WO2022151460A1 - 激光标签识别设备和识别方法 - Google Patents

激光标签识别设备和识别方法 Download PDF

Info

Publication number
WO2022151460A1
WO2022151460A1 PCT/CN2021/072381 CN2021072381W WO2022151460A1 WO 2022151460 A1 WO2022151460 A1 WO 2022151460A1 CN 2021072381 W CN2021072381 W CN 2021072381W WO 2022151460 A1 WO2022151460 A1 WO 2022151460A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
laser
standard
object image
label
Prior art date
Application number
PCT/CN2021/072381
Other languages
English (en)
French (fr)
Inventor
杨莞琳
高煜
谢晖
Original Assignee
罗伯特·博世有限公司
杨莞琳
高煜
谢晖
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 罗伯特·博世有限公司, 杨莞琳, 高煜, 谢晖 filed Critical 罗伯特·博世有限公司
Priority to CN202180090862.6A priority Critical patent/CN117321656A/zh
Priority to EP21918662.4A priority patent/EP4280198A4/en
Priority to PCT/CN2021/072381 priority patent/WO2022151460A1/zh
Publication of WO2022151460A1 publication Critical patent/WO2022151460A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
    • G09F—DISPLAYING; ADVERTISING; SIGNS; LABELS OR NAME-PLATES; SEALS
    • G09F3/00—Labels, tag tickets, or similar identification or indication means; Seals; Postage or like stamps
    • G09F3/02—Forms or constructions
    • G09F3/0297—Forms or constructions including a machine-readable marking, e.g. a bar code
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/40—Extraction of image or video features
    • G06V10/54—Extraction of image or video features relating to texture
    • G—PHYSICS
    • G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
    • G09F—DISPLAYING; ADVERTISING; SIGNS; LABELS OR NAME-PLATES; SEALS
    • G09F3/00—Labels, tag tickets, or similar identification or indication means; Seals; Postage or like stamps
    • G09F3/02—Forms or constructions
    • G09F3/0291—Labels or tickets undergoing a change under particular conditions, e.g. heat, radiation, passage of time
    • G09F3/0294—Labels or tickets undergoing a change under particular conditions, e.g. heat, radiation, passage of time where the change is not permanent, e.g. labels only readable under a special light, temperature indicating labels and the like

