RU2081453C1 - Scanning device which reads and decodes label and label to be read by said device - Google Patents

Abstract

FIELD: optical instruments. SUBSTANCE: label has data lattice which is shaped by manifold of polygons which encodes information and which are arranged in specific geometric structure. Said polygons are arranged in specific geometric structure and have at least two different optical characteristics. When information is encoded for label which consists of information-representing polygons, specific optical characteristics of polygons are set, polygons are arranged in given order and polygons are printed using at least two optical characteristics. When information is output, due to optical scanning of data lattice of information-representing polygons, precise optical copy of digital bit flow which conforms to optical characteristics of information-representing polygons is recorded. This optical copy is decoded and decoded bit flow is restored. Polygons are preferably hexagons. EFFECT: increased functional capabilities. 12 cl, 19 dwg

Description

 The invention relates to a scanning device for decoding an optically readable label containing encoded information, including an optically readable label, means for illuminating a predetermined area through which the label is passed, means for creating an optical image of the zone and generating a bit map corresponding to the brightness of the light reflected from the label, and driving each pixel of the image creating means, means for decoding the bitmap.

 Goods, various components, letters, packaging, containers and the entire gamut of relevant products that are transported or transported often need identification with information related to origin, flight number, destination, name, price, quantity of products and many other types of information. In some cases, reading the encoded information printed on labels attached to these products allows you to automate the reading of numbers related to the sale and inventory or operation of electronic cash registers. In other cases where such coded labels are used, automated tracking and sorting of soil, bags, baggage, etc. as well as placing labels bearing instructions for the manufacture of raw materials or components in the manufacturing process. Labels for products of these types are usually marked with bar codes, one of which is a universal product code. Numerous other code bar codes are also known.

 Commercial bar code systems usually suffer from a significant data density, which must adapt to the present and increasing need in order to encode more and more information on smaller and smaller labels. Attempts to reduce the total size and distance between the strokes in different code systems in order to increase the data density did not solve this problem: optical readers with significant resolution for detecting bar codes that include five mil or fewer strokes are usually not economically feasible are feasible in the manufacture, because the small tolerances inherent in the label printing process, and sophisticated optical equipment necessary to distinguish between bit-bar codes in these areas measures, greatly complicate the process. Otherwise, in order to adapt to increasing quantities, very large labels with a bar code must be produced, which will lead to such labels becoming completely non-compact for attachment to small products. Another important factor is the cost of the paper type label material. A small label has a lower paper cost than a large label; this cost is an important factor in high volume operations.

Alternatives to barcodes include: round shapes that use radially spaced wedge-shaped coded elements (I),
or concentric black and white binary coded rings (2, 3, 4),
grids of rows and columns of coded data squares or rectangles (5),
microscopic spots located in the cells forming a regularly located grid (6),
and densely packed multicolor data fields from dots or elements (7).

 Some of the coding systems described in the above examples and other coding systems known in the art suffer primarily from a lack of data density as in the case of coded circular patterns and grids on rectangular and square boxes. Or, in the case of grids composed of microscopic points or multicolor elements mentioned above, such systems require special means of orientation and transportation, limiting their use only to cases with highly controlled reading conditions.

Due to the size and speed of modern transport systems (using conveyor belts, for example, 0, 9 to 1.2 m wide), with belt speeds approaching 2.5 m / s or more, on which packages of different heights are placed on which coded information labels are affixed and the need to use a small, inexpensive compact label with an area of about one square inch, high voltages occur in optical and decoding systems designed to locate and read labels with data on these fast-moving packages, etc. products. There are difficulties in optical scanning simply detecting the image of the label. In addition, after detection or identification, the label image must be accurately decoded before the next packaging operation in the conveyor system, often this should be done in a split second. These problems have led to the need to create a simple fast and cheap means of signaling the presence of a label with encoded data within the field of view of the optical reader, which is installed so that it is possible to scan along the entire conveyor belt. This means is preferably connected to the high density data grid described below.
Data gratings containing prefabricated targets are known in the art, for example, these are concentric geometric shapes containing rings, squares, triangles, hexagons, and their many variations (8.9).

 The use of symbols containing concentric circles as indicators of recognition and position is also described, and these symbols are attached to products that need to be read optically (10, 11). However, in these systems two different symbols are used to determine the identification of the data field and its position, thereby increasing the complexity of the logic needed to identify the symbols, and also reducing the data capacity of the corresponding data field. Also, when two characters are used, damage to one creates problems in determining the position of the data field and for the operator's ability to recover information from the data field. The latter system uses separate position and orientation labels at opposite ends of the data tracks having linear encoded data labels of only limited data capacity.

 The devices described above mainly use scanning optically with a sensor capable of generating an output video signal corresponding to a change in the brightness of the light reflected from the data grid and the symbols of position and orientation. After quantization, the output video signal in such systems has a specific picture of bits, which can be consistent with a given sequence of bits. However, these systems suffer from the disadvantage of requiring two separate characters first for orientation. Also, due to the need to match the digital output signal of the optical sensor with a given sequence of bits representing symbols and positions, and orientation, erroneous readings are more likely than in the method according to the invention and in the system according to the invention, because in the known label detection systems an inflexible difference is provided recognition level of the target signal.

 In (1), a circular data lattice is shown having a centrally located recognition target containing a sequence of concentric circles. The recognition target provides the means for finding a round label with an optical sensor and determining its geometric center, and hence the geometric center of a round data grid. This is done using a logic circuit that recognizes an impulse diagram representing bull eye configurations for an identification target. However, with regard to barcodes, the data lattice has only a limited data capacity and the system requires a second circular scanning process. Due to the use of both linear and circular scans for a system with such a limited data capacity, an undesirable system complexity can be obtained for the sake of a slight increase in data capacity with conventional barcodes.

 To increase the capacity of these data gratings, codes have been developed that use many color points with high density (7). However, systems (7) require the use of hand-held optical image analyzing devices that are completely unable to record and decode when the data grids are quickly moved on packaging transported on a high-speed conveyor belt. Similarly, for high-density coding systems that use microscopic data coding points, (6) a special transportation means is required, which ensures that the data lattice moves in a given direction, and not just with a random orientation, as might be the case with baggage transported on a conveyor belt or the like. Thus, the coded label must be read strip by strip using a linear reader connected to the label transport means in order to accurately decode the information recorded on the label. Also in this patent it is shown that the position of the card relative to the sensor must be monitored very carefully so that it can be read.

 The technology also uses multiple colors in the manufacture of bar coding systems so that optical problems can be removed when scanning very small strokes. A bar code that uses more than two optical characteristics to encode data in a data lattice, for example, through the use of moving black, gray, and white strokes, is described in (12). However, systems of the type described, although they represent an improvement over previously known barcode systems, nevertheless, cannot boast of the compactness of the data density of the invention described here.

 From the point of view of these disadvantages of the known scanning devices for decoding a machine-readable label, the aim of the invention is to provide a scanning device for decoding a machine-readable label with new and improved compact optically readable labels with high information density, which can be read by an optical sensor when the label is attached to the baggage or otherwise a product transported by a high-speed conveyor system, regardless of the orientation of the baggage on it or the height of the baggage to which the optically readable label is attached, and with this system it was possible to reliably decode the label even in an inclined, curled, warped, partially erased or partially torn manner, and this system also includes correction possibilities so that was to restore incorrectly read or missed information and do so with preference for high priority encoded information, and moreover, this system also uses relatively inexpensive th logic.

 Other chains and advantages of the invention will become apparent from the description of the invention, which follows.

 The device according to this invention consists in a scanning device for decoding an optically readable label, comprising a lighting unit for a given area of the label, an optical image forming unit for the specified area with the label installed, creating a bit map, the elements of which correspond to the brightness values of the light reflected from the label, and exciting each element of the formed optical image, as well as the unit for decoding the bitmap into electrical signals associated with the output of the bl the formation of an optical image, the creation of a bitwise map and excitation of each element, the decoding unit being made in the form of sequentially connected two-dimensional block of clock synchronization signal reconstruction on an optical label, a geometric center detection unit for encoded label polygons when identifying their optical properties and a polygon decoding unit when inverting the process coding.

 In a preferred embodiment of the scanning device of the present invention, the two-dimensional clock synchronization signal recovery unit is made in the form of series-connected non-linear conversion of digital label signals with identification of optical properties transitions between adjacent polygons, a Fourier transform unit whose two-dimensional representation corresponds to the direction, length and brightness of the obtained transitions filtering block of converted digital signals with an exception incorrect direction and spacing of optical property transitions and an inverse Fourier transform, which provides the reduced clock signal.

 In another preferred embodiment of the scanning device of the present invention, a unit for normalizing the image information on the label to predetermined levels for each optical property is connected to the scanning device, associated with the optical image forming unit, creating a bit map, and exciting the elements.

 In one preferred embodiment of the scanning device of the present invention, a scaling unit on a label with the same horizontal and vertical magnification connected to the optical image forming unit, creating a bit map and driving elements is inserted into the scanning device.

 In yet another preferred embodiment of the scanning device of the present invention, a threshold processing and histogram building block is introduced into the scanning device, representing the optical image properties in each polygon of the label associated with the optical image forming unit, creating a bit map and generating elements.

 The essence of one preferred embodiment of the scanning device according to this invention is that the block detecting the geometric centers of the polygons of the label is configured to determine the region with maximum brightness in a given area of the restored clock synchronization signal and to provide a continuous search cycle for all such a signal from the region of highest brightness and at cyclic traversal of each adjacent region with the next highest brightness, with each of the detected regions corresponding etstvuet center polygons.

 The essence of another preferred embodiment of the scanning device according to this invention is that a block for detecting the zone of concentric rings of the optical label is introduced into the scanning device during the correlation processing of optical signals and a signal of a given frequency.

 In a preferred embodiment of the scanning device of the present invention, the optical image forming unit, the bitmap creation and the excitation elements are configured to filter analog signals corresponding to the light brightness values when determining the presence of concentric rings and an optical label in a given area.

 The optically readable label of the present invention consists of an optically readable label containing information-encoded polygons, the geometric centers of adjacent polygons being located at the vertices of a given two-dimensional lattice and the polygons having at least one of two optical properties, the polygons having three, five or more sides, located adjacent or partially adjacent and made with concentric rings in a zone separated from the zone with polygons, and each concentric ring A zo has one of at least two optical properties in an alternating sequence.

 In a preferred embodiment of the optically readable label of the present invention, the polygons are regular hexagons and the two-dimensional lattice is a regular hexagonal lattice.

 In another preferred embodiment of the optically readable label of the present invention, concentric rings are centered on the label.

 In yet another preferred embodiment of the optically readable label of the present invention, the optical properties are black, white, and gray.

 In FIG. 1 shows a top view of an identification target consisting of concentric circles in accordance with this invention; in FIG. 2 is a fragmentary plan view of an optically readable label having adjacent hexagons for encoding data in accordance with this invention; in FIG. 3 is a top view of an assembled optically readable label having adjacent hexagons having three optical characteristics for encoding binary data and containing an identification target, in accordance with this invention; in FIG. 4 is a top view of a three by three cluster of hexagons adjacent to one another, which can serve as a main coding unit in a preferred embodiment of the invention; in FIG. 5 is a cluster map, which graphically shows a data lattice containing 33 rows and 30 columns, forming a grid of 11 rows and 10 columns of the formation of the coding block of hexagons in the form of a figure three into three cells; in FIG. 6 is a schematic illustration of a camera setup system in accordance with the invention for adjusting the position of an optical light sensor in accordance with the height of the measured baggage; in FIG. 7 is a detailed description of the decoding process of the present invention; in FIG. 8 program structure of the encoding and decoding process and data stream; in FIG. 9 a program of a sequence of steps for processing an invention; in FIG. 10 is a top view of a set of adjoining regular hexagons arranged so that the geometric centers of adjacent hexagons lie on the vertices of the hexagonal lattice; in FIG. 11 is a top view of a set of adjoining irregular hexagons arranged so that the geometric centers of adjacent hexagons lie on the vertices of the hexagonal lattice; in FIG. 12 is a top view of a set of partially contacting polygons made essentially in the form of hexagons arranged so that the geometric centers of adjacent polygons lie on the vertices of the hexagonal lattice; in FIG. 13 is a plan view of a set of adjoining polygons essentially in the form of hexagons arranged so that the geometric centers of adjacent polygons lie on the vertices of the hexagonal lattice; in FIG. 14 is a plan view of an optically readable label having substantially adjacent hexagon-shaped polygons arranged so that the geometric centers of adjacent polygons lie on the vertices of the hexagonal array and includes an identification target in accordance with this invention; in FIG. 15 is a top view of a set of adjoining equilateral rectangles arranged so that the geometric centers of adjacent rectangles lie on the vertices of the hexagonal lattice; in FIG. 16 is a top view of a set of non-touching rectangles defining intermediate spaces between these rectangles so that the geometric centers of adjacent rectangles lie on the vertices of the hexagonal lattice; in FIG. 17 is a top view of a set of non-touching pentagons defining intermediate spaces between these pentagons so that the geometric centers of adjacent pentagons lie on the vertices of the hexagonal lattice; in FIG. 18 is a plan view of a set of adjoining squares staggered in rows and columns so that the geometric centers of adjacent squares lie on the vertices of the hexagonal lattice; in FIG. 19 is a top view of a set of partially touching octagons defining intermediate spaces between the polygons so that the geometric centers of adjacent polygons lie on the vertices of the rectangular lattice.

