KR101293703B1 - A system for decoding skewed data matrix barcode, and the method therefor - Google Patents

Abstract

The present invention relates to a data matrix barcode decoding technology, and more particularly, even when a barcode image input by a camera or the like is skewed due to the positional relationship of the cameras, an error is obtained by reading the barcode without changing the shape of the extracted image. A skewed data matrix barcode decoding apparatus and a method therefor for excluding the above are provided.
According to the skewed data matrix barcode decoding apparatus of the present invention and a method therefor, a skew for reading a data matrix even when an axial nonuniformity of the data matrix barcode image or a grid nonuniformity occurs There is a technical effect of providing a data matrix barcode decoding device and a method therefor.
Also, the present invention provides a skewed data matrix barcode which enables a data matrix to be read out even in an image including skew generated by the positional relationship between the camera and the barcode when the barcode image is input by an image means such as a camera. It is effective to provide an extraction and decoding device and a method therefor.
The present invention has a technical effect of providing a skewed data matrix barcode decoding apparatus and a method therefor for minimizing the loss of an image of an original data matrix to read a data matrix that directly reads data from an image of the data matrix.

Description

Device for decoding skewed data matrix barcode and method therefor {A system for decoding skewed data matrix barcode, and the method therefor}

The present invention relates to a data matrix barcode decoding technology, and more particularly, even when a barcode image input by a camera or the like is skewed due to the positional relationship of the cameras, an error is obtained by reading the barcode without changing the shape of the extracted image. A skewed data matrix barcode decoding apparatus and a method therefor for excluding the above are provided.

Among the various types of two-dimensional barcodes, the data matrix barcode is an ISO / IEC standard and is widely used in various applications. The related standards are as follows.

ISO / IEC 16022: 2006 DataMatrix bar code symbology specification

ISO / IEC 154152-D Print Quality Standard

ISO / IEC 15418: 2009 Symbol Data Format Semantics (GS Application Identifiers and ASCMH10 Data Identifiers and maintenance)

ISO / IEC 15424: 2008 Data Carrier Identifiers (including Symbology Identifiers) [IDs fordistinguishing different bar code types]

ISO / IEC 15434: 2009 Syntax for high-capacity ADC media (format of data transferred from scanner to software, etc.)

ISO / IEC 15459 Unique Identifiers

The standards define the printing and reading of the data matrix, and in particular, in the case of ISO / IEC 16022 and ISO / IEC 15415, the data matrix is inclined within a maximum of 7 degrees. Therefore, the data matrix image inputted by the portable terminal or the camera is defined as a category which cannot be read when inputted based on an arbitrary angle. To improve this, the user needs to identify the reference position of the data matrix and input the image. In addition, it is defined as unreadable if there is a non-uniform distortion (skew) of the input barcode image.

As a background art of the present invention, there is a method for detecting a two-dimensional bar code of the Republic of Korea Patent Publication No. 10-2006-0123999. This technology is proposed to solve the problem that it took a lot of time to detect the barcode of the paper by scanning the entire image of the paper by scanning the paper by inserting the paper with the barcode inserted in the unmanned bill acceptor. By analyzing the image of the paper with the scanned paper (Texture Analysis), it is possible to extract only the bar code areas where black and white are densely arranged, thereby reducing the detection time of the bar code area located on the paper. It's about how. This technique includes a first step of scanning paper when paper is loaded, and finding a boundary of the scanned image; Filtering only the high frequency portion; A third step of blurring the image through mean value filtering; A fourth step of performing a threshold holding by comparing a contrast value of each pixel of the filtered image with a reference value, wherein a value larger than the reference value is a white pixel and a small value is a black pixel; A fifth step of expanding the white area by leaving all of the pixels adjacent to white pixels of the image in white, and reducing the white area by leaving all pixels adjacent to the black pixels in black; A sixth step after the operation, finding an area having the largest size calculated by the number of connected white pixels among the remaining white areas; And a seventh step of extracting coordinate values of the area and cutting out an area corresponding to the coordinate value from the image of the paper.

This technology is limited to the technology of extracting only the barcode area of the image of the paper on the scanned paper. When the barcode image is input by an imaging means such as a camera, skew generated by the positional relationship between the camera and the barcode is included. There is a disadvantage that cannot be read from an image. In addition, by comparing the contrast value of each pixel of the filtered image with the reference value, a threshold holding is performed in which a value larger than the reference value is set to a white pixel and a smaller value is set to a black pixel. If the error occurs, an error occurs in the image as a result of the threshold holding due to the fixed reference value.