Definitions

  • the invention relates to a laser label identification device and identification method, and a corresponding machine storage medium.
  • laser labels formed by laser coding have been widely used in automobiles, consumer products, consumer electronics, and so on.
  • Lasers are beams produced by the stimulated radiation of particles, and are known as "the fastest knife”, “the most accurate ruler” and “the brightest light”. Therefore, compared with traditional marking processes such as chemical etching, ink jet coding, and mechanical engraving, laser marking has the advantages of fast marking speed, no wear and no environmental pollution.
  • Laser tags on products are read and identified by barcode readers.
  • Existing code readers mainly focus on improving the ability to identify distorted label images, but there is no ideal solution for confirming the authenticity of labels and products using the labels. In this way, many problems will arise. For example, consumers cannot accurately know whether the product they buy is genuine or counterfeit through laser label verification; it is difficult to determine whether a part is genuine or counterfeit through laser label verification on the production line, resulting in many components must be reworked or changed. into waste.
  • a laser tag identification device which includes: a receiving module configured to receive an object image, the object image including the laser tag image and the laser light on the surface of the object A label area image of the label area; a processing module configured to process the object image to obtain a texture feature of the laser label area on the object surface, the texture feature and the material of the laser label area, using The laser marking device is associated with the laser marking process; and an identification module configured to match the object image with a stored standard image to identify the authenticity of the object image, the standard image containing as a match A standard standard laser label image and a standard label area image containing standard texture features of the label area, wherein the matching includes matching the label image and texture features of the object with the standard label image and standard texture features, respectively.
  • the factors affecting the texture feature include at least one of the following:
  • the laser tag identification device further includes an output module configured to output an identification result in order to determine the authenticity of the product corresponding to the object image.
  • the authenticity of the product includes:
  • the product corresponding to the object image is genuine or fake
  • the requirements for customization include one or more of the pre-specified place of origin, production organization and distribution channel;
  • the product corresponding to the object image belongs to one of the multiple batches of products or does not belong to any one of the multiple batches of products.
  • identifying the authenticity of the object image includes: matching the object image with a standard image representing a batch of products; if the matching is successful, determining that the product corresponding to the object image belongs to the batch; and if the matching fails, it is determined that the product corresponding to the object image does not belong to the batch.
  • identifying the authenticity of the object image includes: matching the object image with a standard image representing each batch of products in a plurality of batches; If the standard image is successfully matched, it is determined that the product corresponding to the object image belongs to the batch; and if the object image fails to match with the standard image representing any batch of products, it is determined that the object is the object The product corresponding to the image does not belong to any one of the multiple batches of products.
  • identifying the authenticity of the object image includes: matching the object image with a standard image of a genuine product; in the case of successful matching, determining that the product corresponding to the object image is genuine ; and if the matching fails, it is determined that the product corresponding to the object image is fake.
  • the matching includes: generating a multi-dimensional object feature vector based on each parameter of the texture feature of the object image and its weight; generating a multi-dimensional standard based on each parameter of the standard texture feature of the standard image and its weight feature vector; calculating a distance between an object feature vector and a standard feature vector, the distance optionally being an Euclidean distance; and determining the authenticity of the object image based on whether the distance complies with a distance threshold.
  • the recognition module realizes the recognition by means of an image recognition model, takes the object image as a model input, and obtains a model output representing the authenticity of the object image after model processing.
  • the image processing model is a machine learning model.
  • the machine learning model performs relearning based on information obtained by the identification device during use.
  • the identification device is provided in a reader for capturing the object image; or the identification device is provided in a verification device for verifying the authenticity of the product corresponding to the object image.
  • Another aspect of the present invention provides a laser tag identification method, optionally performed by the identification device as described above, the method comprising: receiving an object image, the object image including a laser tag image and a laser tag on the surface of the object The label area image of the area; the object image is processed to obtain the texture feature of the laser label area on the surface of the object, and the texture feature is related to the material of the laser label area, the laser marking equipment used and the laser marking area. and matching the object image with a stored standard image to identify the authenticity of the object image, the standard image containing a standard laser label image as a matching standard and a standard texture feature containing the label area. Standard label area image, wherein the matching includes matching the label image and texture feature of the object with the standard label image and standard texture feature, respectively.
  • Yet another aspect of the present invention provides a machine-readable storage medium storing executable instructions that, when executed, cause one or more processors to perform the method as described above.
  • Figure 1 illustrates an exemplary operating environment in accordance with some embodiments of the present invention.
  • Figure 2 shows a reader provided with an identification device according to an embodiment of the invention.
  • Figure 3 shows a schematic block diagram of an identification device according to an embodiment of the present invention.
  • Figure 4 shows an identification process according to an embodiment of the present invention.
  • FIG. 5 shows a flowchart of an identification method according to an embodiment of the present invention.
  • An important aspect of the present invention is the identification and verification of authenticity by means of the surface of the product itself. Specifically, each product leaves a texture feature on the surface during the manufacturing process, which is unique and can be understood as the "natural fingerprint" of the product.
  • Embodiments of the present invention provide an identification scheme by means of the "natural fingerprint" of the laser label, that is, in addition to reading the laser label itself, also read the textural features of the area on the product surface that is laser marked to form the laser label, And match the read information with the stored standard information to realize authenticity identification and verification.
  • the label in the present invention refers to a laser label, that is, a mark formed on the surface of a product by laser marking (eg, laser marking, laser engraving).
  • Laser tags can be implemented in different colors and/or in different styles.
  • Laser labels can be implemented as laser engraved one-dimensional codes, two-dimensional codes, or custom markings (eg, letters, numbers, symbols of custom colors), and the like.
  • the laser label and its label area can be implemented as: a black block area of the two-dimensional code burned on a light-colored material; a blank block area of the two-dimensional code burned on a dark material; two-dimensional code burned on a light-colored material code blank block area; or burn the QR code black block area on dark materials.
  • authentication may include some cases as described below.
  • the products corresponding to the laser labels are genuine or counterfeit products, so that counterfeit products can be identified.
  • "true” corresponds to the product being genuine
  • “fake” corresponds to the product being counterfeit.
  • the customized requirements may include pre-specified origin (eg, in some cases, by identifying the product's origin to identify the grade and specification of the product), production organization (eg, in some cases, by identifying the product's production organization to identify the product's grade and specification), distribution channel (in some cases, by identifying the distribution channel of the product to identify whether the source of the product is compliant).
  • pre-specified origin eg, in some cases, by identifying the product's origin to identify the grade and specification of the product
  • production organization eg, in some cases, by identifying the product's production organization to identify the product's grade and specification
  • distribution channel in some cases, by identifying the distribution channel of the product to identify whether the source of the product is compliant.
  • a batch of products or "a batch of products” are, for example, products corresponding to a batch of labels printed on the same material by the same laser marking equipment.
  • a batch of products or “a batch of products” may also include limitations on other factors, for example, limitations on printing time periods.
  • an identification scheme according to an embodiment of the present invention can be applied in many fields, such as industry, consumer goods, retail, building, agriculture, and transportation.
  • an identification scheme according to an embodiment of the present invention can be used to: (1) identify the authenticity of a product, for example, by providing an identification scheme to reduce the number of counterfeit products of counterfeit brands and minimize reputation loss; (2) identify whether a product is Products belonging to the same batch; (3) in the process of authenticity identification and verification, by transmitting product information to the network, it has the function of tracking and tracing, which helps to improve the transparency of distribution channels; (4) in the process of customer through customer In the process of authenticating verification through the application program on the terminal, it can establish direct contact with customers, thereby improving customer relationship management.
  • FIG. 1 illustrates an exemplary operating environment 100 in accordance with some embodiments of the present invention.
  • object 1 object 1 is the product to be identified and authenticated, eg a screw
  • object 1 has a laser tag, eg the letters "SN".
  • the reader 2 is arranged to be aimed at the laser label area R of the object 1, captures the texture features of the laser label area while reading the laser label, and generates the object image 3 containing the label area image of the label image.
  • a standard image serving as a comparison standard is stored, which includes a standard label image and a standard label area image including standard texture features of the standard label area.
  • the laser label recognition device (hereinafter referred to as "recognition device”) 5 includes a recognition strategy, that is, the object image and the standard image are matched, for example, the label image of the object is compared with the standard label image; after the label comparison is passed, Match the texture features of the label area with the standard texture features; if the matching is passed, it is determined that the object image belongs to the true category; if the matching fails, it is determined that the object image belongs to the pseudo category.
  • the identification device 5 can be implemented in hardware or software or a combination of software and hardware.
  • it can be implemented in one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), data signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs) ), a processor, a controller, a microcontroller, a microprocessor, an electronic unit designed to perform its functions, or a combination thereof.
  • ASICs application specific integrated circuits
  • DSPs digital signal processors
  • DSPDs data signal processing devices
  • PLDs programmable logic devices
  • FPGAs field programmable gate arrays
  • a processor a controller, a microcontroller, a microprocessor, an electronic unit designed to perform its functions, or a combination thereof.
  • microcontroller a microcontroller
  • microprocessor an electronic unit designed to perform its functions, or a combination thereof.
  • the identification device 5 includes a memory and a processor.
  • the memory contains instructions that, when executed by the processor, cause the processor to perform an identification strategy/identification method according to an embodiment of the present invention.
  • Figure 2 shows a reader 200 provided with an identification device 5 according to an embodiment of the invention.
  • the identification device 5 may be implemented as identification software provided in the reader 200 .
  • Reader 200 may be a stationary reader, a handheld reader, or the like.