 The ability to encode information using contrasting colors of adjacent hexagons or "cells" located in a honeycomb structure with a given sequence and array allows you to restore information recorded on the label using an electro-optical sensor.

 Polygonal cells other than hexagons, which are located so that the geometric centers of adjacent polygons lie on the vertices of a hexagonal or other specified lattice, can also be used to encode information on an optically readable label. Such polygonal cells, when they are placed by their respective centers in certain places on a two-dimensional geometric lattice and when encoded in a given sequence by setting various optical characteristics to a plurality of such polygonal cells, can be “read” by an electro-optical sensor and subsequently decoded in accordance with the method of the invention described below.

 The polygonal cells of the invention are information encoded blocks formed by a closed polyline, these cells being placed in a predetermined two-dimensional structure on an optically readable label. Label configurations that utilize a wide variety of polygonal shapes, and lattices of various geometries, and so on, pentagonal, hexagonal, octagonal, rectangular or square lattices, can be applied using the invention. "Neighboring" polygonal cells can be completely in contact, partially in contact, or not in contact with the optically readable label of the invention.

 “Adjacent polygons” are polygons arranged so that the geometric centers of adjacent polygons lie on the vertices of a predefined two-dimensional lattice, and the boundaries of such polygons are in contact with the boundaries of directly adjacent polygons. “Partially contacting polygons” are polygons arranged so that the geometric centers of adjacent polygons lie on the vertices of a predefined lattice, and these polygons are separated in some places along their respective boundaries from other surrounding polygons, resulting in a multitude of intermediate spaces scattered among the polygons optically readable label. “Non-touching polygons” are individual polygons arranged so that the geometric centers of adjacent polygons lie on the vertices of a predefined two-dimensional lattice and they do not contact the boundaries of individual polygons and polygons surrounding a given polygon. In addition, polygonal cells and predetermined two-dimensional grids or lattices on which the centers of adjacent polygons are located may be irregular, have unequal axes, or regular, with equally spaced axes in the configuration. Such two-dimensional lattices have axes independent of the symmetry axes, if any, of polygonal cells.

 The hexagons used in the label of this invention represent certain advantages for encoding the information on the label. These advantages are.

 (1) At a given optical resolution, hexagons can be grouped more compactly than other polygons. For example, for a given resolution, the angles of the squares are considered with great difficulty, so even in other cases unnecessary resolution is required to “read” the squares. Circles would be the most optimal for optical resolution, but the space between adjacent circles would be wasted and the process of processing and printing the image of the label becomes more complicated due to the need to give optical characteristics to the gaps. Hexagons allow optimal packaging of information compared to circles or other polygons containing octagons, squares, triangles, etc. Squares and triangles create problems due to their sharp angles. Circles and octagons create problems due to the presence of unused gaps between adjacent circles or octagons.

 (2) A grid of adjacent hexagons has three axes. Through the use of square or rectangular labels, the main axis of the hexagon can be located in a predetermined ratio to the side of the label. This arrangement of the main axis of the hexagon grid makes it easier to read the data encoded in the hexagon by its relation to this main axis.

 In this application, the “label” includes a discrete part with a suitable sticky back side, attached to the package or article, to the outer surface of the container or other object on which the optically readable information is imprinted in accordance with this invention.

 In this case, an “optically readable data lattice” or “data lattice” means a picture of contiguous hexagons or cells having two or more optical characteristics for encoding in a reversible form, a data set due to the corresponding optical characteristics and the spatial relationship of the hexagons with respect to each other. Hexagons or polygons imprinted with the contents of this recoverable information are hereinafter referred to as hexagons or polygons with “coded information” due to the way the information is encoded on the label.

 This pattern of contiguous hexagons with a maximum number of hexagon-hexagon transition boundaries for optimal reading and maximum information density is called a “honeycomb structure”.

 The contrasting reflective characteristics used to print individual hexagons or data grid cells can vary greatly within the meaning of this invention. In this case, “printing” is the application of materials having predetermined optical characteristics on a substrate, or a change in optical properties when “thermal” printing is used. “Printing” also includes skipping the application of a material having a predetermined optical characteristic onto a portion of the substrate when the substrate itself has a distinctive optical characteristic.

 For example, when printing hexagonal cells in black and white, if the backing is white, then only black cells should actually be printed. Thus, in this case, the white hexagonal cells are also within the definition of the term “print” or “printed”.

 In this case, “optical characteristics” refers to light absorption, reflection and / or refraction of cells printed in various media. When the cells are printed in black (high density black ink), gray (grayscale black) and white (no printing on a white substrate), as in the case of the preferred embodiment of the invention, it is said that the invention has three optical characteristics.

 In this case, as shown in FIG. 1, “a plurality of concentric rings” or “concentric rings” 1 means two or more concentric rings 2, one of which is the inner region of the annular zone 3 defined by the smallest radius “r” of the rings.

In FIG. 2 shows a portion of an electro-optically readable label in accordance with this invention. As seen in FIG. 1, the label contains a plurality of adjacent printed cells in the form of hexagons forming a honeycomb structure. Each of the individual hexagons is indicated by 4 and contains six equal sides 5. The internal angles "a" of the hexagon are also equal to 120 ^o each. In the shown implementation example, the hexagon has a long vertical axis yy and a horizontal axis xx. The xx hexagon is slightly smaller than the yy hexagon 4 due to the geometry of the regular hexagon.

 In the preferred embodiment of the invention shown in FIG. 3, which uses a label 6 having dimensions of approximately one per inch, there will be approximately 888 hexagons or cells 4 (given that in the preferred embodiment, the centers of the label are occupied by an identification target 7 composed of a plurality of concentric rings).

 These contacting hexagons 4 naturally form horizontal rows "R" defined by imaginary line 8 and vertical columns "C" defined by imaginary lines 9. In this example, an inch-by-inch label has only 9 horizontal rows "R" and 30 vertical columns " C "hexagons 6. Each individual hexagon has a" diameter "of the order of 0.8 mm. There are more “R” rows than “C” columns in the square perimeter surrounding the honeycomb structure of the hexagons due to the geometric packaging of the touching hexagons.

 When using the hexagons shown in FIG. 3, it can be noted that the hexagons are staggered with overlapping vertical columns, with alternating vertically spaced hexagons having colinear y-y axes. The y-y axes of the spaced hexagons 4 are linearly aligned with the outer vertical side 5 of the intermediate offset hexagon. The y-y axes of the hexagons 4 20 are parallel to both vertical labels 10 and 11, as shown in FIG. 3. The horizontal rows "R" are measured along the x-x axes at the midpoint of hexagon 4.

 As described below, the hexagons 4 are formed by a printing process in which the hexagons 4 are printed in two or more optical gradations, for example, contrasting colors. These colors may be white 12, black 13, and optionally, but preferably gray 14, as shown in FIG. 3, although other contrasting colors may be used. You can use only two contrasting colors such as white 12 and black 13, as can be seen in figure 2. In a preferred embodiment of the invention, three contrasting colors are used: white, black and gray as shown in FIG. Specific shades of white, black, and gray are selected for optimal contrast when identified using an electro-optical sensor. The gray color level is chosen so that the optical gradations fall approximately equally between the optical characteristics of the white and black colors used to create the label.

 Label 6 shown in FIG. 3 can be obtained by using a discrete label having a preferred area of one square inch, or if an acceptable color background (preferably white) is used, the label can be printed directly onto the packaging surface without the need for a separate label . In view of the necessity and importance of having an optically controlled background for one of the contrasting colors, it is preferable to use a separate label because the background color of the label is easier to control.

 Aligning the hexagons printed on the same label with the sides of the label is important for the subsequent determination of the main axis of the label, as described below. The label is printed so that the y-y axis of the hexagon forming the honeycomb are parallel to the vertical sides of the labels 10 and 11, as shown in FIG.

 When “reading” a hexagonal lattice to decode information contained in individual hexagons, it is important to have a clear color contrast between adjacent hexagons. From the considerations given below, the smaller the optical characteristics used for encoding the hexagons, the simpler the scanning device and the software needed to decode the hexagons can turn out to be. However, with fewer optical gradations, the density of the data on the label also decreases. In the search for a compromise between the amount of decoded information that can be stored on the label and the cost of scanning labels with many optical characteristics, it was found that it is desirable to print coded hexagons with three optical characteristics, namely black, gray and white. If the backing or label has a good white background, when white hexagons can be done in the absence of ink, only gray hexagons need to be printed.

 In a preferred embodiment of the invention, gray hexagonal cells are created by printing the cells in black ink, but only every fifth pixel of the dot matrix printer image element grid is thus printed in the illustrative example described below. This is done by using the halftoning algorithm in a manner known in the art. This allows the printer to print a predetermined proportion of image elements defining a given gray hexagon, while printing a black hexagon requires printing each image element defining this hexagon.

 Black hexagonal cells can be formed by printing with standard black ink. As described below, the software for scanning analysis of the decoding process makes it possible to make strong distinctions between black, gray, and white characteristics, so that accurate color determination is not necessary. On the other hand, if colors other than black, gray, and white are used, or if different gradations of gray are used to create four or five-color data grids, the ink shade contrast should be controlled much more carefully to ensure measurable differences in optical performance between different colors. It should be noted that the use of black ink is the simplest and easiest approach when creating a honeycomb lattice with three optical characteristics of hexagonal cells and is preferred when implementing the present invention.

 Due to the square shape of the label in the preferred embodiment and the nature of the hexagonal cells, the edges of the honeycomb structure contain incomplete hexagons: as shown in FIG. 3, these incomplete hexagons are not used to convey any useful information.

 In a preferred embodiment of the invention, the label also contains an identification target. The identification target 7 shown in FIG. 3 contains a plurality of concentric rings of contrasting colors (shown as black and white). The black rings are respectively indicated by the numbers 15, 16 and 17, and the white rings are respectively indicated by the numbers 18, 19 and 20. The target is preferably located in the geometric center of the label so that it is less susceptible to damage or destruction in whole or in part if the peripheral part of the label is twisted and dirty or damaged. Also, the size of the image buffer (described below) needed to store data from the label before identifying the label is minimized when the identification target is in the center of the label.

 The number of concentric rings used in the identification target can vary, but it has been found that six concentric rings 15, 18, 16, 17, 19, and 20 and their resulting interfaces, varying from white to black, to white, etc., are sufficient.

 The structure correlation technology is used to coordinate the calculated structure, in which concentric rings are assumed in conjunction with the read structure. When matching occurs, the identification target is sought as described in more detail below.

 The identification target may have any total diameter smaller than the data lattice to create an area that can be 25% and preferably equal to about 7% of the data lattice area. Preferably, the identification target is made as small as possible since the area occupied by it on the label cannot be filled with encoded information. In a preferred embodiment, the diameters of the imprinted rings are chosen so that the outer boundary of the outer ring 20 is about 7.45 mm. Thus, in FIG. 3, the area of the identification target 7 is about 7% of the surface area of one square inch of the label 6. In this case, a satisfactory identification target 7 can be printed on the label 6 with an area of one square inch without violating the amount of information that can be encoded in the hexagonal array surrounding the identification target . As is the case with incomplete hexagons on the outer periphery of the label, piecewise hexagons in contact with the outer border of the identification target are not used to encode information. The width of each ring should be almost the same as the size side to side (x-x axis in figure 1) of the hexagons, which allows to improve the resolution. Six rings fit. This is a reasonable number that allows you to improve the location of the rings on the minimum area of the label with a minimum of possible erroneous readings from "fake" marks on the label and other "fake" marks not on a label such as marks somewhere on the conveyor belt.