Another background art of the present invention is a method of reading a two-dimensional bar code of the Republic of Korea Patent Publication No. 10-0275007. This technique photographs a two-dimensional barcode printed on a piece of paper, etc. using an area sensor, and performs distortion correction by projection projection on an image of the two-dimensional barcode obtained from such an area sensor, and brightness irregularities appearing in the image. A method of reading a two-dimensional bar code includes a step of performing correction and a step of decoding a two-dimensional bar code based on the image data of the two-dimensional bar code corrected by the two steps.

Since this technique generates pixels that do not exist in the original image by performing projective transformation on an image of a two-dimensional barcode, noise is added to the original image by the generated pixels, and thus there is a limit that an error in reading may increase. In particular, if there is a difference in contrast due to a shadow or a small image in the single image depending on the camera image input environment, there is no basis and means for determining this as an error. There are disadvantages.

Another background art of the present invention is a two-dimensional bar code reading device technology of the Republic of Korea Patent Publication No. 10-0483393. This technique includes an imaging device for photographing two-dimensional bar codes; Frame memory means for storing an image of a two-dimensional bar code photographed by the imaging device; Obtain the four corner coordinates of the two-dimensional barcode image stored in the frame memory means, obtain four linear equations passing through the four corners, and find the two parts corresponding to the start code and the stop code of the two-dimensional barcode image from the four straight lines. A plurality of scan lines for scanning a two-dimensional bar code image are set at equal intervals between a pair of straight lines crossing all of them, and a plurality of scan lines between the remaining pair of straight lines extending along each of the two portions. By setting a plurality of sampling lines for sampling a plurality of two-dimensional barcode image along each of the at equal intervals, and sampling the two-dimensional barcode image at a plurality of positions where each of the plurality of scan lines intersects the plurality of sampling lines. And a bar code reading device equipped with a bar code reading control means for reading bar code information of the 2D bar code in the 2D bar code image. ,

A bar code reading device for photographing a two-dimensional bar code by an image pickup device and reading bar code information from the captured bar code image includes a slope between four straight lines passing through the four corners of the bar code image at the four corner coordinates of the bar code image. A pair of the four straight lines extending along each straight line a scan line and a sampling line for reading the barcode image in response to the detected distortion of the barcode image by detecting the distortion of the barcode image. The bar code reading device for reading out is arrange | positioned evenly between the straight lines of the summary, The summary is made.

This technique reads the scan line and sampling line evenly between the pair of straight lines among the four straight lines extending along each straight line in response to the skew of the barcode image, resulting in uneven skew in the camera input image. Even if this happens, evenly spaced scan lines and sampling lines have the constraint of increasing reading errors.

Another background of the present invention is a two-dimensional barcode reading apparatus and method technology having a pattern slope calculation function of Korean Patent Publication No. 10-0829108. This technology reads two-dimensional barcodes with a pattern slope calculation function that speeds up the reading of two-dimensional barcodes and minimizes reading errors of barcode patterns by calculating pattern slopes and decoding using calculated slope values. An apparatus and method, the configuration comprising: a CMOS image sensor for capturing a two-dimensional barcode pattern to output image data; and a data storage unit for storing the output image data; controlling and calculating a slope of the captured barcode pattern; A microcontroller for decoding a bar code using a pattern slope value to control the decoded data to be output; A BUS selector for storing the image data in the data storage unit, or performing a bus selection control when outputting the stored image data ; Sequentially search for line pattern by line based on the set value Bar code pattern search unit for determining whether the captured image is a bar code image; If it is determined that the bar code image, it characterized in that it comprises a bar code pattern gradient calculation unit for calculating the slope of the bar code pattern under the control of the microcontroller.

Since this technique calculates a single slope of a barcode pattern and applies it to the reading of the barcode, the barcode to be read is limited to the PDF417 code, which has a limitation that cannot be applied to a distorted data matrix barcode according to a general skew.