  • the identification device 5 when the identification device 5 is implemented in the form of software, the software can be applied to a code reading program or a product verification program, thereby optimizing the authenticity identification function of the code reading program or the product verification program.
  • FIG. 3 shows a schematic block diagram of an identification device 5 according to an embodiment of the present invention.
  • the identification device 5 mainly includes a receiving module 52 , a processing module 54 , an identification module 56 and an output module 58 .
  • each module of the identification device 5 should be understood as a logical description, rather than a limitation on the physical form or arrangement.
  • one or more of the receiving module 52 , the processing module 54 , the identifying module 56 and the outputting module 58 may be implemented in the same chip or circuit, or they may be provided in different chips or circuits respectively, which the present invention does not be limited.
  • Each module of the identification device 5 can be further divided into a plurality of sub-modules, and each sub-module is implemented as a sub-program.
  • FIG 4 shows an identification process 400 according to an embodiment of the present invention.
  • the identification process 400 may be performed in the identification device 5 described above.
  • the receiving module 52 receives an image of the object captured by the reader.
  • the object image includes a laser label image and a label area image, which is an image of a label area on the surface of the object. For example, an image of the surface area R where the laser tag "SN" in Figure 1 is located.
  • the processing module 54 processes the captured image of the object to obtain texture features of a designated area on the surface of the object (i.e., the area on the product label that was laser inscribed to form the laser label).
  • the texture features include natural fingerprint features of the labeled regions on the surface of the object, the natural fingerprint features being associated with at least one of the following.
  • Process parameters in the laser marking process for forming the laser label for example, one or more of the focus spot size, laser power and beam quality.
  • laser marking equipment for example, fiber laser marking machine, end-pump infrared/green light/ultraviolet laser marking machine, laser engraving machine, CO2 laser marking machine
  • performance parameters for example, output power, beam quality, pulse width, pulse repetition rate
  • the identification module 56 matches the object image to a stored standard image (eg, the identification module 56 retrieves the corresponding standard image from the database 4 for identifying the object/product) to identify the authenticity of the object image.
  • the standard image includes a laser marking-based standard label image as a matching standard and a standard label area image that includes standard texture features of the standard label area.
  • the label image and texture feature of the object can be matched with the standard label image and standard texture feature, respectively.
  • the label image of the object is first matched with the standard label image, and after the matching is passed, the texture feature of the label area of the object is then matched with the standard texture feature. It will be appreciated that both the matching on the label image and the matching on the textual features can also be performed simultaneously.
  • the identification module 56 can adopt a suitable image processing and matching scheme, which is not limited in the present invention.
  • the identification module 56 may do so by extracting multi-dimensional feature vectors and computing vector differences (see block 4603).
  • the recognition module 56 calculates the difference between the multi-dimensional feature vector Y A abstracted from sample A and the multi-dimensional feature vector Y B abstracted from sample B. Whether the similarity between them (for example, represented by a vector difference between the two) conforms to a predetermined threshold to determine whether the object image belongs to the true class or the false class.
  • the identification module 56 generates a multi-dimensional feature vector Y A (for example, a standard feature vector) of the sample A based on the multiple parameters characterizing the micro-texture features of the sample A and the weights of the parameters; The multiple parameters and the weight of each parameter generate a multidimensional feature vector Y B (eg, object feature vector) of sample B.
  • the distance between the standard feature vector Y A and the object feature vector Y B eg, the Euclidean distance
  • the N-dimensional feature matrix X A , X A ⁇ R N of sample A is abstracted based on the surface material of the label area of sample A, the process parameters in the laser coding process, and the model of laser coding equipment and its setting parameters ; generate a weight matrix W A , W A ⁇ R M,N+1 that characterizes the weights of the considerations of sample A; and based on the N-dimensional feature matrix X A and weight matrix W A of sample A A A multiplication of these two matrices) to generate a multidimensional eigenvector Y A (eg, a standard eigenvector) of sample A.
  • a multidimensional eigenvector Y A eg, a standard eigenvector
  • the N-dimensional feature matrix X B , X B ⁇ R of sample B is abstracted based on the surface material of the label area of sample B, the process parameters in the laser coding process, and the model of laser coding equipment and its setting parameters.
  • N generate a weight matrix W B , W B ⁇ R M ,N+1 that characterizes the weights of the considerations of sample B ;
  • W B multiply these two matrices) to generate a multidimensional feature vector Y B (eg, object feature vector) of sample B.
  • the similarity between the standard feature vector Y A and the object feature vector Y B is calculated, for example, the Euclidean distance between them is calculated.
  • the similarity between the standard feature vector Y A and the object feature vector Y B can be calculated by the following formula:
  • the similarity (A, B) between the standard feature vector Y A and the object feature vector Y B is less than a predetermined threshold, it is determined that the sample B meets the standard (that is, the determination result is "true"), for example, the sample B is genuine Or sample B and sample A belong to the same batch of products.
  • the recognition module 56 may perform the above-mentioned matching (ie, matching between the object image and the standard image) by means of an image processing model.
  • the image processing model can be implemented by means of artificial intelligence techniques, for example, the model is implemented as a trained machine learning model.
  • the model can improve the robustness of its discriminative ability with the help of a suitable neural network model.
  • the image processing model is trained using a large number of counterfeit images and standard images as samples, so that when the model receives a new image input, it can determine that the new image belongs to the true category (that is, the matching result). to be able to match with the standard image) or pseudo-category (that is, the result of the match is that it cannot be matched with the standard image).
  • the image processing model can also perform re-learning based on the information in the process of use, so as to obtain update parameters periodically, thereby improving the intelligence capability and processing speed of the model.
  • the identification module 56 determines whether the product corresponding to the object image is one of the products in the same batch through matching. In this scenario, it is only necessary to determine whether the object belongs to one of the same batch of products without identifying a specific laser tag or product. In other words, in this scenario, as long as it is determined that the product corresponding to the object image belongs to one of the same batch of products, the recognition result is true; otherwise, the recognition result is false.
  • the identification module 56 can compare the object image with a standard image representing a batch of products to determine the authenticity of the object image, thereby determining whether the product piece corresponding to the object image belongs to the batch of products one.
  • each standard image contains a standard label image and a standard label area image
  • multiple copy images each copy image contains a laser label image and label to prevent region image
  • the model can identify whether the products corresponding to the new input images belong to the same batch of products.
  • the identification device may identify which batch of multiple batches the product corresponding to the object image belongs to, or does not belong to any one of the batches.
  • the recognition module 56 matches the object image with the standard images representing each batch of products in multiple batches; if the object image and the standard image representing a batch of products are successfully matched, it is determined that the object image corresponds to The product belongs to the batch; if the object image fails to match the standard image representing any batch of products, it is determined that the product corresponding to the object image does not belong to any batch of multiple batches of products.
  • the identification module 56 determines the authenticity of the product corresponding to the object image by matching.
  • the identification module 56 matches the object image with a standard image representing a single genuine product to determine whether the object image belongs to a true category or a fake category, thereby determining the authenticity of the product corresponding to the object image.
  • a standard image representing a single genuine product to determine whether the object image belongs to a true category or a fake category, thereby determining the authenticity of the product corresponding to the object image.
  • an image processing model is trained using standard images and a large number of object images of counterfeit or non-compliant products, so that when a new image is input, the model can determine the authenticity of the product corresponding to the new image.
  • FIG. 5 shows a flowchart of an identification method 500 according to an embodiment of the present invention.
  • the identification method 500 may be performed by the identification device 5 described above. Therefore, the above related descriptions also apply here.
  • step S502 an object image is received, the object image includes a laser label image and a label area image of the label area on the surface of the object.
  • step S504 the object image is processed to obtain the texture features of the label area on the object surface.
  • the texture feature is related to the material of the laser label area, the laser marking equipment used and the marking process.
  • step S506 the object image is matched with the stored standard image to distinguish the authenticity of the object image, and the standard image includes the standard laser label image as the matching standard and the standard image including the standard texture features of the label area.
  • Standard label area image wherein the matching includes matching the label image and texture feature of the object with the standard label image and standard texture feature, respectively.
  • the present invention also provides a machine-readable storage medium storing executable instructions which, when executed, cause a machine to perform the identification method or identification process as described above.
  • the identification device described above may be implemented in a number of ways. For example, it may be implemented as hardware, software, or a combination thereof.
  • the identification device may include one or more processors. These processors may be implemented using electronic hardware, computer software, or any combination thereof. Whether these processors are implemented as hardware or software will depend on the specific application and the overall design constraints imposed on the system. As an example, a processor, any portion of a processor, or any combination of processors presented in this disclosure may be implemented as a microprocessor, microcontroller, digital signal processor (DSP), field programmable gate array (FPGA) ), programmable logic devices (PLDs), state machines, gate logic, discrete hardware circuits, and other suitable processing components configured to perform the various functions described in this disclosure. The functions of a processor, any portion of a processor, or any combination of processors presented herein can be implemented as software executed by a microprocessor, microcontroller, DSP, or other suitable platform.
  • DSP digital signal processor
  • FPGA field programmable gate array
  • PLDs programmable logic devices
  • state machines gate logic, discrete hardware circuits, and other suitable processing components configured to perform the various functions described in
  • Computer readable media may include, for example, memory, which may be, for example, magnetic storage devices (eg, hard disks, floppy disks, magnetic stripes), optical disks, smart cards, flash memory devices, random access memory (RAM), read only memory (ROM), memory Programmable ROM (PROM), Erasable PROM (EPROM), Electrically Erasable PROM (EEPROM), Register or Removable Disk.
  • memory may be, for example, magnetic storage devices (eg, hard disks, floppy disks, magnetic stripes), optical disks, smart cards, flash memory devices, random access memory (RAM), read only memory (ROM), memory Programmable ROM (PROM), Erasable PROM (EPROM), Electrically Erasable PROM (EEPROM), Register or Removable Disk.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Multimedia (AREA)
  • Image Analysis (AREA)