 The identification target may take a form other than concentric rings. For example, squares, spirals or hexagons can be used to create transitions of contrasting concentric shapes, since linear intersections on an identification target allow you to create regular, defined and identifiable color transitions that can be sensed by an electro-optical sensor and measured with a suitable filter. It should be noted that although the spiral is not a collection of concentric circles, depending on the size and radius of the spiral, a close approximation of concentric circles can be obtained. A target of concentric circles is preferred because the signal generated by scanning through their center has a frequency that is the same as when crossing in any direction through the center of concentric circles. This makes it possible to simplify the identification of the center, which will be discussed more fully below, and allows identification of the location of the identification target in the one-dimensional search for the analog or digital output signal of the scanning device, although the method according to the invention allows the use of alternately or sequentially two-dimensional digital search for increased accuracy when analyzing digital signal.

In this case, "concentric rings" refers to complete rings, partial rings in the form of semicircles, sectors of concentric rings occupying from 180 to 360 ^o degrees, and concentric spirals that approach the concentric rings.

Since each hexagon can be encoded in three different optical characteristics, in a preferred embodiment, 1585 “bits” of information can be encoded in each hexagon (log ₂ 3). Obviously, if more or less optical characteristics are used than three, the number of bits encoded in each hexagon will vary proportionally. The encoding algorithm is designed to achieve as close as possible the maximum data density and increase the number of transitions of the optical characteristics from cell to cell, which makes it easier to two-dimensional master recovery, described below.

 In FIG. 4 shows a 3x3 cell set including nine hexagonal cells 21, which is the main coding unit used in the preferred embodiment of the invention. This is a desirable coding approach within the scope of the invention. As shown in more detail below, sets of 3x3 cells of hexagons 21 are mapped to encode 13 bits of information if the set contains a complete set of 9 hexagons or less than 13 bits if the set is incomplete due to the presence of unused hexagons. 1292 bits of information can be recorded on a single square inch label with a data grid containing about 888 hexagons and an identification target, which occupies about 7% of the label area.

When encoding each set, the outer, lower hexagons 22 and 23 in each set 21, as shown in FIG. 4, are limited in their respective optical characteristics, so that they are always determined to be different from the intermediate and contacting hexagon 24. Thus, only one bit hexagon can be encoded in hexagons 22 and 23. In this case, it is possible to encode 13 bits of information in set 21 by encoding 11 bits for the remaining seven hexagons. Since when mapping 7 hexagons, there are more possible combinations than used (for example, 3 ⁷ = 2187 combinations versus 2 ¹¹ = 2048 combinations), some combinations are rejected, such as all black, all gray, all white or essentially all black, gray or white combinations. The reasons for the contrasting colors of the hexagons 22 and 23 as compared to the hexagon 24 are the need to guarantee the transitions needed for the master recovery and arbitrary normalization described below, as well as the need for assistance in determining the horizontal alignment of the data grid, as described below. In cases where the coding sets have 7 and 8 hexagons, 7 useful hexagons are encoded with 11 bits, and the eighth hexagon, if any, is encoded with one bit. For all other partial sets, 3 bits are encoded on each pair of hexagons and 1 bit on each remaining single hexagon, which is described more fully below.

Therefore, it can be seen that the label is a particularly efficient, easy-to-read label (using appropriate scanning equipment and analytical software) for encoding very high-density information on a relatively inexpensive, easy-to-print label. As already noted, in a preferred embodiment, hexagons are used in a format of 33 rows • 30 columns in a one-inch label, with the identification target accounting for approximately 7% of the total surface area of the label. In practice, 13 bits of information are obtained from a set of 9 hexagons, so 1.44 bits of information are allocated per cell. This is less than the theoretical 1,585 bits per hexagon due to other conditions of the encoding algorithm, since all 3 ⁷ structures are not used, and some of the at least optically desirable cell-to-cell transitions are excluded.

 From the following considerations, in a preferred embodiment of the invention, it is desirable to introduce a certain amount of error protection into the encoding of the label, so that the actual amount of information being restored in the label is reduced in favor of a high degree of data integrity during decoding.

 The description of an example implementation of a label in which hexagonal cells are used is directly applicable to optically readable labels in which other polygonal cells are used. The described methods of "printing" the optical characteristics of a hexagon are equally applicable to printing the optical characteristics of other polygonal cells, both in black, white, gray (grayscale), and in other colors. Similar inconveniences and advantages with respect to data density are inherent to labels printed with polygonal cells other than hexagons when the optical characteristics of black, white and optionally gray are used to print polygonal cells. As with hexagon-containing labels, labels printed with other polygonal coding cells can be “read” with a smaller scanning device when only two optical characteristics are used to encode information in polygonal cells, in particular black and white therefore, with these colors maximum contrast is achieved.

 The information coding procedures and algorithms described for labels containing hexagons are directly applicable to labels printed with different polygonal cells. Similarly to hexagon-containing labels, cells with incomplete polygons that may appear on the border of an optically readable label, or which are obtained due to partial erasure by an identification target containing a sequence of concentric rings, are not used to encode information.

 The "honeycomb structure" comprises a lattice of adjacent hexagons 25, the geometric centers 26 of which also lie on the vertices 26A of the "hexagonal grid" or "hexagonal lattice" 27, as shown in FIG. 10. Regular hexagons, that is, hexagons having six equal sides and six equal internal angles, form hexagonal lattices, which are also correct in configuration, and have three equally spaced axes (A1, A2 and A3), which are located at an angle of 120 degrees relative to each other.

If the label hexagons 28 are irregular but symmetrical, for example, if the hexagons are stretched along the parallel sides 29, 30, the geometric centers 31 of the adjacent hexagons will describe the irregular hexagonal grid 32 shown in FIG. 11. Such an irregular hexagonal lattice will still have three axes (A1, A2 and A3) corresponding to the three symmetry axes of irregular shape of the hexagons, however, all three axes will not be equally spaced, i.e. the three axes will not be located relative to each other under angle of 120 ^o .

 Although the hexagonal lattice shown in FIG. 11 is not correct in origin, it is nevertheless a two-dimensional geometric grid or lattice having axes with a given location. Thus, the locations and intervals of the geometric centers of the hexagons located at the vertices of the intersecting axes of the hexagonal lattice are also given. Then, the geometry of the hexagonal lattice is used in the decoding process described below. In particular, the filtering step performed on the converted digital data corresponding to the image measured by the optical sensor is adjusted to reflect the predetermined geometry of the label, so that the digital representation of the measured label can be used to accurately restore the original grid. In the recovery process, missing points are also detected from the hexagonal grid. Missing grid points arise due to the fact that transitions of the optical characteristic did not occur between polygons with similar optical characteristics.

 With incorrect hexagonal grids of the type described in connection with FIG. 11, it is desirable to perform the step of determining the main axis, step (3) (e) in FIG. 7 of the decoding process carried out after the Fourier transform step of the process to identify the main axis of the optically readable label. The main axis of the label will have the geometric centers of the polygons lying along this axis at different intervals than the other two axes.

 Label configurations of the invention approximating a preferred embodiment comprising polygonal cells, as described above, are possible using certain polygonal cells. 12 shows a cell configuration in which polygonal cells 33 are used, which are very similar to hexagons, but which are 20 sided polygons, not hexagons. Similarly created polygons with the number of sides greater or less than 20 can also be printed. Polygons 33 are partially contiguous, in contrast to the imaginary contiguous hexagonal cells 34 in which they are depicted.

 Intervals 35 in the example label of FIG. 12 may or may not be printed with optical characteristics different from the encoded polygons. The intermediate intervals do not carry encoded information, therefore their presence leads to a lower data density. However, if the intermediate intervals scattered between the polygons have a different optical characteristic as opposed to neighboring polygons, then more transitions between the optical characteristics of the polygons and the intermediate intervals can be noticed by the optical sensor and therefore a signal of higher energy will appear in the conversion region during the encoding process, in detail described below, but the overall noise level of the system will also increase.

 Due to the fact that the polygons of the label shown in FIG. 12 are located on a hexagonal grid having three equally spaced axes, the geometric centers 36 of the polygonal cells 33 lie on the vertices of the hexagonal lattice 37. Since the patterns are symmetrical, the location, location and spatial orientation of the sides of the polygons turn out to be predefined and can be determined in the field of transforming the decoding process with only small modifications necessary in a two-dimensional recovery synchronization program. In particular, small modifications should be made with filters used to filter the converted digital data representing the coordinates, interval and brightness of the optical characteristics that were measured by the optical sensor in the illuminated image of the label. Such modifications will immediately become apparent to a person skilled in the art.

 Shown in FIG. 12 label used with hexagons. Individual polygons have a number of symmetry axes greater than three, however, due to the fact that they are so close to a hexagon, an optical sensor with a moderate resolution can "read" them as hexagons. The geometric centers 36 of the polygons 33 lie, however, at the vertices of the three equally spaced axes (A1, A2 and A3) of the hexagonal lattice 37.

 On Fig shows a similar shape (compared with the polygon 33 in Fig) polygonal figure 38, which can be made completely in contact. These polygons 39 can be described by the imaginary hexagon 39 shown in FIG. 12, but intermediate intervals cannot be found between the actual polygons (35 in FIG. 12). Such a construction in contact is desirable to simplify the decoding process, but it is not mandatory in the practice of using the invention. Polygons 38 are shown in a state where their respective geometric centers 40 lie on the vertices of the hexagonal lattice.

 And again, as with polygons 33 in FIG. 13, the polygons 38 are substantially hexagonal in shape, and with moderate optical resolution, they can be mistaken for hexagons.

 In FIG. Figure 14 shows the swelling of the label, which would have been possible if it had been printed with a dot matrix printer printing 200 pixels per inch. Polygons 40 in FIG. 14 indicate the shape of the geometric shape that will actually be printed instead of the hexagon by such a dot matrix printer due to the density of the image elements of the printer. Printers with a higher pixel density should provide a closer approximation to the hexagon than the polygons shown in FIG. 14. Thus, the polygons 38 in FIG. 13 and 40 in FIG. 14 are probably by-products due to the intrinsic limitations of some printers in the process of printing labels containing hexagonal cells or are obtained as a result of persistent attempts to print such polygons essentially in the form of hexagons in a first approximation. The shape of such polygons is essentially a hexagon and allows them to function in a practical sense as equivalents of coding cells in the form of touching hexagons.

 As in the case of FIG. 3, the optically readable label shown in FIG. 14 also contains an identification target 41 containing a plurality of concentric rings 42 47. Like the hexagons on the label of FIG. 3, the polygons 40 are substantially hexagonal in FIG. 14 are arranged in rows “R” a and pillars “C”, surrounded by imaginary lines 48 and 49 and 50 and 51, respectively. As with the hexagons in FIG. 3, the polygons in FIG. 14 have their geometric centers lying on the vertices of a hexagonal lattice defined by the equally spaced axes A1, A2, and A3. Thus, labels with the configuration shown in FIG. 14 are immediately encoded and decoded in accordance with the methods described below.

 If alternative label geometry using square, rectangular, pentagonal or octagonal mesh or similar structures is used, adjustments should be made to the two-dimensional synchronization restoration process described below. For various geometry of the predefined lattice, changes are required that must be made in the filters used in the filtering stage of the two-dimensional synchronization restoration process. Filters work with converted digital data corresponding to the optical characteristics of polygons read by the sensor in the image area. Such small adjustments to the filtering scheme can be easily made by one of ordinary skill in the art. In the case where the predetermined two-dimensional lattice has uneven axes, i.e., when it is incorrect in configuration, it may be desirable to determine the main axis of the label before performing the Fourier transform on digital data representing the optically measured invention. This is because the intervals of the geometric centers of the polygons will not be the same in the transformation region.

 Non-touching polygons can also be used to create an optically readable label in accordance with this invention. In FIG. 15 shows a hexagonal lattice of squares 52, which are arranged without contact, and their corresponding geometric centers 53 lie on the vertices of the hexagonal lattice formed by the three equally spaced axes A1, A2 and A3. Obviously, in this case, the configuration is a hexagon-based configuration of imaginary hexagons 54 described around polygons 52, whereby intermediate intervals 55 are formed.

 Similar grids for square 52 shown in FIG. 15 can line up using rectangles. In FIG. 16, a plurality of rectangles 56 are shown in which the geometric centers of adjacent rectangles lie on the vertices of the hexagonal lattice formed by the intersecting axes A1, A2 and A3. Again, the pattern of the hexagonal structure is complemented by imaginary hexagons 57 in FIG. 16 described around non-touching rectangles 56, whereby intermediate intervals 58 between the rectangles 56 are created. In FIG. 17 also shows a label lined with non-touching pentagons 59, in which the geometric centers of adjacent pentagons 59 lie along three equally spaced axes A1, A2 and A3. The geometry of non-touching pentagons is even easier to reproduce by describing pentagons 59 by imaginary hexagons 60, while forming intermediate intervals 61 between pentagons 59.

