RU2777951C1 - Label, as well as a method and system for automated product recognition using it - Google Patents

Abstract

FIELD: scanning devices.

SUBSTANCE: group of inventions relates to a label, as well as to a method and system for automated recognition of products by such a label. The essence of the claimed group of inventions is as follows. The label includes a base with a graphic code applied, which corresponds to the identification number of the recognized product. In this case, the graphic code contains at least four areas located in different parts of the label, each of which encodes a part of the identification number. At least two graphic elements are placed in each of the areas, one of which encodes at least one character of the identification number. In this case, another graphic element encodes the position of the specified symbol in the identification number. The product recognition system according to the claimed label includes at least one scanning device and a data processing unit. The method for recognizing the product according to the claimed label provides for scanning the label, obtaining images or visual displays of areas of the graphic code, recognizing graphic elements of the code, decoding data, determining the identification number for further identification of the product.

EFFECT: increase in the speed and accuracy of automated recognition of the product when bending or jamming the label on the product.

9 cl, 2 dwg

Description

[01] Field of technology

[02] The group of inventions relates to a label, as well as a method and system for automated recognition of products by such a label. The inventions can be used at customer self-service checkouts, as well as at automated warehouses when sorting goods.

[03] State of the art

[04] At self-service checkouts, an automated product recognition system is often used without the direct participation of the buyer and the store employee. In such systems, the buyer only puts the goods on the cash register, and they themselves move to the scanning area, where they are automatically recognized by scanning the graphic code applied to the product (barcode, QR code, etc.) from different angles. For goods by weight or goods produced directly by the store, the graphic code is printed on a label that is affixed to the product packaging.

[05] If the label is bent or jammed, for example, as a result of sticking it to the corner of the package or when the flexible package is deformed, automatic recognition of the graphic code is significantly more difficult. In this case, the processing unit receives several separate images or images of the label, on which there are only parts of the graphic code. In such a situation, the cash register data processing unit needs to match these parts of the image to obtain a single code that identifies the product.

[06] Algorithms for "stitching" pieces of graphic code are widely known in the art. In particular, in US patent US 10635877, 04/28/2020, a method for reading barcodes during automated product recognition is proposed, in which a plurality of code images are generated applied to the product, determining the three-dimensional position of each label segment, and linking the segments to read the code.

[07] This method is quite complicated and takes a lot of time. In order to speed up the recognition of a graphic code from several images of a label, it is advisable to abandon the stitching of code images and preliminarily divide the code into small areas that will encode parts of the identification number. Such areas will be guaranteed to be scanned entirely.

[08] Given this principle, the closest analogue of the claimed group of inventions is a label, as well as a method and system for product recognition using it, disclosed in the publication of US application US 2021073496, 03/11/2021. The label includes a base and a graphic code applied to it, which is divided into several physically separated areas. Each of these areas contains graphic elements that correspond to the symbols of the product identification number. In addition, the label has an augmented reality trigger in the form of a graphic element, which, in addition to visual functionality, can also set the position of recognized characters in the identification number. When reading the label, the processing unit receives an image of the parts of the graphic code and, using an augmented reality trigger, connects them in the correct sequence, receiving an identification number. However, in the specified analog, the method of dividing the graphic code into separate areas is used only for a decorative purpose in order to “hide” the code in the label image. In addition, all areas of the graphic code are located in the central part of the label, which does not allow it to be used for automated recognition of products where the label can be bent or jammed. At the same time, using an augmented reality trigger to determine the location of parts of the code is a complex and time-consuming process.

[09] Disclosure of the Invention

[010] The main technical problem to be solved by the considered group of inventions is the complexity of automated recognition of products by the labels pasted on them in case of deformation (bending, twisting, jamming, etc.) of such labels

[011] The technical result of the group of inventions is to increase the speed and accuracy of automated recognition of the product when the label on the product is bent or jammed.

[012] The specified problem is solved, and the technical result is achieved in the label attached to the product for automated recognition of the product, due to the fact that it includes a base and a graphic code applied to the base corresponding to the product identification number, the graphic code contains at least four areas , each of which encodes a part of the product identification number. These areas are located in different parts of the label. Each of said areas contains at least two graphic elements, one of which encodes at least one character of the identification number, and the other encodes the position of the specified character in the identification number.

[013] In addition, according to private label implementations:

[014] - the base of the label has a rectangular shape, and the areas of the graphic code are located in different corners of the base;

[015] - in the center between the areas of the graphic code there is an additional area containing graphic elements to check the recognition of graphic elements of the areas of the graphic code or to restore these elements in one of the areas,

[016] - the identification number symbol is a digit or letters, and the graphic element of the graphic code is an alphabetic or numeric symbol, or an icon, or a set of dots, or a set of strokes.