Abstract

一种激光标识识别设备（5）和方法（500）。识别设备（5）包括：接收模块（52），被配置成接收对象图像（3），对象图像（3）包含激光标签图像和对象表面上激光标签区域（R）的标签区域图像；处理模块（54），被配置成对对象图像（3）进行处理，以获得对象表面上激光标签区域（R）的纹理特征，纹理特征与激光标签区域（R）的材质、采用的激光打标设备和打标过程相关；以及识别模块（56），被配置成将对象图像（3）与存储的标准图像进行匹配，以识别对象图像（3）属于真类别还是伪类别，标准图像包含作为匹配标准的标准激光标签图像和包含标签区域（R）的标准纹理特征的标准标签区域图像。

Description

激光标签识别设备和识别方法 技术领域
本发明涉及一种激光标签识别设备和识别方法，以及相应的机器存储介质。
背景技术
近年来，通过激光打码形成激光标签被广泛应用于汽车、消费品、消费电子，等等。激光是由粒子受激辐射产生的光束，被誉为“最快的刀”、“最准的尺”和“最亮的光”。因此，相比于诸如化学腐蚀，油墨喷码、机械雕刻之类的传统打标工艺，激光打标具备标记速度快、永不磨损、无环境污染的优点。
产品上的激光标签通过读码器读取并被识别。现有的读码器主要专注于提升辨识畸变标签图像的能力，但是，在确认标签以及采用该标签的产品的真伪的安全性方面，尚不存在较为理想的解决方案。这样，会引起诸多问题，例如，消费者通过激光标签验证无法准确获知其购买的产品是正品还是赝品；在生产线上难以通过激光标签验证确定出零件是正品还是赝品，导致许多组件必须重工或变成废料。
发明内容
鉴于现有技术中存在的问题，本发明的一个方面提出了一种激光标签识别设备，其包括：接收模块，其被配置成接收对象图像，所述对象图像包含激光标签图像和对象表面上激光标签区域的标签区域图像；处理模块，其被配置成对所述对象图像进行处理，以获得所述对象表面上激光标签区域的纹理特征，所述纹理特征与所述激光标签区域的材质、采用的激光打标设备和激光打标过程相关；以及识别模块，其被配置成将所述对象图像与存储的标准图像进行匹配以识别出所述对象图像的真伪，所述标准图像包含作为匹配标准的标准激光标签图像和包含标签区域的标准 纹理特征的标准标签区域图像，其中，所述匹配包括将对象的标签图像和纹理特征分别与标准标签图像和标准纹理特征进行匹配。
根据一实施方式，其中，影响所述纹理特征的因素包括以下至少一项：
-所述激光打标过程中的工艺参数；
-采用的激光打标设备的类型及其性能参数；和
-采用的激光打标设备和所述激光标签区域的材质的组合。
根据一实施方式，所述激光标签识别设备还包括输出模块，其被配置成输出识别结果，以便确定所述对象图像对应的产品的真伪。
根据一实施方式，所述产品的真伪包括：
(1)对象图像对应的产品为正品或赝品；
(2)对象图像对应的产品是否符合定制的要求，可选地，所述定制的要求包括预先规定的产地、生产机构和分销渠道中的一项或多项；
(3)对象图像对应的产品是否属于一批次产品；
(4)对象图像对应的产品属于多个批次产品中的一个批次或不属于多个批次中的任一批次品。
根据一实施方式，识别所述对象图像的真伪包括：将所述对象图像与表示一批次产品的标准图像进行匹配；在匹配成功的情况下，确定为所述对象图像对应的产品属于该批次；以及在匹配失败的情况下，确定为所述对象图像对应的产品不属于该批次。
根据一实施方式，识别所述对象图像的真伪包括：将所述对象图像与表示多个批次产品的各批次产品的标准图像进行匹配；在所述对象图像与表示一批次产品的标准图像匹配成功的情况下，确定为所述对象图像对应的产品属于该批次；以及在所述对象图像与表示任一批次产品的标准图像都匹配失败的情况下，确定为所述对象图像对应的产品不属于所述多个批次产品的任一批次。
根据一实施方式，识别所述对象图像的真伪包括：将所述对象图像与类型为真的产品的标准图像进行匹配；在匹配成功的情况下，确定为所述对象图像对应的产品为真；以及在匹配失败的情况下，确定为所述对象图像对应的产品为伪。
根据一实施方式，所述匹配包括：基于所述对象图像的纹理特征的各参数及其权重生成多维的对象特征向量；基于所述标准图像的标准纹理特征的各参数及其权重生成多维的标准特征向量；计算对象特征向量与标准特征向量之间的距离，所述距离可选地为欧式距离；以及基于所述距离是否符合距离阈值确定所述对象图像的真伪。
根据一实施方式，所述识别模块借助于图像识别模型实现所述识别，将所述对象图像作为模型输入，并经过模型处理后得到表示所述对象图像真伪的模型输出。
根据一实施方式，所述图像处理模型为机器学习模型。
根据一实施方式，所述机器学习模型基于所述识别设备在使用过程中获得的信息执行再学习。
根据一实施方式，所述识别设备设置在用于捕捉所述对象图像的读取器中；或者所述识别设备设置在用于验证所述对象图像对应的产品的真伪的验证设备中。
本发明的另一个方面提出了一种激光标签识别方法，可选地由如上所述的识别设备执行，所述方法包括：接收对象图像，所述对象图像包含激光标签图像和对象表面上激光标签区域的标签区域图像；对所述对象图像进行处理，以获得所述对象表面上激光标签区域的纹理特征，所述纹理特征与所述激光标签区域的材质、采用的激光打标设备和激光打标过程相关；以及将所述对象图像与存储的标准图像进行匹配以识别出所述对象图像的真伪，所述标准图像包含作为匹配标准的标准激光标签图像和包含标签区域的标准纹理特征的标准标签区域图像，其中，所述匹配包括将对象的标签图像和纹理特征分别与标准标签图像和标准纹理特征进行匹配。
本发明的又一个方面提出了一种机器可读存储介质，其存储有可执行指令，所述指令当被执行时，使得一个或多个处理器执行如上所述的方法。
为了能更进一步了解本发明的特征以及技术内容，请参阅以下详细说明和附图，然而附图仅提供参考与说明用，并非用来对本发明加以限制。
附图说明
图1示出了根据本发明的一些实施方式的示例性操作环境。
图2示出了设置有根据本发明的实施方式的识别设备的读取器。
图3示出了根据本发明的实施方式的识别设备的示意性框图。
图4示出了根据本发明的实施方式的识别过程。
图5示出了根据本发明的实施方式的识别方法的流程图。
具体实施方式
本发明的一个重要方面在于借助于产品本身的表面进行真伪识别和验证。具体而言，每一个产品在制造过程中都会在表面留下纹理特征，这一特征独一无二，可以理解为产品的“自然指纹”。本发明的实施例提供了借助于激光标签的“自然指纹”的识别方案，即，除了读取激光标签本身以外，还读取产品表面上被激光打标以形成激光标签的区域的纹理特征，并将读取的信息与存储的标准信息进行匹配，以实现真伪识别和验证。
本发明中的标签指的是激光标签，即，通过激光打标(例如，激光打码，激光刻码)而在产品表面形成的标识。激光标签可以实现为不同的颜色和/或不同的样式。激光标签可以实现为激光刻出一维码、二维码、或定制标记(例如，定制颜色的字母、数字、符号)，等等。
在一实施例中，激光标签及其标签区域可以实现为：浅色材 料上烧刻二维码黑色块区域；深色材料上烧刻二维码空白块区域；浅色材料上烧刻二维码空白块区域；或深色材料上烧刻二维码黑色块区域。
在本发明中，“真伪”可以包括如下所述的一些情况。
(1)激光标签对应的产品为正品或赝品，由此可以鉴别出伪造品。在该情况下，“真”对应于产品为正品；“伪”对应于产品为赝品。
(2)激光标签对应的产品是否符合定制的要求。在该情况下，“真”对应于产品符合定制的要求；“伪”对应于产品不符合定制的要求。