 An alternative hexagonal lattice can be constructed when the axes A1, A2 and A3 of the lattice are equally spaced but do not correspond to the symmetry axes of the polygonal figures. Instead, the geometric centers of adjacent polygons lie on the vertices of the intersecting axes of the lattice. Such a device is shown in FIG. 18, where there is a plurality of adjoining squares 62, in which the geometric centers of 63 adjacent squares lie along the axes A1, A2 and A3.

 Higher-order polygons can likewise line up on a given two-dimensional grid. In FIG. 19 shows a collection of partially contacting octagons 64 defining a plurality of intermediate intervals 65 between octagons 64. The centers 66 of adjacent octagons 64 are located at the vertices of the intersecting axes A1 and A2, thereby forming a lattice of octagons 64, which can be used in the practice of the invention. Intervals 65 can be imprinted with an optical characteristic different from that used for octagons 64. However, this is not necessary in practice when using the invention, since it is the location, orientation, and brightness of the optical characteristic in the center of octagons 64 lying in a predetermined position on the hexagonal the lattice formed by the axes A1 and A2 are the most important in the decoding process.

 It should be noted that although the preferred embodiment of the label is shown and described, numerous variations of the label are possible without departing from the scope of the invention. For example, a label does not have to have an area of one square inch. One square inch was chosen as a reasonable label size, allowing an acceptable data density of 100 alphanumeric characters of information with a high degree of error protection without creating an excessively large label size. It is advisable to have a one-square-inch label to reduce paper costs and other costs associated with printing, transporting and processing such labels. Conventional barcode labels of similar sizes had significantly lower data densities. Using 4, 5 or more optical characteristics or colors to highlight hexagons, you can pack significantly more information in a given space of hexagons of a predetermined size, but the complexity of the mathematical software and the sensitivity of the scanning system necessary to restore this information grows. Thus, for practical reasons, a coding system having three optical characteristics is highly desirable: black, gray and white. Also, the dimensions of the hexagons and the identification target can vary widely within the scope of this invention.

 Although the "bundling" of hexagons into sets of 3 • 3 cells is described above, other set structures can be applied or bundling can be completely eliminated, and the encoding algorithm can be targeted specifically to the individual structure of the hexagons. The relative amount of coded message information may also vary widely within the scope of this invention, as opposed to error correction.

The following describes the encoding process of the present invention in relation to a preferred embodiment of the label. It should be understood that the preferred embodiment is disclosed in the following description, but numerous combinations, variations and permutations are possible within the scope of this invention $
The process can begin with a predefined sequence of data that needs to be encoded on the label. In a preferred embodiment, the label is a freight label, and the data is divided into two fields defined as “high priority message” and “low priority message”. However, it should be understood that the invention is not limited to two different messages or priority levels. Many messages and priority levels can be created within the quantitative limits of labels of a given size and number of cells.

 For example, when the label should be a cargo label, a “high priority message” may specify nine characters representing the zip code of the recipient of the baggage, parcel or letter. We are talking about nine digits because although many individuals and firms have five-digit postal codes, nine-digit postal codes are becoming more common. Therefore, when handling baggage at the check-out, the most important part of the information is the postal code. It determines the main value of baggage and allows you to use various scanning and packaging control systems to direct the packaging to the right place on the truck, on the plane, in the container system, etc.

 A low priority message may contain, for example, a name and a loading address, including a zip code, the recipient of certain baggage, and account information.

 The basis for creating a high priority message and a low priority message is the need to protect the high priority message with excessive error correction, which allows you to place (encode) the high priority message in the more central area of the label, where it is less likely to be damaged or destroyed, and also repeat and distribute a high priority message in a low priority message so that even if the high priority message is partially destroyed, there is a great chance Note that a high priority message can recover from a low priority message. Due to the location of the high priority message in the central region, it may be necessary only to decode the high priority message for some purposes, so that only part of the label will have to be processed, which will speed up the processing. This can happen, for example, when the parcel is on the conveyor and you only need to determine the zip code to find out which of the several conveyor paths the parcel should go through.

 Due to the low priority, the low priority message is not shown twice on the label. However, as described below, both low and high priority messages may contain various error protection codes and correction devices in order to maximize the likelihood that both messages can be accurately restored.

 The use of error protection symbols as part of the encoded information may, in a preferred embodiment of the invention in combination with a corresponding recorded program and computer, cause the system to correct errors during the decoding process as described below. The use of error correction codes is well known in the art and is within the competence of a person skilled in the art.

 In practice, the label-sharing operator can manually enter data into an appropriate computer terminal, which is intended, as shown below, to turn on a printer that prints a label with a high priority message and a low priority message encoded accordingly in the hexagons of the label. It is not essential that a high priority message and a low priority message are actually created, but this is desirable in order to maximize the likelihood that the most important data to be encoded will be restored. In a preferred embodiment, the label is also printed with a centrally located identification target containing a plurality of concentric rings of two alternating contrasting colors, the colors being preferably selected from two colors used to print individual hexagons, and most preferably black and white to provide maximum contrast.

 The operator manually entering this data makes the appropriately programmed computer encode each character of the input message using a field identifier in order to create a binary bit sequence in the managed computer that represents the message characters and is encoded by the field to indicate messages of high priority and low priority and relative position everyone. This operation is carried out by the program "TEXTIN.C" (Fig. 8).

 Or the process may begin with the information to be encoded already contained in the binary bit sequence, because, for example, it was received from a storage device or created in some other way. Therefore, the message to be encoded can exist in a form that is manually (electronically) converted to a binary bit sequence or that starts as a binary bit sequence.

 Once a binary bit sequence has been created or an error-protected bit sequence has been produced in the step that will be discussed later, the bit sequence must be mapped in accordance with a specific mapping structure to encode the hexagonal honeycomb structure of the present invention. In FIG. 5 shows a “set map", which depicts individual hexagonal cells in sets of 3x3 cells aligned into a grid or honeycomb structure containing 33 rows and 30 columns of hexagons. Each row is numbered and each column is numbered. The row numbers go from 1 to 33, and the column numbers from 1 to 30. It can be noted that some of the hexagons marked along the top surface and the right-side surface of the map within the geometric center of the grid are denoted by Xs. This indicates that these hexagons do not contain bitmapped information. This is due to the fact that the outer Xs represent partial hexagons on the edge of the label, which leads to the presence in each of these rows of one smaller hexagon. The internal hexagons denoted by Xs represent gaps either occupied by the identification target or incomplete hexagons along the perimeter of the identification target, so that these internal hexagons denoted by Xs are not bitmapped. All hexagons that are not indicated by Xs are capable of recording information. In a preferred embodiment, each of these spaces should be occupied by black (B), e m (W) or (G) hexagons. As noted above, although various packaging and mapping techniques can be used, the application of this invention allows the use of packets of 9 hexagons in 3 rows and 3 hexagons in each to specify specific bits of information, and as described above, 113 bits of information are preferably encoded in each such pack of 9 hexagons.

 In a data grid containing 33 rows and 30 columns of contiguous hexagons, it is formed and can be seen in connection with FIG. 5 a grid of 11 rows of 10 columns of sets of hexagons, each of which contains a structure in the form of 3 • 3 cells of touching hexagons. It should be noted, however, that each set of 3 rows of cells per 3 columns of cells within the grid in the form of 11 sets x 10 sets contains a set of 7 or 8 hexagons due to the geometric packaging of the hexagons and the number will vary from row to row. Thus, 6 sets containing 8 hexagons, and 5 sets containing 7 hexagons, are obtained with this arrangement. Also, a centrally located identification target allows you to create additional incomplete sets. Thus, in FIG. 5 shows a graphical representation of the used sets of hexagons available for encoding with bits of information in the data lattice in the form of 33 rows of 30 columns of contacting hexagons.

As shown in FIG. 4, sets with nine usable hexagons are encoded using the following algorithm:
Take eleven bits of information and map their sequence of seven hexagons, designated as a, c, d, e, f and h.

 Hexagons g and i are used to represent 1 bit each so as to ensure that each one is different from hexagon h.

 Thus, thirteen bits of information are encoded in a complete bunch of 3 • 3 cells from nine contiguous hexagons.

For partial sets of 7 or 8 hexagons used:
Take eleven bits of information and map them into a sequence of the first seven hexagons used.

 The eighth hexagon, if any, is used to represent one bit.

For all other partial cells:
Map three bits of information in as many pairs of hexagons as possible.

 Any remaining single hexagons are used to represent one bit.

Since mapping seven hexagons produces more combinations than eleven bits (that is, 3 ⁷ 2187 versus 2 ¹¹ 204809, some combinations of hexagons must be rejected. Rejected combinations become those that provide the least number of transitions. To do this, lookup tables have been created for mapping sets in accordance with Fig. 5. The creation and use of these lookup tables is quite within the reach of a skilled programmer. In accordance with Fig. 9, the program for creating lookup tables "BI NHEH.LUT "and" HEXBIN.LUT "are designated as" MK HEX LUT ".

 The application of this bitmapping scheme allows the coding of 1292 bits of information in a data lattice of 33 rows x 30 columns of contiguous hexagons.

The sequence in which high priority information and low priority information is placed across the entire set map is preliminary, depending on:
(a) the size of the high priority message,
(b) the size of the low priority message,
(c) an optimal location for a high priority message in a secure location.

 Using the kit map shown in FIG. 5, as a template, the “MKMAPS. C” program with a mapping recording, operating on digital data contained in a storage medium, makes a preliminary determination of how to distribute information: both a high priority message and a low priority message throughout the dialing map, as will be fully described below.

 In order to minimize the likelihood of error and to provide the possibility of error correction, the preferred embodiment of the invention should include extensive protection against error and the possibility of correction. For example, in a preferred embodiment having 1292 bits of information that can be encoded in a one square inch hexagon array, having 33 rows x 30 columns of hexagons, and an identification target occupying about 7% of the label area, it is desirable to use 36 bits of high priority message information to encode a 9-bit zip code plus one additional alphanumeric character that can represent the loading code. In this example, it would also be necessary to use 120 check bits for a high priority message. This is determined by the number of necessary error correction capabilities. Similarly, in the example shown, 560 bits of a low priority message, this includes 40 bits of a high priority message that is included in the low priority message. In the example shown, 576 check bits of the low priority message are supplemented to maintain security and facilitate the recovery of the low priority message. This example illustrates the much more generous use of check bits to protect and recover a high priority message as opposed to a low priority message. It should be understood that the information provided is only an example, and a high priority message could be longer or shorter, a low priority message would be longer or shorter, the number of check bits is more or less, depending on the particular application of the invention.

 The "systematic code" takes a specific message sequence and adds a different error checking sequence to the message sequence. The “unsystematic code” receives a sequence of a specific message and includes an error checking sequence with a message sequence, so that the message is no longer different, but is of course recoverable. The use of either systematic or unsystematic coded for protection against error is within the competence of this invention. The following describes the systematic code.

 It has already been shown that the stage of "superposition of error detection symbols" includes systematic and / or unsystematic coding systems.

 Various systematic linear cyclic error protection codes are known in the art, for example, BCH codes, Reed-Solomon codes are separately included to protect the integrity of high and low priority messages. Reed-Solomon codes are very effective and most useful when multi-bit characters are checked for errors. Reed-Solomon codes are well known and it must be understood that this is simply the preferred option, although many other error correction codes could be used in the invention.

The following examples provide some information about the Reed-Solomon code. The specific characteristics of the Reed-Solomon code can be specified by the following parameters:
m is the number of bits in each character,
n number of characters in block 2 ^m 1
k number of message characters (number of message bits k)
t correction ability in number of characters (nk) / 2
A nine-digit postal code and a single alphanumeric character for further recognition purposes require 36 bits without error protection in the example described below. The Reed-Solomon code with the following parameters was selected for the high priority message.

m 6 (6-bit characters)
n 2 ⁶ 1 63
t = 10
Therefore kn 2t 43
Since only six 6-bit characters are required to represent a 36-bit message, the remaining 37 characters (43-6) characters are padded characters, which are implied between the encoders and decoders they do not need to be put on the label. Thus, the total number of bits required on the label for a high priority message is (63-37) • 6 or 156 bits.