KR 10-1055853 B1 KR 10-2011-0042557 A KR 10-0944272 B1 WO 2006078359 A1 US 7497380 B2

 ISO / IEC 16022: 2006- Data Matrix bar code symbology specification  ISO / IEC 15415- 2D Print Quality Standard  ISO / IEC 15418: 2009- Symbol Data Format Semantics  ISO / IEC 15424: 2008- Data Carrier Identifiers  ISO / IEC 15434: 2009- Syntax for high-capacity ADC media  ISO / IEC 15459- Unique Identifiers  GS1 DataMatrix An introduction and technical overview of the most advanced GS1 Application Identifiers compliant symbology

The present invention provides a skewed data matrix barcode decoding device and a method therefor for reading a data matrix from an image even if axial nonuniformity or grid nonuniformity occurs. do.

Also, the present invention provides a skewed data matrix barcode which enables a data matrix to be read out even in an image including skew generated by the positional relationship between the camera and the barcode when the barcode image is input by an image means such as a camera. It is a technical problem to provide an extraction and decoding apparatus and a method therefor.

It is an object of the present invention to provide a skewed data matrix barcode decoding device and a method therefor for minimizing the loss of an image of an original data matrix to read a data matrix that directly reads data from an image of the data matrix. .

In order to solve the above problems, an apparatus for receiving and decoding a skewed data matrix barcode from a camera for capturing an image, comprising: a first frame memory for storing color images and data from the camera and the camera; A terminal comprising a display for outputting image data of the first frame memory, a control unit, a 2D matrix reading unit, a DMA controller, a second frame memory, an operation frame memory, a gray level converter, a dual linear interpolation unit, an area filter unit A skewed data matrix barcode decoding device including a binary image converter, a connection area extractor, a peripheral gradient extractor, a finder pattern detector / rotator, a symbol calculator, and a 2D matrix memory and a method thereof are provided as technical solutions.

According to the skewed data matrix barcode decoding apparatus of the present invention and a method therefor, a skew for reading a data matrix even when an axial nonuniformity of the data matrix barcode image or a grid nonuniformity occurs There is a technical effect of providing a data matrix barcode decoding device and a method therefor.

Also, the present invention provides a skewed data matrix barcode which enables a data matrix to be read out even in an image including skew generated by the positional relationship between the camera and the barcode when the barcode image is input by an image means such as a camera. It is effective to provide an extraction and decoding device and a method therefor.

The present invention has a technical effect of providing a skewed data matrix barcode decoding apparatus and a method therefor for minimizing the loss of an image of an original data matrix to read a data matrix that directly reads data from an image of the data matrix.

Figure 1 is an example of various two-dimensional barcodes
Figure 2 shows the construction of a 2 x 2 data matrix
Figure 3 shows the basic pattern of a 2 x 2 data matrix
4 is an error that occurs when printing or scanning a data matrix.
5 is a blurring phenomenon that occurs when correcting an image of a skewed data matrix.
6 is a block diagram of the skewed two-dimensional barcode decoding apparatus of the present invention
7 is a diagram in which an image of a second frame memory is converted to gray levels and stored.
8 is a diagram illustrating an example in which image data of a second frame memory is reduced and stored in an operation frame memory.
9 is a process of processing gray level image data stored in an operation frame memory.
10 is an example of edge value binarization image data processing of reduced image data;
11 is a configuration for extracting a candidate region of a barcode
12 is a configuration for obtaining the length and the slope of the candidate region of the barcode
13 is a configuration for extracting an image of a second frame memory corresponding to a barcode candidate region of an operation frame memory.
14 is an example of converting a barcode candidate region image of a second frame memory.
15 is a configuration for obtaining a slope including all the symbols placed in the right vertical direction with respect to the binary barcode image
16 is an embodiment of a flowchart for obtaining a vertical center symbol gradient initial value Cv.
FIG. 17 is a general flowchart for obtaining the vertical center symbol gradient initial value Cv. FIG.
18 is a configuration for obtaining coordinates of symbols in a vertical center direction from a barcode image
19 is a configuration for obtaining a slope including all the symbols placed in the upper horizontal direction with respect to the binary barcode image
20 is an embodiment of a flowchart for calculating a horizontal center symbol gradient initial value Ch
21 is a general flow chart for calculating the horizontal center symbol gradient initial value Ch.
22 is a configuration for obtaining coordinates of symbols in a horizontal center direction from a barcode image
23 shows the arrangement relationship between the linear equations in the horizontal direction and the linear equations in the vertical direction of the barcode.
24 is a cross relationship between a single horizontal linear equation and a vertical linear equation among the straight lines connecting the symbols in the horizontal-vertical direction.
25 is a transformation relationship for coordinates of a second frame memory that intersects in the horizontal-vertical direction.
26 shows an example of generating a 2D barcode in a 2D matrix memory.
27 is a configuration relationship for decoding data in a 2D matrix memory through a 2D matrix reading unit.
28 through 30 are flow charts of the skewed data matrix barcode decoding method of the present invention.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the following description of the present invention, when it is determined that a detailed description of a related known function or configuration may unnecessarily obscure the subject matter of the present invention, the detailed description thereof will be omitted.