该定制的要求可以包括预先规定的产地(例如，有的情况下通过鉴别产品的产地来鉴别产品的等级和规格)、生产机构(例如，有的情况下通过鉴别产品的生产机构来鉴别产品的等级和规格)、分销渠道(有的情况下通过鉴别产品的分销渠道来鉴别产品来源是否合规)中的一项或多项。
(3)激光标签对应的产品是否属于一批次产品。在该情况下，“真”对应于产品属于该批次产品；“伪”对应于产品不属于该批次产品。
(4)在多个批次产品的情况下，可以识别出激光标签对应的产品属于多个批次中的哪一个批次，或者不属于多个批次中任一批次。在该情况下，“真”对应于产品属于多个批次产品中的一个批次；“伪”对应于产品不属于多个批次中任一批次。
“一批次产品”或“一个批次的产品”例如是由同一台激光打标设备在同一材质的材料上打印出的一批标签对应的产品。
例如，将激光打标设备A在材料B上打印的一批激光标签对应的产品设定为批次Ⅰ；将激光打标设备A在材料C上打印的一批激光标签对应的产品设定为批次Ⅱ；将另一激光打标设备D在材料A上打印的一批激光标签对应的产品设定为批次Ⅲ；将该另一激光打标设备D在材料C上打印的一批激光标签对应的产品设定为批次次Ⅳ。
另外，除了对于打印设备和打印材料的限定，“一批次产品”或“一个批次的产品”还可以包括其他因素的限定，例如，还包括打印时段的限定。
根据本发明实施例的识别方案可以在工业、消费品、零售、楼宇、农业和交通等诸多领域得到应用。例如，根据本发明实施例的识别方案可以用于：(1)鉴别产品真伪，例如，通过提供识别方案减少仿造品牌的假冒产品数量，并最大限度地减少名誉损失；(2)鉴别产品是否属于同一批次的产品；(3)在真伪识别和验证过程中，通过将产品信息传输到网络，从而具备追踪和追溯功能，有助于提高分销渠道的透明度；(4)在客户通过客户端的应用程序来实现真伪验证的过程中，能够与客户建立直接的联系，从而提升客户关系管理。
以下，结合附图来说明本发明的具体实施方式。
图1示出了根据本发明的一些实施方式的示例性操作环境100。在环境100中，对象1(对象1为待识别和验证的产品，例如，为一螺钉)具有激光标签，该激光标签例如为字母“SN”。读取器2设置成对准对象1的激光标签区域R，在读取激光标签的同时捕捉激光标签区域的纹理特征，并生成包含标签图像的标签区域图像的对象图像3。在数据库(资料库)4中存储有用作比对标准的标准图像，其包含标准标签图像和包含标准标签区域的标准纹理特征的标准标签区域图像。激光标签识别设备(以下简称“识别设备”)5包含识别策略，即，将对象图像和标准图像进行匹配，例如，将对象的标签图像和标准标签图像进行比对；在标签比对通过之后，将标签区域的纹理特征与标准纹理特征进行匹配；在匹配通过的情况下判定为对象图像属于真类别；在匹配不通过的情况下判定为对象图像属于伪类别。
识别设备5可以用硬件或者软件或者软件与硬件相结合的方式来实现。对于硬件实现的部分，可以在一个或多个专用集成电路(ASIC)、数字信号处理器(DSP)、数据信号处理器件(DSPD)、可编程逻辑器件(PLD)、现场可编程门阵列(FPGA)、处理器、 控制器、微控制器、微处理器、被设计以执行其功能的电子单元、或它们的组合中实现。对于以软件实现的部分，可以借助于微代码、程序代码或代码段来实现，还可以将它们存储在诸如存储组件之类的机器可读存储介质中。
在一个实现方式中，识别设备5包括存储器和处理器。存储器包含指令，该指令在被处理器执行时使得处理器执行根据本发明的实施例的识别策略/识别方法。
图2示出了设置有根据本发明实施例的识别设备5的读取器200。识别设备5可以实现为设置在读取器200中一识别软件。读取器200可以是固定式读取器、手持式读取器，等等。
可以理解的是，在识别设备5以软件形式实现的情况下，该软件可以适用于读码程序或产品验证程序，由此优化读码程序或产品验证程序的真伪识别功能。
图3示出了根据本发明的实施方式的识别设备5的示意性框图。如图3所示，识别设备5主要包括接收模块52、处理模块54、识别模块56和输出模块58。
应当注意，识别设备5的各模块的命名应当被理解为逻辑上的描述，而不是对物理形态或设置方式的限定。例如，接收模块52、处理模块54、识别模块56和输出模块58中的一个或多个可以实现在同一芯片或电路中，它们也可以分别设置于不同的芯片或电路中，本发明对此不进行限定。识别设备5的各模块还可以进一步划分为多个子模块，每个子模块实现为一个子程序。
图4示出了根据本发明的实施方式的识别过程400。该识别过程400可以在上述识别设备5中执行。
在框402中，接收模块52接收由读取器捕捉的对象图像。对象图像包含激光标签图像和标签区域图像，该标签区域图像是对象表面上的标签区域的图像。例如，图1中的激光标签“SN”所位于的表面区域R的图像。
在框404中，处理模块54对捕捉的对象图像进行处理，以获得该对象表面上的指定区域(即，产品标签上被激光烧刻而形成 激光标签的区域)的纹理特征。
纹理特征包含对象表面上标签区域的自然指纹特征，所述自然指纹特征与以下至少一项关联。
(1)对象表面上的标签区域的材质。
(2)形成该激光标签的激光打标过程中的工艺参数，例如，聚焦光斑大小、激光功率和光束质量中的一项或多项。
(3)采用的激光打标设备的类型(例如，光纤激光打标机、端泵红外/绿光/紫外激光打标机、激光雕刻机、CO2激光打标机)及其性能参数(例如，输出功率、光束质量、脉冲宽度、脉冲重复频率)。
(4)采用的激光打标设备以及打印材料的组合。考虑到不同的打印设备在同一材料上打印，会产生不一样的纹理特征；同一打印设备在不同的材料上打印，会产生不一样的纹理特征。因此，打印设备与打印材料的组合是影响纹理特征的重要因素。
在框406中，识别模块56将对象图像和存储的标准图像(例如，识别模块56从数据库4获取用于鉴别该对象/产品的相应标准图像)进行匹配，以识别该对象图像的真伪。标准图像包含作为匹配标准的基于激光打标的标准标签图像和包含标准标签区域的标准纹理特征的标准标签区域图像。
在该匹配中，可以将对象的标签图像和纹理特征分别与标准标签图像和标准纹理特征进行匹配。例如，首先将对象的标签图像与标准标签图像进行匹配，在该匹配通过之后，接着将对象的标签区域的纹理特征与标准纹理特征进行匹配。可以理解的是，关于标签图像的匹配以及关于文理特征的匹配两者也可以同时执行。
关于对象的标签图像与标准标签图像之间的匹配方法，识别模块56可以采用适合的图像处理和匹配方案，本发明不进行限定。
关于对象的纹理特征与标准纹理特征之间的比对，识别模块56可以通过提取多维特征向量并计算向量差的方法来实现(参见 框4603)。
在一实施例中，以标准图像作为样本A，对象图像作为样本B为例，识别模块56通过计算从样本A抽象出的多维特征向量Y A与从样本B抽象出的多维特征向量Y B之间的相似度(例如，通过两者的向量差来表示)是否符合一预定阈值来判断对象图像属于真类别还是伪类别。具体而言，识别模块56基于表征样本A的微文理特征的多个参数以及各参数的权重生成样本A的多维特征向量Y A(例如，标准特征向量)；基于表征样本B的微文理特征的多个参数以及各参数的权重生成样本B的多维特征向量Y B(例如，对象特征向量)。接着，计算标准特征向量Y A与对象特征向量Y B之间的距离，例如，欧式距离。接着，判断计算出的距离是否大于距离阈值。在计算出的距离大于距离阈值的情况下，判定为对象图像属于伪类别；在计算出的距离小于等于距离阈值的情况下，判定为对象图像属于真类别。