 This error coding scheme will be able to correct a maximum of 60 (10 • 6) bit errors, which is up to 38.5% of the bits used. Due to the large number of implied padding characters, the great ability to detect errors in this Reed-Solomon coding makes it extremely improbable to read a high priority message incorrectly.

The low priority message was encoded with a Reed-Solomon code with error protection having other parameters, namely:
m 8 (8-bit characters)
n 2 ⁸ 1 255
t 36
kn 2t 183
Since there are 1292 bits for encoding on the label in accordance with this example, a total of 1336 bits (1292 156 bits of a high priority message and check bits) are available for the coding and verification bits of a low priority message. Thus, the remaining 904 bits (255 • 8 - 1136) should be implied by padding bits. This allows 560 bits (183 • 8 904) to be received for the information content of the low priority message and 576 check bits.

In order to better ensure recovery of a high priority message, it is also contained in a low priority message. The Reed-Solomon error protection code used for low priority messages allows encoding an additional 86 6-bit alphanumeric characters and has a maximum error correction capacity of about 25.4%
Using the described Reed-Solomon encoding with error protection, you can get a total number of 1292 bits of information on the illustrative label, distributed as follows:
36 bits of high priority information,
120 check bits of high priority,
560 low priority information bits (including 40 bits of a high priority message inserted in a low priority message),
576 low priority check bits.

 The household data sequence, including the corresponding check bits for storing information, is for individual hexagons on the set map of FIG. 5.

It should be noted that a wide variety of distribution schemes can be applied, given that the following important criteria that need to be defined are:
(1) the safe location of the high priority message near the identification target (if presented on the data grid),
(2) the creation of a structure that is reasonably capable of re-assembly upon reading.

 The specific error coding program used in the illustrated example is contained under the name "ERRCODE.C".

Reed-Solomon coding requires the multiplication of the message code vector by the generator matrix. The matrix is multiplied using the arithmetic of the Galois field. Adding any two field elements is obtained by performing the excluded "or" operation between both elements. Multiplication is done using logarithms in the Galois field. The logarithm and antilogarithm are obtained by using lookup tables derived from primary polynomials, especially for high priority messages: 1 + x ⁶ ; and for a low priority message: 1 + x ² + x ³ + x ⁴ + x ⁸ . As shown in FIG. 9, the auxiliary program "GF.0c" produces reference tables necessary for arithmetic of the Galois field.

Reference tables are computed and stored in the "GF.LUT" file for use during encoding and decoding. The generator polynomial g (x) for the Reed-Solomon code is determined by the following equation:
g (x) (x + a) (x + a ² ) (x + a ^2t )
Where
"a" is the base element of the Galois field.

The generator machine for the Reed-Solomon code is formed by conducting a long division for each of the rows of the generator matrix. Each row of the generator matrix is given by the remainder after a long division of x ^nk-1 by g (x).

 The calculation of generator polynomials g (x), as well as generator matrices for messages of high priority and low priority, is carried out in accordance with the auxiliary program "MKRSLUT.C".

 In a preferred embodiment of the invention, labels containing hexagons are printed with standard printing equipment, which is always affordable and affordable. A printer with a matrix of 300 • 300 dots per square inch gives good results when printing tri-color (black, white, gray) labels with 888 hexagons plus a centrally located identification target. The printer with this resolution is represented by the Hewlett-Packard Laser Jet 11 series model with a memory of 0.5 megabytes and a graphics resolution of 300 dpi. A grid of elements is depicted 300 • 300 having a density of 90,000 image elements per hexagon in a preferred embodiment. Each image element is assigned a value of 0 or 1, representing a black or white image element. This printer is used to print a two-color data grid from black or white hexagons. It can also be used to print a tri-color lattice of data from black, white and gray hexagons if the halftone algorithm is used to create gray hexagons, as described above.

 As shown in FIG. 8, using the recorded program “MKMAPS.C”, a lookup table “REGIONS.LUT” was created from 34 rows of 30 columns, which is similar to FIG. 5, but which is a device for indicating the choice of black or white for the rings of the identification target. Individual hexagons are encoded as black, white or gray, or as unused. A separate reference table "HEX MAP. LUT" was created by the recorded subroutine of the program "MKMAPS.C", which sets the belonging of each of the 300x300 image elements to the grid of image elements to specific areas in "REGIONS.LUT", that is, about 90 image elements per hexagon. Image elements belonging to the orientation rings are encoded in either black or white. The rings of the identification target are printed, firstly, by generating a hexagonal structure on a row of each region, and then generating rings. Areas partially or completely covered by search rings are considered unused in "REGIONS.LUT".

The encoded error protection bit sequence is mapped in accordance with a predetermined sequence into an array of 11 x 10 hexagons. As shown in FIG. 8, the sequence is defined by a custom lookup table “ORDER.LUT” generated by an auxiliary recorded program called “ORDER.C.”
The recorded program "PRLABEL.C" was used to set the values to 0, 1 or 2 areas available for printing on the label, while leaving the area with a value of 3 unchanged. The gray levels for each of the hexagons in the set of 3 • 3 cells are set in connection with a recorded program called "CELL CODE.C".

 The preference for memorizing a high priority message in an area close to the identification target, where it will be less prone to label degradation, is inserted into this auxiliary custom program. Therefore, the program "LABEL. C" is used to generate a sequence of bits suitable for input into a laser printer.

 You may notice that the use of black, gray and white colors allows for a simple label printing procedure, because black ink is only needed when using the standard halftone algorithm in a manner well known in the art. If other color combinations are used (which is quite feasible), the need to print in other colors obviously leads to significant difficulties in comparison with the tri-color black-gray-white version or the two-color black and white version.

 Thus, when a black or white value is assigned to each image element of the printer, labels can be printed to create an encoded format, as shown in FIG. 3, in which some hexagons are white, some gray and some black, and in which the identification area The target is preferably black and white concentric rings formed in the geometric center of the label.

 Having described how data is written to the label and how it is printed, it is necessary to describe the subsequent interpretation of the label or the decoding process. It should be noted that it is desirable to perform the function of interpreting the label at a very high speed, for a time of the order of a split second, in order to increase the efficiency of the baggage handling process (or other processing or reading from the label). There are two main approaches that can be used to capture an image while reading from a label. The label can be read at a relatively low speed using a static fixed focus manual scanning device. Or it is highly desirable to use an electro-optical sensor with a servo-controlled focusing mechanism that allows dynamic scanning of rapidly moving baggage of varying sizes and heights to achieve high-speed operation. The decoding process and device described below have been demonstrated in connection with a fixed focus scanning device. A process having the basic capabilities described here regarding a static fixed-focus scanning device is applicable to a dynamic scanning system with certain modifications of the optical system noted below. When processing baggage at high speed, it is desirable to have a high-speed scanning mechanism that allows you to read labels coming at a linear speed of about 2.5 m per second or more and passing under the installation site of a fixed reader. Thus, the image processing function comprises the following steps. 7 shows a diagram of the steps of the decoding process.

1. Lighting labels
When a package, parcel or letter passes on a high-speed conveyor, the area to be illuminated is large enough because the size of the package that is adjustable on the conveyor can be quite large and variable. For example, for baggage handling systems, a 1 m wide conveyor and packaging from a few inches to a meter wide (and similar heights) are unusual. Therefore, one square inch label can be anywhere across the conveyor. It is likely that the packages will also be located at oblique angles relative to the axis of movement of the conveyor belt. Parcels, packages, letters or similar items can have different heights, so that the scanned labels can be, for example, one inch or less above the conveyor, on the one hand, or up to 90 cm or more with respect to the maximum height of the package that can perceive the described system, on the other hand.

 In order to adequately illuminate the labels in accordance with this invention, especially taking into account the wide range of packaging widths, heights and viewing angles of the labels, it is desirable to use a high-brightness light source that reflects well on the basis of two or more optical characteristics selected for the label. The light may be infrared, ultraviolet, or visible light, and the light spectrum of the applied visible lighting may vary. The light measurement technique preferably includes measuring light reflected from black, white, or gray hexagons on the label.

The light source must provide enough reflected light to the light sensor (for example, a charge-coupled device described below) so that the light sensor can reliably distinguish among the black, gray and white or any other shade the optical characteristics of the hexagons to be measured. In a dynamic scanning system, an LED array can be used to create an illumination level of about 10 mW / cm ² in the label illumination area at the label level. LEDs can be located in a volumetric array without focusing lenses or in a linear array using a cylindrical focusing lens. A laser light source transmitted through an appropriate optical system creating a linear light source can also be used in this invention.

 The choice of the light source and the properties of the light source for the application in question is within the competence of specialists in this field. It must be recalled that since the label being searched has only one square inch at the maximum dimension, located at a height of up to 90 cm on a tape 1 m wide, running at a speed of the order of 2.5 m per second, it is very important to be able to illuminate the labels well with a view to being quite frisky identifying and finding labels.

In the case of a static sensor with a fixed focus used in the shown example, the level of illumination of the order of 2 mW / cm ² was sufficient for the application of the invention. This was accomplished through a luminescent light source.

2. Optical measurement of the reflected image of the label
The second stage of the decoding process recognition part is an optical measurement of the illuminated area by an electrically controlled sensor. The hardware-light sensor used in the illustrative example for a fixed-focus static scanning system contains an industrial high-quality color television camera on a charge-coupled device connected to a television lens, including a 5 mm extension tube, focus 50 mm, aperture 1.3, supplied by the company D.O. Industries, Inc. (Japan), and manufactured under the trademark NAVITRON ^TM . The camera is connected to an image capture card, designated as model number DT-2803-60, supplied by Data Translation Inc.

 Optical measurement may include viewing the entire label using a spatial sensor such as the camera and image capture card described above, or alternatively, it can be carried out with a linear array sensor containing a charge-coupled device on a chip in which the second measurement of label scanning is performed packing movement account (and labels). A suitable charge-coupled device chip is Thomson-CSF THX 31510 CDZ, element 4096 of a high-speed linear image sensor of a charge-coupled device supplied by Thomson-CSF, Division Tubes Electronics, France.

 In dynamic systems comprising moving baggage with a label in a conveyor system, it is desirable to have a long optical path between the recognizable labels and the light source. The primary goal of creating a long optical path is to reduce the change in apparent size or increase the label measured by the remote light sensor. For example, if the optical path is, say 1 m 20 cm, then the visible size of the labels located 2.5 cm above the conveyor belt will be very different from the size of the labels 90 cm above the conveyor belt. If an optical path of about six meters is used, the apparent size of the same labels will be the same. Due to this, it is possible, regardless of height, to fill the entire or almost the entire area of the light sensor with the area to be measured, which allows to achieve a high resolution image. If a spatial sensor is used rather than a linear sensor, the same principle applies. This can be done using the long optical path shown in FIG. 6.

 In order to be able to focus on labels of packages of different heights, a height sensor is needed. An ultrasonic sensor may be used or the sequence of light rays may be interrupted by baggage as a sensor. Any of these systems is applicable and can trigger an appropriate adjustable open-loop or closed-loop focusing mechanism for measuring and adjusting the position of the optical measuring elements (for example, lenses or sensors) relative to each other on a continuous basis, shown in Fig.6.

 In FIG. 6 is a schematic illustration of a camera focusing and adjusting system operating in accordance with the image for adjusting the position of the camera light sensor in accordance with the height of checked baggage. 6 shows a suitable lens 67, coil drive, height sensor and feedback loop in accordance with the image. The height sensor 68 shown in FIG. 6 may be an ultrasonic height sensor or a light beam that is interrupted by each package traveling along the conveyor. The output signal of the height sensor is supplied to the microprocessor 69, which, in turn, is connected to the drive 70 of the coil moving the coil 71 on which the charge-coupled device 72 or other suitable light sensor is mounted. The shaft position sensor 73 measures the position of the coil 71 and its output to the microprocessor 69 completes the feedback loop for measuring and adjusting the position of the coil 71.

 The sensor must be able to measure the reflected light coming from the illuminated label, and must also produce a tax signal corresponding to the brightness of the reflective characteristics of the label recorded by individual image elements of the electro-optical sensor.

 A suitable light source described above may be mounted on a mounting surface above the conveyor and covering an area extending across the entire width of the conveyor with light of a predetermined quality and brightness. The light reflected from the label can be bent by a sequence of reflectors, and then measured by an electro-optical sensor.

 The purpose of a curved optical path is to create a compact and therefore rigid system.