[017] The specified technical result is also achieved in the method of automated product recognition, due to the fact that it includes scanning a label using at least one scanning device, obtaining images or images of areas of a graphic code from at least one scanning device, recognizing graphic code elements in the specified areas on the images or images, decoding the data of the graphic elements, determining the product identification number by substituting characters in the appropriate positions of the identification number based on the decoded data, and matching the identification number to the database to identify the product.

[018] According to particular embodiments of the method, each of the obtained images or images contains only a part of the areas of the graphic code, while the label may be in a bent or wrinkled state during scanning.

[019] The technical result is also achieved in a system for automated recognition of products, which includes at least one scanning device, configured to scan the label in an automated mode, and a data processing unit associated with the scanning device. The data processing unit is configured to receive images or images of graphic code areas from at least one scanning device, recognize graphic code elements in said areas on images or images, decode graphic element data; determining the identification number of the product by substituting characters in the appropriate positions of the identification number, based on the decoded data, and matching the identification number with a database to identify the product.

[020] According to a particular embodiment of the system, the scanning device may be a laser scanner, an LED scanner, an imager, a camera, or a video camera.

[021] Unlike analogs, in the group of inventions under consideration, a label is used in which the graphic code is divided into four areas located in different parts of the label. This guarantees the possibility of separate recognition of each of these areas from different angles in almost any bend or wrinkle of the label in any plane. At the same time, in each of these areas, not only the value of the symbol of the identification number is encoded, but also its location (digit) in this number. This makes it easy to set the identification number in an automated mode when receiving several images (images) of the label from different angles by substituting the decoded characters in the appropriate position (digit) of the identification number. Since each of the areas is small (consists of two graphic elements), the possibility of recognizing only a part of such an area is excluded and there is no need to "stitch" image fragments. As a result, the speed and accuracy of product recognition is increased.

[022] Brief Description of the Drawings

[023] The invention is illustrated by the figures, where:

figure 1 shows an example of a label according to the invention in question,

figure 2 shows a scheme for recognizing a product by a label.

[024] The elements in the figures are indicated by the following reference numbers:

1 - label,

2 - base of the label,

3 - product,

4 - graphic code;

4a, 4b, 4c, 4d graphics code areas;

5 - identification number;

6 - a graphical element encoding the position of the symbol of the identification number;

7 - a graphic element encoding the symbol of the identification number;

8 - additional area;

9 - scanning device;

10 - data processing unit.

[025] the Implementation of the invention

[026] The claimed label (1) includes a base (2) that can be glued to the product (3), and a graphic code (4) applied to the base (1), which encrypts the identification number (5) of the product. Said graphic code (4) has at least four areas (4a, 4b, 4c, 4d) located in different parts of the base (2) of the label (1). The base of the label preferably has a rectangular shape, and the areas (4a, 4b, 4c, 4d) of the code (4) are placed in the corners of the base (2).

[027] In each of the indicated areas (4a, 4b, 4c, 4d) at least two graphic elements are placed, one (7) of which encodes the character of the identification number (5), and the other (6) the position (digit) of the specified character in the identification number (5). The graphic elements (6) and (7) can be, for example, alphabetic characters of various alphabets, numbers, icons, a set of dots, strokes, etc. The symbols of the product identification number (5) can be, for example, letters and/or numbers.

[028] As an example, in FIG. 1 shows the identification number (5) of the product as a four-digit number. In the graphic code (4) on the label (1), elements A, B, C and D (7) encode the numbers "1", "2", "3" and "4" of the number (5), respectively, and the elements γ, π, τ, ω (6) are the positions of these digits in the number (5), i.e. the digit of a four-digit number, where γ corresponds to thousands, π to hundreds, τ to tens, and ω to ones. In this example, it is possible to specify 10,000 identification number options. If more options are needed, letters can be added to the identification number or several symbols of the number (5), such as two-digit numbers, can be encoded in one graphic element (7).

[029] Between the areas (4a, 4b, 4c, 4d) of the code (4) in the center of the label (1), an additional graphic element (8) or several such elements can be placed, which make it possible to check the correct recognition of graphic elements (6, 7 ) code (4) or restore the values of elements (6, 7) in one of the areas (4a, 4b, 4c, 4d), for example, if part of the label (1) is damaged. Element (8) can be a symbol, an icon, a set of dots, strokes, etc. This element (8) may, for example, encode a checksum of the values of elements (6, 7) in areas (4a, 4b, 4c, 4d) for the purpose of checking values or recovering missing data. Such elements are widely used in barcode or QR code recognition and are well known to those skilled in the art.