例如，基于样本A的标签区域的表面材质、激光打码过程中的工艺参数、和激光打码设备的型号及其设置参数而抽象出样本A的N维特征矩阵X A，X A∈R N；生成表征样本A的考量因素的权重的权重矩阵W A，W A∈R M,N+1；并基于样本A的N维特征矩阵X A和权重矩阵W A(例如，将X A和W A这两个矩阵相乘)生成样本A的多维特征向量Y A(例如，标准特征向量)。类似地，基于样本B的标签区域的表面材质、激光打码过程中的工艺参数、和激光打码设备的型号及其设置参数而抽象出样本B的N维特征矩阵X B，X B∈R N；生成表征样本B的考量因素的权重的权重矩阵W B，W B∈R M,N+1；并基于样本B的N维特征矩阵X B和权重矩阵W B(例如，将X B和W B这两个矩阵相乘)生成样本B的多维特征向量Y B(例如，对象特征向量)。接着，计算标准特征向量Y A与对象特征向量Y B之间相似度，例如，计算它们之间的欧式距离。标准特征向量Y A与对象特征向量Y B之间相似度可以通过如下公式来计算：
Figure PCTCN2021072381-appb-000001
当标准特征向量Y A与对象特征向量Y B之间相似度Similarity(A,B)小于预定阈值时，判定为样本B符合标准(即，判定结果为“真”)，例如，样本B为正品或者样本B与样本A属于同一批次的产品。
识别模块56可以借助于图像处理模型来完成上述匹配(即，对象图像与标准图像之间的匹配)。
该图像处理模型可以借助于人工智能技术来实现，例如，该模型实现为经训练的机器学习模型。该模型可以借助于适合的神经网络模型来提升其鉴别能力的鲁棒性。
在一实施例中，将大量的仿造图像和标准图像作为样本，训练该图像处理模型，以使得该模型能够在接收到新图像输入时，判断出该新图像是属于真类别(即，匹配结果为能够与标准图像匹配上)还是伪类别(即，匹配结果为无法与标准图像匹配上)。
在该实施例中，该图像处理模型还可以基于使用过程中的信息来进行再学习，以便周期性地获得更新参数，从而提升该模型的智能能力和处理速度。
下面描述一些场景下的匹配过程。
在一场景下，参见框4061，识别模块56通过匹配，判断对象图像对应的产品是否为同一批次的产品之一。在该场景下，只需要判断出对象是否属于同一批次产品之一，而无需识别出特定的激光标签或产品。换言之，在该场景下，只要判断为对象图像对应的产品属于同一批次产品之一，识别结果就为真；反之，识别结果为伪。
在该场景下，识别模块56可以将对象图像与表示一批次产品的标准图像进行比对，以判断对象图像的真伪，从而判断对象图像对应的产片是否属于所述一批次产品之一。
例如，采用同一批次的多个产品的多个标准图像(每个标准图像都包含标准标签图像和标准标签区域图像)以及多个仿制图像(每个仿制图像都包含防止的激光标签图像和标签区域图像)来训练图像处理模型，以使得该模型在输入新图像时，能够判断 出该新图像对应的产品是否属于该批次，并输出判断结果。
可以理解的是，即便是同一批次产品的图像之间也可能存在细微差别，而不会完全一致。通过大量批次样本的训练，使得模型能够鉴别出输入的新图像对应的产品是否属于同一批次的产品。
在另一种情况下，识别设备可以识别出对象图像对应的产品属于多个批次中的哪一个批次，或者不属于这些批次中的任一批次。
例如，识别模块56将对象图像与表示多个批次产品的各批次产品的标准图像进行匹配；在对象图像与表示一批次产品的标准图像匹配成功的情况下，确定为对象图像对应的产品属于该批次；在所述对象图像与表示任一批次产品的标准图像都匹配失败的情况下，确定为所述对象图像对应的产品不属于多个批次产品的任一批次。
在关于“批次”的实施例中，由于无需将对象图像与一批次产品对应的多个图像分别对比，可以节省大量的算力，提升识别效率。在另一场景下，参见框4062，识别模块56通过匹配，判断对象图像对应的产品的真伪。
在该场景下，识别模块56将对象图像与表示单个正品的标准图像进行匹配，以判断对象图像属于真类别还是伪类别，从而确定对象图像对应的产品的真伪。关于产品真伪的一些可能实例，参见上面相关描述，在此不赘述。
例如，采用标准图像和大量伪造品或不合规的产品的对象图像来训练图像处理模型，以使得该模型在输入新图像时，能够判断出该新图像对应的产品的真伪。
图5示出了根据本发明的实施方式的识别方法500的流程图。该识别方法500可以由上述识别设备5来执行。因此，以上相关描述同样适用于此。
在步骤S502中，接收对象图像，所述对象图像包含激光标签图像和对象表面上标签区域的标签区域图像。
在步骤S504中，对所述对象图像进行处理，以获得所述对象表面上标签区域的纹理特征。所述纹理特征与所述激光标签区域的材质、采用的激光打标设备和打标过程相关。
在步骤S506中，将所述对象图像与存储的标准图像进行匹配，以别所述对象图像的真伪，所述标准图像包含作为匹配标准的标准激光标签图像和包含标签区域的标准纹理特征的标准标签区域图像，其中，所述匹配包括将对象的标签图像和纹理特征分别与标准标签图像和标准纹理特征进行匹配。
本发明还提供机器可读存储介质，其存储有可执行指令，当所述指令被执行时使得机器执行如上所述的识别方法或识别过程。
可以理解，以上描述的方法或过程中的所有操作都是示例性的，本发明并不限制于方法或过程中的任何操作或这些操作的顺序，而是应当涵盖在相同或相似构思下的所有其它等同变换。
可以理解，以上描述的识别设备可以通过多种方式来实施。例如，可以被实施为硬件、软件、或其组合。
识别设备可以包括一个或多个处理器。这些处理器可以使用电子硬件、计算机软件或其任意组合来实施。这些处理器是实施为硬件还是软件将取决于具体的应用以及施加在系统上的总体设计约束。作为示例，本发明中给出的处理器、处理器的任意部分、或者处理器的任意组合可以实施为微处理器、微控制器、数字信号处理器(DSP)、现场可编程门阵列(FPGA)、可编程逻辑器件(PLD)、状态机、门逻辑、分立硬件电路、以及配置用于执行在本发明中描述的各种功能的其它适合的处理部件。本发明给出的处理器、处理器的任意部分、或者处理器的任意组合的功能可以实施为由微处理器、微控制器、DSP或其它适合的平台所执行的软件。
软件可以被广泛地视为表示指令、指令集、代码、代码段、程序代码、程序、子程序、软件模块、应用、软件应用、软件包、例程、子例程、对象、运行线程、过程、函数等。软件可以驻留 在计算机可读介质中。计算机可读介质可以包括例如存储器，存储器可以例如为磁性存储设备(如，硬盘、软盘、磁条)、光盘、智能卡、闪存设备、随机存取存储器(RAM)、只读存储器(ROM)、可编程ROM(PROM)、可擦除PROM(EPROM)、电可擦除PROM(EEPROM)、寄存器或者可移动盘。尽管在本发明给出的多个方面中将存储器示出为是与处理器分离的，但是存储器也可以位于处理器内部(如，缓存或寄存器)。
以上描述被提供用于使得本领域任何技术人员可以实施本文所描述的各个方面。这些方面的各种修改对于本领域技术人员是显而易见的，本文限定的一般性原理可以应用于其它方面。因此，权利要求并非旨在被局限于本文示出的方面。关于本领域技术人员已知或即将获知的、对本发明所描述各个方面的元素的所有结构和功能上的等同变换，都将通过引用而明确地包含到本文中，并且旨在由权利要求所覆盖。