 The output analog video signal of the sensor is filtered. An analog electrical signal is used in conjunction with an analog bandpass filter to detect the presence of an identification target on the data gratings. Then, the analog signal is inverted into a digital signal using a conventional analog-to-digital converter inserted into the image capture card described below, or other known means. Instead of an analog bandpass filter, you can substitute a digital filtering circuit to determine the presence of an identification target by comparing the digital data representing it with the quantized output signal of an analog-to-digital converter, which will be described in more detail below.

 An example of a spatial sensor having a chip with a charge-coupled device with multiple detectors that was used in accordance with the invention is the previously described color television camera on a charge-coupled device of the Panasonic WV-CD130 type. The sensor output analog signal was applied to the DT-2803-60 type data capture image capture card described earlier, containing a 6-bit monochrome analog-to-digital video conversion for quantization and subsequent processing. Using a suitable recorded routine, the ordered digital output of the image capture card was stored in the memory as an exact replica of the image recorded by the optical sensor.

3. Reflected image processing
The most important part of the invention is the processing of an optically measured image in order to recreate and orient with accuracy the initial label configuration and color (optical characteristics) of each hexagon. This is done through the use of the following steps, after which the known structure by which the label was originally encoded and bit-mapped can be used to decode the information contained on the label.

 (a) Finding the center of the target.

 Before using the above-described television camera on a charge-coupled device and the image capture board shown in FIG. 10, the initialization program “DTINIT.C” 74 was run to set the image capture board to a known ready state and load the output lookup tables, followed by program "DTLIVE.C" 75, which sets the image capture card to "live mode". Then the "DTGRAB.C" program is controlled by the operation of the image capture board by quantizing the view in the image memory into 250 rows and 256 columns, and the samples are stored as 6-bit values aligned to the right in bytes.

 The auxiliary programs "DTSAVE.C" and "DTLOAD.C" allow you to transfer screen images to and from the storage device.

 Upon first familiarization with the label image, a conventional analog band-pass filter can be used to identify two or more characteristics of the concentric rings of the identification target. Both of these optical characteristics are preferably black and white because the largest contrast produces the highest energy signal. In order to find a fixed structure of the transition from black to white again to black, etc. it is desirable that when scanning linearly across the identification target and passing through the center of the target, a uniform frequency response is generated regardless of the orientation of the label. Thus, the target rings are optimally composed of contrasting concentric rings. Then the sensor output signal is bifurcated and selected in two directions of detection. In one direction, all the energy in the output signal is detected, and in the other, energy is measured at the ring frequency. When both output signals are compared, the energy in the ring detector is closest to the energy in the whole energy detector when the sweep is measured at the center of the identification target. The center of the identification target is found when this closest approximation occurs. However, in the dynamics of the preferred embodiment of the invention, the first filtering step preferably uses an analog bandpass filter or, moreover, a selective analog bandpass filter, although a digital filter is applicable.

 It should be noted that the step of finding the identification target, designated as “FIND.C” 76 in FIG. 9, is indicated as optional in FIG. 7, because a handheld scanning device can be used in the method of the invention, in which case the operator can accurately place scanning device to ensure proper alignment of the sensor. This, of course, turns out to be much slower than using an automatic sensor, and the use of an automatic sensor is preferred for high speed operation. If an automatic sensor (not manual) is used in the work, target detection is a necessary step in the process.

 As an alternative to the analog filter described above, a digital band-pass filter can be installed using the Parks-Mikklennan algorithm, which is supplied with the software "Software for the construction of digital filters for personal computers".

 A one-dimensional digital band-pass filter is used in connection with this invention to filter a normalized digital sequence of bits, as described further below, through the following filter routine. The filter band is the estimated frequency of the ring. A one-dimensional digital bandpass filter was designed for a sampling frequency of 400 pixels per inch and a length of 125 pixels (or 0.3125 inches), and is designed to operate on the basis of the size of the printed rings of the identification target, as shown in FIG. The frequency was 300/16 linear pairs per inch, with the output of the normalized frequency (where 400 pairs of lines per inch 1) to 300/16 • 400 or 0.046875. A filter with a band extending 5% below this frequency and 15% higher was chosen because distortion of the label usually leads to image shrinkage and therefore to an increased frequency. Stop bands were made from 15% below the frequency down to 0 and from 25% above the frequency of the ring to 0.5 (Nyquist limit). The filter coefficients were stored in the file "IMPULSE.LUT" 77, for Fig. 9, for subsequent operations, the first 62 coefficients are excluded, because the filter is symmetrical.

 A filter of 25 image elements along the length was performed by sampling a band-pass filter at the output intervals corresponding to the measured horizontal gain. For example, if the horizontal image gain is 80 pixels per inch, each fifth filter sample will be used (400/80 5 pixels). For non-integer steps, linear interpolation of adjacent filter samples is used.

 A second two-dimensional filter 25 by 15 pixels was also used. The sample values for this two-dimensional filter were based on the Euclidean distance of each point from the center of the filter, which were scaled for the corresponding horizontal and vertical amplifications. Then, for non-integer sampling intervals, linear interpolation is used.

 The output signal of the specified one-dimensional filter was squared and smoothed by a first-order recursive low-pass filter, which provided an exponential window of the past process. When the output of the smoothing filter exceeds a predetermined threshold, an optional two-dimensional filtering step was used to confirm the existence of the target and determine its exact location, as described below. In the first part of two-dimensional filtering, a reduced-size filter of 10 pixels per 10 pixels was used to save computations. This filter scans a rectangular area around the location defined by the one-dimensional filter. If the maximum two-dimensional correlation exceeds a predetermined threshold, then the final stage of two-dimensional filtering with a full filter of 25 by 25 image elements was used for a small square window around the maximum. If the best result of this filter exceeds a predetermined threshold, the center is detected. If none of the thresholds was exceeded, the program partially “discharged” the smoothing filter and returned to one-dimensional scanning. If the one-dimensional scan was completed without detecting the presence of the identification target, the program exited with an error return. (b) Normalization of the measured image.

 The brightness of the reflected light recorded by the optical sensor used in the process may vary due to variations in light, print density, paper reflectivity, camera sensitivity and other reasons, including label damage, such as bending, warping, etc. As an optional (but desirable) stage of reflections, the light measured by the sensor and transmitted to the memory can be normalized by the usual procedure. Using the technology known in the art, the recorded normalization program "NORM.C" 78 shown in Fig. 9 was used to analyze the brightness levels of the light reflected from the label, which is recorded by blocks of image elements in the scanning device, to find the minimum and maximum brightnesses of the reflected light, data recorded for the lattice The ordered digital output of the specified scanning device and combinations with the image capture card were loaded from memory for further processing using the specified recorded program frames of normalization.

 Using the equations y mx + b, where the minimum brightness substituted for x will give y 0, and the maximum brightness substituted for x will yield y 63, the recorded brightness of reflected light for each image element is adjusted so that black and the whitest colors present in the recorded image were set as the standard, and other shades of black, white and gray were adapted to these standards. Thus, the normalization step facilitates the processing of the measured image.

 (c) Recalculation of the image.

 For subsequent calculations, the recorded reflected image of the label is recalculated to create an image with equal horizontal and vertical amplification. Again, this is an optional step, but it makes it easier to quickly and accurately recover encoded information. The recounting operation was carried out to give the image uniform horizontal and vertical placement of the sample, for example, 150 pixels per inch, which is shown in the static example of the invention with a fixed focus.

 The recount operation occurs by calculating the fractional addresses of the rows and columns of the samples by 1/150 inch, based on the known horizontal and vertical amplification. Each point in the new uniform recalculated image is then allocated from the corresponding set of points in the image repeated in memory. To approximate the magnitude of points in fractional addresses, two-line interpolation is used. By recounting, the center of the label is placed in a known position in memory. The recalculated image is stored for later use in the search phase. It is then applied at all subsequent stages of the process that the recalculated image of the label centers at a known position on the grid, but it should be noted that this does not indicate the orientation of the label, which may be asymmetric with respect to the sensor.

(d) Two-dimensional synchronous recovery
The subsequent set of process steps is called collectively "two-dimensional synchronous recovery." The steps are performed using a suitable recorded program and routines called "CLOCK.C" 79, indicated in FIG. 9.

 This operation is performed in two dimensions on a recalculated image to accurately determine the position of the hexagon on the original data grid. The purpose of synchronous recovery is to locate and correct the effects of warping, curling, or bevelling the label, since the label cannot be completely flat. This is an important part of the process and it is not limited to hexagonal coded labels. This operation is applicable to other processes for decoding an encoded label that includes a regular two-dimensional grid, such as squares, triangles, etc.

 One-dimensional synchronous recovery is a concept that is well understood in the field of signal processing. Two-dimensional synchronous recovery is a continuation of this process and will become clear after some thought for a qualified specialist. You need to understand that the term "synchronous recovery" is a bit confusing for an expert, since it does not apply to synchronization.

 (i) Edge reinforcement and non-linear machining.

 The first stage in the implementation of synchronous recovery can be carried out using various nonlinear mapping operations known in the art for creating signal components at a given synchronization frequency, which are missed at the quantized image output from the optical sensor and image capture board. The goal of non-linear mapping is to take (preferably) a normalized and recalculated image that exists at this point during the process, and to form from it a two-dimensional non-linear map that enhances the transitions between adjacent contrasting hexagons. In a preferred embodiment of the invention, this is done by mapping standard deviation. This step can also be carried out by filtering with an image differentiation kernel, some of the means for which are known in the art, such as Laplace or Sobel kernels, and then the absolute value is determined or the results are squared.

 When mapping the standard deviation, an image with undifferentiated cell-to-cell edges is stored in memory. Then a standard deviation map is created to determine the location of the edges of the contacting contrasting hexagons by determining the standard deviations of a combination of 3 x 3 groups of image elements (this differs from sets of 3 x 3 cells) to determine standard deviations of the brightness of image elements. Standard deviation calculations are performed to determine the areas of image elements having a fixed color (smallest standard deviations) representing the inside of a hexagon or the interface between one-color hexagons, as opposed to groups of image elements having higher standard deviations, which represents transitions from a hexagon of the same color to an adjacent hexagon contrasting color. Because adjacent hexagons often have the same color, the standard deviation map does not completely highlight each hexagon. Due to the fact that during the standard deviation mapping process it is not possible to distinguish between interfaces between hexagons of the same color, missing borders or edges between hexagons are obtained. Other aspects of the synchronous recovery process are aimed at regenerating these missed transitions.

 The decoding process of this invention can be applied to any of the examples described above. Coding units of various geometries can be easily adapted and these encoded polygonal cells can be placed by the geometric centers of adjacent polygonal cells at the vertices of a known predetermined two-dimensional lattice.

 When the optically readable labels of the present invention are “read” by the optical sensors of the types described herein, the specific geometry or shape of the individual coding blocks or polygonal cells is not detected by the optical sensor. Instead, the sensor simply samples the optically readable label with a known number of samples per inch and records the brightness of the reflected light corresponding to the optical characteristic of the particular sample region that is shown. Then these values are stored in a storage medium for further processing. In other words, the electro-optical sensor allows you to record the average brightness of the light in the area beyond the area on the entire surface of the label, regardless of whether something is printed on the label or not. This is what is implied when recording images with edges that remain unchanged from cell to cell in memory. For this purpose, the decoding process immediately adapts to reading optically readable labels with large-range configurations, since the centers of polygonal coding blocks lie at predetermined intervals and directions on a two-dimensional lattice.

 It has been found in practice that variations of the hexagonal mesh coding system, as in the case of label examples using polygons essentially in the form of hexagons shown in FIG. 14, can lead only to modest decreases in the entire signal, and therefore to small decreases in the information capacity of the system. Using polygonal shapes with poor packaging or lattices of partially touching or non-touching polygons, rather than touching packaging, you can get a weaker, but nevertheless useful signal for many cases. However, at some point, the signal-to-noise ratio of the system due to the strongly pronounced shape of the polygonal coding cells, inefficient packing of cells and given two-dimensional gratings, leading to large intermediate intervals between the polygons, will drop to an unacceptably low capacity for storing information and recovery.

 The acceptability of the system depends on the quality of the signal restored by the electro-optical sensor. By altering the measuring system, for example, by increasing the number of samples per unit area on the surface of the label, it is possible to improve the recovery of the signal recorded by the sensor and improve the storage of information and the recovery characteristics of partially contacting and non-touching label configurations. Such adjustments to make less desirable label configurations usable are within the power of the signal processing professional.