[030] The system for automated recognition of products (3) by label (1) includes one or more scanning devices (9) and an associated data processing unit (10). The scanning devices (9) may be laser scanners, LED scanners, imagers (photo scanners), still cameras, video cameras, or a combination thereof. At the same time, the devices (9) provide the possibility of obtaining images (for an imager, photo or video camera) or images (for laser and LED scanners) of the label (1) from different angles. For this, several devices (9) located at different points in space, or one rotating device, or an omnidirectional device can be used. The system may also include a scanning tunnel and a conveyor for moving products (3) along said tunnel.

[031] The data processing unit (10) may be a structurally single or distributed device of several nodes. For example, the recognition functions may be performed by the scanner hardware, which will be part of a distributed data processing unit.

[032] The product recognition method (3) can be implemented as follows.

[033] The product (3) moves along the conveyor, and the label (1) glued to it falls into the field of view of one or more scanning devices (9). It is scanned and the data processing unit (10) receives images or images of areas (4a, 4b, 4c, 4d) of the graphic code (4) of this label. If the label (1) is folded over, as shown in FIG. 2, different areas (4a, 4b, 4c, 4d) of code (4) are recognized from different scanning angles, i.e. each image or image does not contain all areas, but only part of them. Next, block (10) performs recognition of graphic elements (6, 7) in each area (4a, 4b, 4c, 4d) in random order and decodes them, determining the symbols of the identification number and their length (receives parts of the identification number 5). Based on this data set, the data processing unit (10) restores the correct order of the symbols of the identification number (5). After that, block (10) compares the identification number with the database and determines the specific product (3).

[034] If necessary, the block (10) by means of scanning devices (9) can also recognize the elements of the additional area (8) and check the correctness of the code (4) or restore the data value in one of the areas (4a, 4b, 4c, 4d).

Claims (22)

1. A label (1) attached to a product (3) for automated product recognition, which includes a base (2) and a graphic code (4) applied to the base, corresponding to the identification number (5) of the product (3), and the graphic code (4) contains at least four areas (4a, 4b, 4c, 4d), each of which encodes a part of the identification number (5) of the product (3), characterized in that these areas (4a, 4b, 4c, 4d) are located in different parts labels (1), and in each of the indicated areas (4a, 4b, 4c, 4d) there are at least two graphic elements (6, 7), one of which (7) encodes at least one character of the identification number (5) , and the other (6) is the position of the specified character in the identification number (5). 2. The label according to claim 1, characterized in that the base (2) has a rectangular shape, and the areas (4a, 4b, 4c, 4d) of the graphic code (4) are located in different corners of the base (2). 3. The label according to claim 2, characterized in that at least one graphic element (8) is placed in the center between the areas (4a, 4b, 4c, 4d) of the graphic code (4) to check the recognition of graphic elements (6, 7) areas (4a, 4b, 4c, 4d) of the graphic code (4) or to restore data elements (6, 7) in one of the areas. 4. The label according to claim 1, characterized in that the identification number symbol (5) is a number or letters, and the graphic element (6, 7) of the graphic code (4) is an alphabetic or numeric symbol, or an icon, or a set of dots , or a set of strokes. 5. A method for automated recognition of a product (3) by a label (1) attached to it, made according to any one of claims 1-4, including: - scanning the label (1) using at least one scanning device (9), - receiving from at least one scanning device (9) images or images of areas (4a, 4b, 4c, 4d) of the graphic code (4), - recognition of graphic elements (6, 7) of the code (4) in the indicated areas (4a, 4b, 4c, 4d) on images or images, - decoding data of graphic elements (6, 7); - determining the identification number (5) of the product by substituting characters in the appropriate positions of the identification number (5), based on the decoded data, and - comparison of the identification number (5) with the database for product identification (3). 6. The method according to claim 5, in which each of the received images or images contains only a part of the areas (4a, 4b, 4c, 4d) of the graphic code (4). 7. The method according to claim 6, wherein the label (1) is in a bent or crumpled state when being scanned. 8. System for automated recognition of products by label (1) according to any one of claims 1-4, including at least one scanning device (9) configured to scan the label in an automated mode, and a data processing unit (10) associated with the scanning device, which is configured to: - obtaining images or images of areas (4a, 4b, 4c, 4d) of the graphic code (4) from at least one scanning device (9), - recognition of graphic elements (6, 7) of the code (4) in the indicated areas (4a, 4b, 4c, 4d) on images or images, - decoding data of graphic elements (6, 7); - determining the identification number (5) of the product (3) by substituting characters in the appropriate positions of the identification number, based on the decoded data, and - comparison of the identification number (5) with the database for product identification (3). 9. The system of claim 8, wherein the scanning device (6) is selected from the group consisting of a laser scanner, an LED scanner, an imager, a camera, and a video camera.

Images