Claims (12)

  1. 一种激光标签识别设备，包括：
    接收模块，其被配置成接收对象图像，所述对象图像包含激光标签图像和对象表面上激光标签区域的标签区域图像；
    处理模块，其被配置成对所述对象图像进行处理，以获得所述对象表面上激光标签区域的纹理特征，所述纹理特征与所述激光标签区域的材质、采用的激光打标设备和激光打标过程相关；以及
    识别模块，其被配置成将所述对象图像与存储的标准图像进行匹配以识别出所述对象图像的真伪，所述标准图像包含作为匹配标准的标准激光标签图像和包含标签区域的标准纹理特征的标准标签区域图像，
    其中，所述匹配包括将对象的标签图像和纹理特征分别与所述标准标签图像和所述标准纹理特征进行匹配。
  2. 如权利要求1所述的激光标签识别设备，其中，影响所述纹理特征的因素包括以下至少一项：
    -所述激光打标过程中的工艺参数；
    -采用的激光打标设备的类型及其性能参数；和
    -采用的激光打标设备和所述激光标签区域的材质的组合。
  3. 如权利要求1或2所述的激光标签识别设备，其中，所述激光标签识别设备还包括输出模块，其被配置成输出识别结果，以便确定所述对象图像对应的产品的真伪。
  4. 如权利要求3所述的激光标签识别设备，其中，所述产品的真伪包括：
    (1)对象图像对应的产品为正品或赝品；
    (2)对象图像对应的产品是否符合定制的要求，可选地，所 述定制的要求包括预先规定的产地、生产机构和分销渠道中的一项或多项；
    (3)对象图像对应的产品是否属于一批次产品；
    (4)对象图像对应的产品属于多个批次产品中的一个批次或不属于多个批次中的任一批次品。
  5. 如权利要求1-4中任一项所述的激光标签识别设备，其中，识别所述对象图像的真伪包括：
    将所述对象图像与表示一批次产品的标准图像进行匹配；
    在匹配成功的情况下，确定为所述对象图像对应的产品属于该批次；以及
    在匹配失败的情况下，确定为所述对象图像对应的产品不属于该批次。
  6. 如权利要求1-4中任一项所述的激光标签识别设备，其中，识别所述对象图像的真伪包括：
    将所述对象图像与表示多个批次产品的各批次产品的标准图像进行匹配；
    在所述对象图像与表示一批次产品的标准图像匹配成功的情况下，确定为所述对象图像对应的产品属于该批次；
    在所述对象图像与表示任一批次产品的标准图像都匹配失败的情况下，确定为所述对象图像对应的产品不属于所述多个批次产品的任一批次。
  7. 如权利要求1-6中任一项所述的激光标签识别设备，其中，识别所述对象图像的真伪包括：
    将所述对象图像与类型为真的产品的标准图像进行匹配；
    在匹配成功的情况下，确定为所述对象图像对应的产品为真；以及
    在匹配失败的情况下，确定为所述对象图像对应的产品为伪。
  8. 如权利要求1-7中任一项所述的激光标签识别设备，其中，所述匹配包括：
    基于所述对象图像的纹理特征的各参数及其权重生成多维的对象特征向量；
    基于所述标准图像的标准纹理特征的各参数及其权重生成多维的标准特征向量；
    计算对象特征向量与标准特征向量之间的距离，所述距离可选地为欧式距离；以及
    基于所述距离是否符合距离阈值确定所述对象图像的真伪。
  9. 如权利要求1-8中任一项所述的激光标签识别设备，其中，所述识别模块借助于图像识别模型实现所述识别，将所述对象图像作为模型输入，并经过模型处理后得到表示所述对象图像真伪的模型输出，
    可选地，所述图像处理模型为机器学习模型，
    进一步可选地，所述机器学习模型基于所述识别设备在使用过程中获得的信息执行再学习。
  10. 如权利要求1-9中任一项所述的激光标签识别设备，其中，所述识别设备设置在用于捕捉所述对象图像的读取器中；或者
    所述识别设备设置在用于验证所述对象图像对应的产品的真伪的验证设备中。
  11. 一种激光标签识别方法，可选地由如权利要求1-10中任一项所述的识别设备执行，所述方法包括：
    接收对象图像，所述对象图像包含激光标签图像和对象表面上激光标签区域的标签区域图像；
    对所述对象图像进行处理，以获得所述对象表面上激光标签区域的纹理特征，所述纹理特征与所述激光标签区域的材质、采 用的激光打标设备和激光打标过程相关；以及
    将所述对象图像与存储的标准图像进行匹配以识别出所述对象图像的真伪，所述标准图像包含作为匹配标准的标准激光标签图像和包含标签区域的标准纹理特征的标准标签区域图像，
    其中，所述匹配包括将对象的标签图像和纹理特征分别与标准标签图像和标准纹理特征进行匹配。
  12. 一种机器可读存储介质，其存储有可执行指令，所述指令当被执行时，使得一个或多个处理器执行如权利要求11所述的方法。
PCT/CN2021/072381 2021-01-18 2021-01-18 激光标签识别设备和识别方法 WO2022151460A1 (zh)