 Therefore, the process allows you to widely push the boundaries of the useful signal-to-noise ratio. Thus, polygonal cells, whether they are regular or irregular in shape, can be used as coding units on the optically readable label of the invention. In addition, since the spacing and the direction of the centers of the polygons are known relative to adjacent polygonal cells, the polygonal coding cells can lie on a given lattice, and not on a hexagonal lattice, and the polygons can be placed in contact, with partial contact or even without contact on an optically readable label.

 As will be shown in more detail below, the nonlinear mapping technology, in particular the standard deviation mapping technology described here with respect to the preferred embodiment, facilitates the reconstruction of missing transitions or edges between polygonal cells with similar optical characteristics. Moreover, the same feature can overcome the loss of transitions between polygons and the intermediate intervals between polygons with similar optical characteristics. This is a situation where label configurations containing partially touching or non-touching polygons are used in practice. This feature is carried out using the subsequent fast Fourier transform, filtering and inverse fast Fourier transform.

 Through the use of optional technology in a preferred embodiment of the invention, it is possible to reduce the amount of calculation necessary to produce a map of standard deviations. Normally, to calculate the sum of nine image elements in a block of 3 • 3 image elements, eight addition operations would be required. This can be halved by replacing each element of the invention of the image itself with the sum of it and the image elements left and right of it. This will require two additions per image element. Then, the same operation is performed on the new image, excluding the amount calculated for the image elements immediately above and below. This will require two more additions, bringing the total number of additions to four. It can be shown that at the end of these steps each image element is replaced by the sum of himself and his eight immediate neighbors.

 Mapping of standard deviation is a necessary technology to create this map of hexagons corresponding to the original data grid, but only with gaps in the transition between the original hexagons of the same color.

 (ii) Crop.

 The following routine, called coding, is optional. Coding has been used in practice to apply the invention to reduce the brightness of borders that are not related to the contours of the hexagon. These boundaries arise at two points: the target ring and the uncontrolled invention surrounding the label. The weighing function is used to reduce the brightness of these areas. The details of how to use framing as a precursor to the fast Fourier transform are within the competence of a specialist.

 (iii) Two-dimensional fast Fourier transform.

 The two-dimensional fast Fourier transform of digital quantities corresponding to an (optionally) cropped standard deviation map is then performed under the control of commercially available recorded programs. During operation, the computer performs a fast Fourier transform of the image developed at the previous stage to provide a two-dimensional representation of the location, direction and brightness of the transition boundaries of the contrasting hexagons identified at the stage of standard deviation mapping. Simply put, the fast Fourier transform is a measure of the placement, direction, and brightness of the borders between the hexagons when they are known. Thus, the regular placement and orientation of the boundaries of the hexagons will allow certain points in the field of conversion to acquire a high energy level. The brightest point will be 0.0 in the transformation plane, corresponding to the DC component in the image. The six points surrounding the center point represent the placement, direction, and brightness of the borders between the hexagons.

 One of skill in the art will appreciate that, with respect to hexagons, the two-dimensional representation of the spacing, direction, and brightness of the interfaces of the contrasting polygons recognized at the previous stage of standard deviation mapping can be calculated by performing fast Fourier transform of the digital data corresponding to the measured label image. Thus, the gap and the directivity of the boundaries of the polygon will lead to the fact that some points in the transformation region will have high energy. The number of high-energy points surrounding the center point in the coordinates 0,0 of the transformation plane will depend on the geometry of the particular polygonal coding cell used in the manufacture of the optically readable label. For hexagons, however, such points surrounding the center point will represent the spread, direction and brightness of the edges between the polygons or the edges between the polygons and intermediate intervals if the label configuration is either partially in contact or in contact.

 Since the image is a real (and not complex) quantity, the transformation region is a point symmetric about the origin. Thus, only half the plane of the transformation region must be calculated, thereby saving almost half of the machine time. The exclusion of these calculations also reduces the number of attempts required during subsequent filtering of the image and inverse steps of the fast Fourier transform. The fast Fourier transform program used in connection with an illustrative example of a static system with a fixed focus was an available subroutine "R2DFFT" from the software package 87 "FFT-2" company Microway, Inc. Kingston Massachusetts.

 (iv) Image filtering.

 Next, a filtering process is needed to reconstruct the complete scheme of all the hexagons in the image area using the converted digital data. This can be done by eliminating any points of the transformation region that do not correspond to the specified placement and direction of the boundaries of the hexagons identified at the standard deviation mapping step. Six great points in conversion are due to the hexagonal honeycomb label design. In the transformation domain, only three points are actually identified, because the image is symmetrical in points relative to the origin, and the second three points can be implied from the first three. In a preferred embodiment, the filtering is carried out in three stages to exclude transitions from the step of mapping the standard deviation, which are too spaced, too close and / or in the directional direction.

 Firstly, high-pass filtering is performed by resetting to zero all points within a given circle around the origin of the transformation region, but at a certain distance outward from the origin, insufficient for six wonderful points located in the shape of a hexagon in the graphical transformation region. These points correspond to gaps larger than the gaps of the hexagons, and therefore carry information related to missing transitions in the label image. To recreate the missing transitions in the label image, it is necessary to exclude information about the missing transitions in the Fourier transform domain.

 Then, all points outside a certain radius, in addition to six remarkable points in the transformation region, are reset. This corresponds to false transitions that are too close together. This operation is combined with the first to form a ring from the remaining points. Creating this ring is equivalent to performing spatial band-pass filtering. The inner and outer radii of the ring are determined by the estimated variation in the contours of the hexagons. Since the “diameter” of the hexagon is supposed to be 5 image elements in the described example, and for a conversion length of 256 image elements, the vertices of the hexagon in the transformation area will be in 256/5 51.2 image elements from the center. Accordingly, a ring was used with an inner radius of 45 image elements and an outer radius of 80 image elements, which corresponds to the diameters of the hexagon 3,5 5,69 image elements. A filter with the advantage of transmitting higher-order frequencies was used because label deformations such as warpage and skew cause image shrinkage.

 After performing spatial band-pass filtering as described above, there is a ring with six remarkable points, with each point having an equal angular position with respect to the center / point 0,0 / transformation area. To complete the task of rejecting unnecessary information in the transformation field, the directional filtering step is used. Any point at a too large angular distance from the remarkable areas in the transformation area is reset. This leads to the fact that any borders that do not occur in one of the three directions dictated by the hexagonal honeycomb oblique structure are removed in the image area.

For conducting directional filtering, it is necessary to find the most remarkable points remaining after spatial band-pass filtering. Presumably this point is one of six remarkable points of the region and transformation similar to the vertices of a hexagon. Five other remarkable points on the same radius from the center and with angular placement of factors under 60 ^o are also obvious in the field of transformation. Therefore, all other points with an angular distance greater than 10 ^o from any of these points are eliminated. The six edges of the ring remain. Due to this step of directional filtering, any information about the incorrect placement or direction of the image area is eliminated. Eliminating this incorrectly placed information allows you to restore the complete picture of each hexagon in the transformation area.

 The previous discussion of the filtering scheme used for the preferred embodiment of a label containing contacting hexagons required modification when various predetermined two-dimensional gratings were used for the optically readable label. Nevertheless, it should be noted that only small changes to the filtration scheme are sufficient for a person skilled in the art to adapt to the various label configurations discussed above and shown in the accompanying drawings.

 As soon as they think about individual coding cells, it is predetermined that their respective boundaries will have certain angular positions and a given number of sides of a given length. Then it is necessary to determine the ratio of adjacent polygons, for example, whether they are in contact, partially in contact. It is also necessary to install a geometric lattice on which the geometric centers of the polygons will be located. Since the specified label geometry is predetermined, one skilled in the art can create an appropriate filtering scheme for filtering energy points in the transformation region, so that only the brightest points corresponding to the desired spacing and direction of the boundaries of the polygons are processed with a fast Fourier transform in the form of a subroutine.

 Regarding the actual filters made, it is necessary to understand that it is necessary to make an appropriately measured spatial band-pass filter based on a given diameter of polygonal coding cells. Then it is desirable to make a directional filter for filtering energy points different from the most remarkable points corresponding to the vertices of the polygonal coding cells. This eliminates any information regarding incorrect spacing or direction of polygonal coding cells in the image area and intermediate intervals, if any. By eliminating such incorrect information, a complete picture of the polygonal coding points in the image area is created. The digital data is then ready for the inverse fast Fourier transform in accordance with the process steps described below.

 (v) Processed Fast Fourier Transform.

 For a real return to the image region, thereby restoring the image image of the touching hexagons of the data lattice, it is desirable to perform two-dimensional inverse fast Fourier transform (2D-IFFT) on the filtered data of the transformation region. The inverse transform is carried out using the standard two-dimensional inverse Fourier transform (R2DIFT) routine available in package 87FFT-2 from Microsoft, Inc. Kingston Massachusetts. Upon completion of the inverse transformation stage, the picture of each hexagon is restored in the image area. In the new image, the centers of the hexagons are of high importance. The actual value of the spots at the centers of the hexagons depends on how many edges were in the neighborhood. More edges lead to more energy at the allowed frequencies and, therefore, to points with a high value. Fewer edges result in points with a lower value. Point value is a good measure of confidence level in synchronous recovery at any given point.

 (e) Definition of the major axis.

 The hexagonal image is now recreated, but it is necessary to determine its orientation.

The hexagonal honeycomb structure of the invention has three "axes" spaced 60 ^° apart. The direction of these axes is set at the brightest point in the transformation region after spatial bandpass filtering. Now you can verify which of these three axes is the main axis. This step is optional. If this step is not performed, the label should be decoded three times using each of the three axes, with only one axis allowing a meaningful message. The main axis is arbitrarily selected as the axis that runs parallel to the two sides of the label, as described in FIG.

 If the borders of the square label are determined based on knowledge of the main axis, then most of the energy in the restored hexagonal structure will be inside these borders of the square.

To determine the main axis, each of the three axes is assumed to be the main. The following figure of the square label is determined for each axis tested, and the total energy of the master structure is determined from the output digital energy data from the inverse conversion routine. A proper test is characterized by the highest energy. Then the angle of this major axis is recorded for the initialization step and other search operations. In this regard, it is not yet known whether the recorded angle is in the right direction or at 180 ^o from the right direction.

 It should be noted that it is not necessary to determine in general all three areas of the label, since the energy in the areas common to the three squares should not be determined.

 (f) Search.

 A recorded program called “SEARCH.C” 80, shown in FIG. 9, combines the converted and regenerated center information with the recorded brightness levels of the original image so that the gray level of each hexagon can be determined. Search is carried out in such a way as to minimize the possibility of "losses" in the search. The end result is to obtain a gray level matrix for each hexagon of the data grid.

 During the first part of the SEARCH.C program, four important information arrays are created, the CVAL (setpoint) array stores the quality measure of the reconstructed clock signal for each hexagon, while the GVAL array allows you to store the value (0 to 63) of the gray level in the center of each hexagon. The remaining IVAL and JVAL arrays allow you to remember the location of the row and column of the center of each hexagon.

 (i) Initialization steps.

 Based on the angle of the main axis determined in step (e) and the known placement of the hexagons (5 image elements) in the example, the estimated horizontal and vertical offsets from the center of one hexagon to the centers of the surrounding six hexagons can be calculated using a computer.

 Following these calculations, the SEAPCH.C program acts on the recovery driving signal received from the memory and the recalculated label image also received from the memory. The main goal of the initialization routine is to merge and condense information from these two sources and generate a data matrix that provides a gray scale value for each hexagon.

 The search initialization step is limited to the square around the center of the label with a side of about 1/3 of an inch. Within this region, the point with the highest value located in the reconstructed lattice of the driving signal is a good starting point. Then determine the location of this starting point relative to the center of the label. This starting point is the point where the driving signal is powerful and distinguishable, as well as the point relatively close to the center of the label. A powerful distinguishable signal is necessary to ensure that the search starts from the active center of the hexagon, and it is desirable that the point is close to the center of the label so that its absolute location can be determined without serious influence of twisting or skewing. A measure of the quality of a point in the synchronization restoration structure is the value of the point plus the size of the eight points surrounding it. The rectangular coordinates of the starting point are converted to a polar shape, the polar coordinates are adjusted relative to a predetermined angle of the main axis, and this result is converted back to a rectangular shape. These coordinates are scaled in accordance with the intended arrangement of the rows (4.5 image elements) and the arrangement of the columns (5 image elements) coming to the input position on the hexagon matrix. The synchronization qualities, gray levels, and placement corresponding to the initial hexagon are then introduced into the corresponding grids CVAL, GVAL, IVAL, JVAL.

 (ii) Main search cycle.