1 shows examples of various two-dimensional barcodes. Since the two-dimensional bar code can record a larger amount of data than the conventional one-dimensional bar code, and thus various applications are possible, its use is increasing. The two-dimensional barcode has a QR code having three reference figures for reading, an Aztec code and a Maxi code having a reference reference figure in the center, and a rotational reading path around the center. Shotcode, data matrix with reading bar around the code, EZ code, PDF417 code consisting of two-dimensional bar code (Codablockcode), Code 16K code, and various other modification codes have been proposed and used.

Among these barcodes, the data matrix is a bar code formed by arranging black and white squares in two dimensions. The normal data size is 2Kbyte, and one data matrix can record 2,335 numeric and alphanumeric characters. Can be. The symbol size of the data matrix may range from 10 X 10 to 144 X 144, and a data matrix obtained by combining a plurality of data matrixes from a single data matrix to a plurality of columns horizontally and vertically is commercialized.

2 shows the configuration of a 2 × 2 data matrix to be recognized in the present invention. The 2 X 2 data matrix 101 of the drawing is a combination of two single data matrices in a vertical and horizontal direction. The left and right of the data matrix can be distinguished from the left pinter pattern 103 and the lower side of the data matrix. The pinter pattern 104 is located.

3 shows the basic pattern of a 2 × 2 data matrix. The 2 × 2 data matrix 101 of the figure is a configuration in which two single data matrices are combined horizontally and vertically, and each single data matrix has a finder pattern at the leftmost and bottommost. In addition, each single data matrix has a pattern indicating the size of the basic symbol for reading the barcode pattern at the rightmost and the topmost so as to be divided into '0' and '1'. clock track). As shown in FIG. 7, the 2 X 2 data matrix 101 which combines the single data matrix 2 horizontally and vertically is the left and right sides of the data matrix, which can distinguish the beginning of each bar code. The finder pattern 103 and the lower finder pattern 104 are located, and the clock track 102 is arranged at the rightmost and the topmost to indicate the size of the basic symbol for reading the barcode pattern divided into '0' and '1'. do. In addition, a finder pattern and a clock track, which can distinguish a single data matrix, are positioned at the center of the horizontal light longitudinal direction of the 2 × 2 data matrix 101.

Figure 4 shows a typical error that occurs when printing or scanning a data matrix. First, Figure 4 (a) shows a barcode in a steady state, while the diagram in (b) shows a data matrix in which the contrast is incorrectly adjusted. Occurs when the brightness and contrast settings or the scanner's brightness level settings are incorrect at the time of reading.

Figures (c) and (d) show errors due to the spreading of the ink or the printing area according to the print target lipid in the printer printing the data matrix. The error by figure (c) is shown. FIG. 4E shows the loss of the pattern generated when the resolution of the printer or the resolution of the data matrix is set too low or the feeding speed is uneven. Figure (f) shows an axial nonuniformity data matrix that occurs when the printer's feeding speed is incorrect or when the data matrix and the printer's pixel size do not match. Figure (g) shows the grid nonuniformity of the data matrix due to the printer feeding slip or uneven feeding speed. The error shown in FIG. 4 is described as a case of printing or scanning a data matrix. However, an error occurs even when the data matrix is read by a camera attached to a fixed camera or a portable terminal.

The bleeding of the data matrix and the loss of the pattern shown in (e) shown in FIGS. 4C and 4D can be overcome by increasing the resolution of the camera and the auto focus function. . However, the error due to contrast shown in (b) may not only occur depending on the brightness of the surroundings, even if the camera has an autoexposure / auto iris function. As a result, unevenness in brightness occurs within a single data matrix image. In addition, the axial nonuniformity phenomenon in (f) and the grid nonuniformity phenomenon in (g) are caused by a camera lens that is not parallel with the mobile terminal and the plane where the data matrix is located or is too small. This is a phenomenon that can always occur. The present invention provides an apparatus and method for enabling the reading of a data matrix from an image even when the above axial nonuniformity and grid nonuniformity occur.