Priority Applications (3)

Application Number Priority Date Filing Date Title
CN202180090862.6A CN117321656A (zh) 2021-01-18 2021-01-18 激光标签识别设备和识别方法
EP21918662.4A EP4280198A4 (en) 2021-01-18 2021-01-18 LASER LABEL IDENTIFICATION DEVICE AND METHOD
PCT/CN2021/072381 WO2022151460A1 (zh) 2021-01-18 2021-01-18 激光标签识别设备和识别方法

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2021/072381 WO2022151460A1 (zh) 2021-01-18 2021-01-18 激光标签识别设备和识别方法

Publications (1)

Publication Number Publication Date
WO2022151460A1 true WO2022151460A1 (zh) 2022-07-21

Family

ID=82446794

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/072381 WO2022151460A1 (zh) 2021-01-18 2021-01-18 激光标签识别设备和识别方法

Country Status (3)

Country Link
EP (1) EP4280198A4 (zh)
CN (1) CN117321656A (zh)
WO (1) WO2022151460A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117725942A (zh) * 2024-02-06 2024-03-19 浙江码尚科技股份有限公司 用于标签纹理防伪的识别预警方法及系统

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118218779B (zh) * 2024-03-26 2024-10-01 中捷通信有限公司 针对箱体的标签设置方法、装置以及电子设备

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6723259B1 (en) * 1999-02-05 2004-04-20 Tesa Ag Anticounterfeit, laser-markable label produced by diffraction gratings
CN102646363A (zh) * 2012-04-20 2012-08-22 李峰 文字防伪、条码组合标识结构及方法
CN103177268A (zh) * 2013-04-19 2013-06-26 海南大学 一种基于dct变换的智能纹理防伪方法
CN203870645U (zh) * 2014-01-03 2014-10-08 微玺科技股份有限公司 防伪标示结构
CN107907124A (zh) * 2017-09-30 2018-04-13 杭州迦智科技有限公司 基于场景重识的定位方法、电子设备、存储介质、系统
CN109858582A (zh) * 2018-12-26 2019-06-07 胡黄龙 一种数码与其图像特征相结合的防伪方法
CN110544424A (zh) * 2018-05-28 2019-12-06 北京柯斯元科技有限公司 具有检测标识的随机纹理防伪标识结构

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110428028B (zh) * 2019-07-22 2023-01-10 杭州沃朴物联科技有限公司 基于准动态镭射标签的识别验伪方法、装置、设备及介质

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6723259B1 (en) * 1999-02-05 2004-04-20 Tesa Ag Anticounterfeit, laser-markable label produced by diffraction gratings
CN102646363A (zh) * 2012-04-20 2012-08-22 李峰 文字防伪、条码组合标识结构及方法
CN103177268A (zh) * 2013-04-19 2013-06-26 海南大学 一种基于dct变换的智能纹理防伪方法
CN203870645U (zh) * 2014-01-03 2014-10-08 微玺科技股份有限公司 防伪标示结构
CN107907124A (zh) * 2017-09-30 2018-04-13 杭州迦智科技有限公司 基于场景重识的定位方法、电子设备、存储介质、系统
CN110544424A (zh) * 2018-05-28 2019-12-06 北京柯斯元科技有限公司 具有检测标识的随机纹理防伪标识结构
CN109858582A (zh) * 2018-12-26 2019-06-07 胡黄龙 一种数码与其图像特征相结合的防伪方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP4280198A4 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117725942A (zh) * 2024-02-06 2024-03-19 浙江码尚科技股份有限公司 用于标签纹理防伪的识别预警方法及系统

Also Published As

Publication number Publication date
EP4280198A1 (en) 2023-11-22
CN117321656A (zh) 2023-12-29
EP4280198A4 (en) 2024-03-13

Similar Documents

Publication Publication Date Title
US12026860B2 (en) System and method for detecting the authenticity of products
US11449698B2 (en) Scanner with control logic for resolving package labeling conflicts
JP6308370B2 (ja) 識別方法、識別システム、照合装置及びプログラム
AU2015383137B2 (en) Methods and a computing device for determining whether a mark is genuine
WO2022151460A1 (zh) 激光标签识别设备和识别方法
CN1661627A (zh) 具有随机出现特征的防假冒和防篡改标签
US20170200247A1 (en) Systems and methods for authentication of physical features on identification documents
US20180033020A1 (en) System and apparatus for detecting forgery features on identification documents
CN107633413A (zh) 一种防伪溯源系统及方法
CN102292700A (zh) 用于增强安全印刷的系统和方法
CN112585615A (zh) 物件双重材料-数字防伪保护
CN112534775A (zh) 数字文件防伪保护
CN112308046A (zh) 图像的文本区域定位方法、装置、服务器及可读存储介质
Ishiyama et al. mIDoT: Micro identifier dot on things—A tiny, efficient alternative to barcodes, tags, or marking for industrial parts traceability
US20210042486A1 (en) Two-dimensional barcode generating method, verification method, server and two-dimensional barcode
WO2021102770A1 (zh) 用于验证产品的真伪的方法和设备
Rahman et al. An approach to facilitate business system by multiple barcode detection using faster rcnn
WO2023210228A1 (ja) コード認証システム及びコード認証プログラム
Smiatacz Face recognition: shape versus texture
WO2014177220A1 (en) Detection of presence of a bar code in image data
CN117893227B (zh) 一种饰品防伪标识生成、验证方法、设备及可读存储介质
Wang et al. A Recognition Method of Ceramic Microcosmic Images Based on SURF and Blockchain
CN111210238A (zh) 质保登记方法、装置、设备及介质
CN116415968A (zh) 用于实现对假冒产品的稳健并具有成本效益的大规模检测的方法和系统
Duerksen et al. System and method for authentication

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21918662

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 202180090862.6

Country of ref document: CN

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2021918662

Country of ref document: EP

Effective date: 20230818