 The main search cycle determines the location of the centers of the remaining hexagons. The cycle ends when the estimated number of hexagons is found. The search order by the center of the hexagons is extremely important. The increased reliability of the decoding process in the face of label degradation comes from the particular search technology used, described below.

 Each iteration of the search cycle begins with a sample of the location of the synchronization recovery point of the largest magnitude, the neighbors of which were not found because of their strongest magnitudes. From this known point, the search will continue by one hexagon in each of the six directions. The effect is to erect the search structure along the trajectory from best to worse search structure along the trajectory from best to worst quality of the restored synchronization. Thus, if there is a weak region of restored synchronization, for example, in the center of the label or in the erased region, the search algorithm bypasses it, and does not follow through it. By circumventing these weak areas and preserving them in the end, they greatly reduce the possibility of losses and by preserving them in the last place greatly reduce the possibility of losses on the grid. Since losses are as bad as incorrect grayscale reading, this feature of the search algorithm is very powerful.

 The subroutine is responsible for finding neighbors of the best quality of the synchronization value in the main loop. The routine runs six times, one for each hexagonal neighbor of the hexagon in question. First calculate the location of the neighbor. If this neighbor is outside the label boundary, the loop iteration stops. If not, the neighbor is checked to see if he has already been found from another direction. The loop iteration will end if a neighbor is searched, because the algorithm makes earlier searches more reliable than later ones. If the neighbor remains outside this test, the estimated location of the center of the neighbor in the synchronization recovery structure is calculated. At this point, carry out a gradient search for the synchronization signal of the greatest magnitude. Eight elements of the image surrounding the restored position are sorted in order to see if a higher synchronization value is found. If found, then the best neighboring point has eight checked neighbors to check if there is any better value. This gradient sorting provides a degree of adaptation, which is essential if you need to read twisted or rickety labels. Then the routine goes to the next neighbor or returns when all neighbors are checked.

 As noted in step (d) above, as a result of the data conversion processes, the recreated image of the label now carries information related to the geometric centers of the polygonal coding cells. A polygon with more edges, i.e. transitions, is revealed to have more energy in the centers. The centers will lie on a predetermined two-dimensional lattice having a given number of equally- or unequally distributed axes, depending on the circumstances. Information related to the spatial enrichment of the axes of a given two-dimensional lattice can optionally be used at the stage of orientation of the main axis.

 However, it should be noted that the algorithm can be modified accordingly so that the decoding process determines the actual geometry of the two-dimensional lattice, and from this definition it follows in the direction of determining the filtering scheme, the so-called main axis of the label (that is, the axis of the two-dimensional lattice, which is parallel to two sides square optically readable label described here) and provides the necessary coordinates for the search routine.

 Regardless of whether the geometry of the label is determined by such an optional step as described above, or simply introduced into the decoding process by means of appropriate modifications of the two-dimensional process, restoration synchronization, the many label configurations described and shown here can be easily adapted by a person skilled in the art. it should be noted that the number of axes along which the centers of individual adjacent polygonal coding cells are located, and their corresponding orientation, can be substituted into the step of determining the main axis for all three axes of the hexagonal lattice in the preferred embodiment. Therefore, the main axis of a given two-dimensional lattice can be determined without checking and analyzing the errors described above in step (e).

 As for the hexagonal lattice of the preferred embodiment, the information from the step of determining the main axis and the known spacing of the polygons can be used to calculate the estimated horizontal and vertical displacements from the center of one polygon to the centers of the surrounding polygons. Following these calculations and after making the necessary adjustments to the search subroutine, the search, including the initialization phase, and the main search cycle step can be performed to configure the specific label that is used in this case. It should be noted that such minor adjustments to the SEARCH.C 80 enumeration program in the attached list of source codes are within the competence of a mid-level specialist in this field.

 After the completion of the subroutine, note the current location of the center so as not to sort it again. The effect is that this location is excluded as a candidate whose neighbors went through a search. For each iteration of the cycle, from 0 to 6 new candidates are added and one candidate is excluded. With good tools, a data structure can be used in which candidates are stored in order of magnitude as input and exclusion operations are performed. One such structure is called a priority queue.

It is known that a linear search algorithm requires about n ² operations, while a well-organized priority queue using a balanced tree or disordered structure requires about n log n operations. An n-order sorting algorithm based on sorting by groups can also be used if the values of the restored synchronization are scaled and reduced to a small range of integers.

 (g) Generating a histogram and setting a threshold.

 After the end of the main search cycle, the locations of the centers of all the hexagons are determined, and the gray values of all the hexagons that are recorded turn out to be completely filled. The next step is to limit the digital values of the gray level in the range of 0 to 63 discrete levels, for example, black, black, gray and white (for black, white and gray labels). This is done by constructing a histogram of the brightness values of the label image from the centers of the hexagons. Quantized levels can be determined by finding dips in the histogram.

 (h) Rough grid correction and final orientation.

 After limiting the discrete levels, there may be two distortions. First, the grill may be out of center. This can happen if at the initial stage of the search the location of the synchronization signal of the best quality relative to the center of the label is not correctly determined. The second possibility is that the entire label is effectively read from top to bottom, since the angle of the main axis has an ambiguity of 180 degrees.

 The recorded subroutine allows the function to determine whether the label is offset from the center. If the label is positioned correctly, the coordinates of the center row should go through the center of the label. To determine whether a vertical layout error was made, the rows above the hypothetical central row are checked to see which ones form the line closest to the center of the label. If the row above or below is closer than the hypothetical central row, then produce a corresponding shift up or down. If the left straightening of short rows was not done correctly, this is adjusted by shifting the short rows one position to the right.

 Errors of horizontal position and reading upside down are checked using the information entered into the label and known as coarse mesh information. Information is distributed in sets of 3 • 3 hexagon cells, as described above. Since the label can be, for example, with a grid of 33 rows of 30 columns, these sets form a grid of 11 by 10. The lower central hexagon from each complete set of 3 • 3 cells has a special property that is laid down during encoding. There is a guaranteed transmission on either side of this hexagon, as previously mentioned in connection with figure 4. For example, if the lower central hexagon is black, the lower left and lower right hexagons must be either gray or white. The recorded routine takes advantage of the transfer property to remove the final two possible distortions. First, a lattice is created, where each lattice element indicates whether a transition has occurred between two horizontal adjacent hexagons. Then, the lattice is checked for every 9 hypothetical slides of the coarse grid located in the 3–3 structure around the supposed slip 0. One of these slides will show the best agreement between the real and the supposed transitions and this slip position will be preserved. Then the same hypothesis is tested with the assumption that the label is read upside down. This will happen if the angle of the main axis is indeed directed from right to left relative to how the label was printed, and not from left to right.

 If the label was simply turned upside down, that is, the upper rows swapped with the lower ones, and the upper columns with the lower columns, then the slip results are also inverted. However, for the label to flip correctly, one important conversion is necessary. When reading short (length 29) rows, the left ones are checked, so when the label is turned over, these labels must be checked correctly. Regulation has been made, and this is precisely the procedure that will make the results of the slip hypothesis not easy to turn over. In fact, the best slip test results will be better than any previous tests if the label was actually read upside down.

 By determining whether the label has been read upside down and if there is any slippage in the absolute location, the label matrix can be encoded. With the correct definition of the image and slip, the image processing functions are completed and data decoding processes begin.

 4. Decoding.

 The recorded program "RD.LABEL.C", shown in Fig. 8, allows you to read the file generated by the enumeration program and generates a file of a sequence of bits, which in the preferred embodiment is 1292 bits. In this case, the recorded CELL DEC.C subroutine is used with Fig. 8 for masking unused hexagons and for applying a decoding program, which is an inverse of the encoding program.

 The first step in the decoding process is to generate a sequence of bits from the information of the hexagons using the hexagon-bit mapping process, which is the reverse of the bit-hexagon mapping process used in the encoding operation.

 The bit (information) sequence is then bifurcated by the program into a bit sequence of a high priority message and a bit sequence of a low priority message, or as many bit sequences as are used for label coding.

Then, error correction must be applied to each bit sequence using the error coding technology that was used in the label coding process. For example, if Reed-Solomon coding is used, when correcting the error on the bit sequence generated by the enumeration program, an output signal is generated that exists in the same format as previously described for encoding the input file. Error correction can be performed in the following sequence:
1. Calculate the syndromes.

 2. Calculate the error finder polynomial using the Berkelekamp-Messi algorithm.

 3. Calculate the location of the error using Chen enumeration.

 4. Calculate the error values using the Forney algorithm.

 The last step is performed only if the correctable number of errors is detected in steps 2 and 3. The number of detected errors is also calculated. If an uncorrectable number of errors is detected or if the error is located in a significant packing (described above), then check the box. The specific error coding procedure used in the example shown is indicated by “ERRDEC.C” in FIG.

 5. Conclusion.

 When following baggage (when identifying its location on the conveyor), a high priority message indicating the destination zip code can be used to turn on suitable guide levers or conveyors to direct baggage to the appropriate truck, plane or baggage car delivering the baggage for its intended purpose.

 Although the invention can be used on a conveyor-deflecting system, it is obvious that it can be applied in a wide range of information collection, baggage processing and production operations, in which it is desirable to read the label on baggage, letter, part, machine or similar device and force the system to perform processing baggage or a manufacturing operation, for example, having a label. The invention allows these operations to be performed with high speed, high accuracy, associated with a significant amount of label information, and even protect most of the information from loss due to label breaks or other damage.

 As shown in FIG. 8, the program “TEXTOUT. C” can be used to periodically display a decoded message on a computer terminal.

Claims (12)

 1. A scanning device for decoding an optically readable label, comprising a lighting unit for a predetermined label propagation zone, an optical image forming unit for a predetermined area with an installed label, creating a bit map, the elements of which correspond to the brightness values of the light reflected from the label, and exciting each element of the generated optical image, as well as a unit for decoding a bit-wise card into electrical signals associated with the output of the optical imaging unit, Creating radix maps and excitation of each element, characterized in that the decoding unit is formed as a series of blocks of two-dimensional signal recovery clock on the optical label neometricheskih center detection block encoded polygons label for identification of their optical properties and polygons decoding unit for inverse encoding process.  2. The device according to claim 1, characterized in that the block of two-dimensional reconstruction of the clock synchronization signal is made in the form of series-connected block of nonlinear conversion of digital signal labels with identification of transitions of optical properties between adjacent polygons, a Fourier transform unit, the two-dimensional representation of which corresponds to the direction, length and brightness of the received transitions, the filtering block of the converted digital signals with the exception of the wrong direction and spacing transition optical optical properties and the inverse Fourier transform block providing the reconstructed clock synchronization signal.  3. The device according to p. 2, characterized in that the unit is introduced to normalize the image information on the label to the specified levels for each optical property associated with the unit for generating an optical image, creating a bit map and exciting elements.  4. The device according to claim 2, characterized in that the unit for zooming on the label with the same horizontal and vertical magnification is introduced, connected with the block for generating an optical image, creating a bit map and exciting elements.  5. The device according to claim 2, characterized in that a threshold processing and histogram building block is introduced that represents the optical image properties in each polygon of the label associated with the optical image forming unit, creating a bit map, and generating elements.  6. The device according to claim 1, characterized in that the block detecting the geometric centers of the polygons of the label is configured to determine the area with maximum brightness in a given area of the restored clock synchronization signal and to provide a continuous search cycle for all such a signal from the area of greatest brightness and with a cyclic bypass of each adjacent areas with the next highest brightness, with each of the detected areas corresponding to the center of the polygon.  7. The device according to claim 1, characterized in that the unit for detecting the zone of concentric rings of the optical label is introduced during the correlation processing of input signals and a signal of a given frequency.  8. The device according to claim 1, characterized in that the unit for generating an optical image, creating a bit map and driving elements is configured to filter analog signals corresponding to light brightness values when determining the presence of concentric rings and an optical label in a given area.  9. An optically readable label containing information-encoded polygons, wherein the geometric centers of adjacent polygons are located at the vertices of a given two-dimensional lattice and the polygons have at least one of two optical properties, characterized in that the polygons have three, five or more sides, are adjacent or partially adjacent and made with concentric rings in a zone separated from the zone with polygons, and each concentric ring has one of at least two optical voystv in an alternating sequence.  10. The label according to claim 9, characterized in that the polygons are regular hexagons, and the two-dimensional lattice is a regular hexagonal lattice.  11. The label according to claim 9, characterized in that the concentric rings are placed in the center of the label.  12. The label according to claim 9, characterized in that the optical properties characterize black, white and gray colors.

Images