thermos cup.

[0086] Since the concave printing circle B of the thermos cup steel bottom G is too shallow, the feature is not obvious. In order to solve the problem that the feature of the concave printing circle B is not obvious, the edge of the concave printing circle B is regionally pixel enhanced.

[0087] Specifically, local cosine calculation is performed on the edge of the gravure circle B, the pixel value of the edge of the gravure circle B is enhanced, and the pixels near the edge of the gravure circle B are weakened.

[0088] Step S43: using the RANSAC algorithm (Random Sample Consensus) to fit the gravure edge of the gravure circle B after pixel enhancement, and generate a gravure edge fitting circle.

[0089] Step S44: extract the outline of the marking character "YETI" from the image of the steel bottom of the thermos cup with the marking character, and use the angle discrimination method to filter out the outermost point set of the outline of the marking character "YETI".

[0090] Step S45: Use the circular caliper algorithm tool to fit the outermost point set of the outline of the marking character "YETI" to generate a marking character outline circle.

[0091] Step S46: Compare the distance between the center of the gravure edge fitting circle and the center of the marking character outline circle.

[0092] That is, the distance between the center of the gravure edge fitting circle generated by the gravure edge of the gravure circle B after pixel enhancement and the center of the marking character contour circle generated by the outermost point set of the contour of the marking character "YETI" is compared.

[0093] If the distance between the center of the gravure edge fitting circle and the center of the marking character outline circle is greater than the preset threshold, it is determined that the marking character "YETI" on the steel bottom G of the thermos cup is offset, otherwise the marking character "YETI" on the steel bottom G of the thermos cup is not offset.

[0094] like Figure 5 As shown, in one embodiment, it is detected that the marking character "YETI" on the steel bottom G of the thermos cup is offset. Figure 6 As shown, in one embodiment, the marking characters "YETI" on the steel bottom G of the thermos cup detected are not offset, that is, the marking characters "YETI" on the steel bottom G of the thermos cup conform to the standard.

[0095] After completing the detection of whether the marking characters "YETI" on the steel bottom G of the thermos cup are offset, the visual inspection of the marking characters on the steel bottom of the thermos cup is ended.

[0096] The present invention performs trap detection on the marked character "YETI" on the steel bottom G of the thermos cup based on machine vision, which can reduce the investment in labor costs, greatly improve work efficiency, reduce costs and increase efficiency, and also reduce the hazards of laser work positions.

[0097] According to an embodiment of the present invention, a visual detection system for marking characters is provided, which is used to execute a visual detection method for marking characters on a steel bottom of a thermos cup provided by the present invention.

[0098] According to an embodiment of the present invention, a computer storage medium is provided for storing computer execution instructions. The computer execution instructions are used to execute a visual detection method for marking characters on the steel bottom of a thermos cup provided by the present invention.

[0099] There are a few points to note:

[0100] (1) The drawings of the embodiments of the present invention only involve structures related to the embodiments of the present invention. Other structures may refer to conventional designs.

[0101] (2) For the sake of clarity, in the drawings used to describe the embodiments of the present invention, the thickness of layers or regions is exaggerated or reduced, that is, these drawings are not drawn according to the actual scale. It is understood that when an element such as a layer, film, region or substrate is referred to as being "on" or "under" another element, the element may be "directly" "on" or "under" the other element or there may be intermediate elements.

[0102] (3) In the absence of conflict, the embodiments of the present invention and the features therein may be combined with each other to obtain new embodiments.

[0103] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. The protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for visually detecting marked characters, characterized in that: The visual detection method comprises the following steps: S1. Mark the steel bottom of the thermos cup; S2, obtaining an image of the steel bottom of the thermos cup with marking characters, and detecting whether the steel bottom of the thermos cup has marking characters; If the steel bottom of the thermos flask has a marking character, then the process proceeds to step S3, otherwise the visual inspection of the marking character on the steel bottom of the thermos flask ends; S3. Check whether the marking characters on the steel bottom of the thermos cup are complete; If the marking characters on the steel bottom of the thermos flask are complete, then the process proceeds to step S4, otherwise the visual inspection of the marking characters on the steel bottom of the thermos flask ends; S4, detecting whether the marking characters on the steel bottom of the thermos flask are offset, and ending the visual inspection of the marking characters on the steel bottom of the thermos flask.

2. The visual inspection method according to claim 1, characterized in that: In step S2, the following method is used to detect whether the steel bottom of the thermos cup has marked characters: S21, performing edge similarity detection between the marking characters in the image of the steel bottom of the thermos cup with the marking characters and the marking characters of the standard product; S22, according to the window search strategy, find the edge contour feature with the largest similarity score between the marking characters in the steel bottom image of the thermos cup with the marking characters and the marking characters of the standard product; S23, determining the angle relationship between the marking characters in the image of the steel bottom of the thermos cup with the marking characters and the marking characters of the standard product by scaling and affine transformation; Determine the angle value of the marking character in the image of the steel bottom of the thermos cup with marking characters according to the angle relationship between the marking character in the image of the steel bottom of the thermos cup with marking characters and the marking character of the standard product; S24. When the angle value of the marking characters in the image of the thermos cup steel bottom with the marking characters engraved thereon does not exceed the preset threshold, it is determined that the thermos cup steel bottom has the marking characters, otherwise it is determined that the thermos cup steel bottom does not have the marking characters.

3. The visual inspection method according to claim 1, characterized in that: In step S3, the marking characters on the steel bottom of the thermos cup are checked to see if they are complete by the following method: S31, extracting each character outline of the marked characters from the image of the steel bottom of the thermos cup with the marked characters; S32, removing noise points outside each character outline by using a black hat algorithm and a ring mask algorithm; S33, performing feature matching on each character contour after removing noise points and each standard character contour in the standard product; S34. Determine whether the marking characters on the steel bottom of the thermos cup are complete according to the maximum matching algorithm.

4. The visual inspection method according to claim 1, characterized in that: In step S4, the marking characters on the steel bottom of the thermos cup are detected to see if they are offset by the following method: S41, extracting the gravure edge of the steel bottom of the thermos cup from the image of the steel bottom of the thermos cup with the marking characters engraved thereon; S42, performing regional pixel enhancement on the concave printing edge of the steel bottom of the thermos cup; S43, using the RANSAC algorithm to fit the gravure edge after pixel enhancement, and generate a gravure edge fitting circle; S44, extracting the outline of the marking character from the image of the steel bottom of the thermos cup with the marking character, and using the angle discrimination method to filter out the outermost point set of the marking character outline; S45, using a circular caliper algorithm tool to fit the outermost point set of the marking character contour to generate a marking character contour circle; S46, comparing the distance between the center of the gravure edge fitting circle and the center of the marking character outline circle; If the distance between the center of the gravure edge fitting circle and the center of the marking character outline circle is greater than a preset threshold, it is determined that the marking characters on the steel bottom of the thermos flask are offset; otherwise, the marking characters on the steel bottom of the thermos flask are not offset.

5. A visual inspection system for marking characters, characterized in that: The visual inspection system is used to execute the visual inspection method described in any one of claims 1 to 4.

6. A computer storage medium, characterized in that: The computer storage medium is used to store computer-executable instructions; the computer-executable instructions are used to execute the visual inspection method described in any one of claims 1 to 4.