5 illustrates a blurring phenomenon that occurs when correcting an image of a distorted data matrix. In order to correct an image of the data matrix distorted by a camera skew or the like into a rectangular data matrix in a rectangular shape at right angles, the image of the data matrix is stretched or reduced for each pixel. However, in the above correction process, a pixel that does not exist in the image of the original data matrix is newly created or existing pixels are reduced to a smaller number of pixels than the original number of pixels. There is a limit. FIG. 5 shows that the image of the 2 × 2 data matrix 102 caused by the correction is lost and the image is generally blurred, and the data read therefrom is inevitably increased in error. Accordingly, the present invention provides an apparatus and method for directly reading data from an image of a data matrix by minimizing loss of an original data matrix image.

6 shows the configuration of the skewed two-dimensional barcode decoding apparatus of the present invention. The two-dimensional barcode extracting apparatus included in the image of the present invention includes a camera 110 photographing the barcode 100 and a first frame memory 120 and the first frame memory storing color image data from the camera 110. Based on the terminal consisting of a display 130 for outputting the image data of (120). The first frame memory 120 is for storing a general color image, and may be configured to store data for each pixel when a color image such as an RGB method or a YUV method is provided as data from the camera 110. . Although omitted in the drawings, the terminal includes an A / D converter between the camera 110 and the first frame memory 120 and a D / A converter between the first frame memory 120 and the display 130. And a separate processor for controlling the camera 110, the A / D converter, the first frame memory 120, the D / A converter, and the display 130.

Hereinafter, functions and operations of the components of FIG. 6 will be described.

-DMA controller 140

The direct memory access (DMA) controller 140 transmits the image data from the camera 110 stored in the first frame memory 120 to the second frame memory 150 to be described later by a transfer command of the controller 160 to be described later. Save it.

Second frame memory 150

The second frame memory 150 is used to store image data of the first frame memory 120 and has the same pixel resolution as that of the first frame memory 120 described above, and the data of the second frame memory 150. Finally read with a barcode.

Control unit 160

Data movement between the frame memories and individual processing units are controlled and the data matrix barcode is finally read.

Matrix reader 160a

The barcode reading result in the second frame memory 150 finally reads the 2D matrix barcode of the 2D matrix memory 260 stored as '0' and '1'.

Calculated frame memory 170

The arithmetic frame memory 170 is configured to perform arithmetic processing while preserving data of the second frame memory 150 as a storage device for arithmetic processing the data of the second frame memory 150.

Gray level converter 180

The image from the camera 110 stored in the second frame memory 150 is converted to a gray level.

Dual linear interpolator 190

The image data of the second frame memory 150 converted to gray level is reduced and stored in the operation frame memory 170. The reduction of the image is performed at the ratio of integer multiples.

Area filter unit 200

Allocate a memory of 3 X 3 pixel size to be used as a mask for the gray level image in the operation frame memory 170, and read the pixel values of 3 horizontally and vertically based on the position of the current pixel to be calculated. Find the minimum value and compute the difference as the edge value.

Binary image converter 210

Gray level image data of the operation frame memory 170 or the second frame memory 150 is converted into binary images of '1' and '0'.

-Connection area extraction unit 220

The connection area extractor 220 determines adjacent 'white' pixel groups for the binary image in the operation frame memory 170, assigns a unique number to each group, and performs labeling. Extract the 'white' pixel group, which consists of the largest number of pixels among the distinct 'white' pixel groups. The connection region extracting unit 220 extracts the candidate region of the barcode by substituting 'black' for the pixel values of the remaining groups except for the 'white' pixel group composed of the extracted most pixels. When labeling adjacent 'white' pixel groups, 'white' pixels facing in the horizontal or vertical direction with respect to one 'white' pixel are regarded as the same group, and there are 'black' pixels in the horizontal or vertical direction. If there are 'white' pixels in the diagonal direction, it is determined that they are not a group.

Peripheral slope extraction unit 230

The peripheral gradient extractor 230 'black' the pixel values of the remaining groups except the 'white' pixel group, which is composed of the largest number of pixels for the binary image in the operation frame memory 170. Calculate the gradient (vertical gradient: Gradient Vert) between the boundary coordinates of the upper half and the lower boundary of the center vertically. Calculate the gradient (horizontal gradient Gradient Horz) between the 1/2 right boundary coordinate and the 1/2 left boundary coordinate around the center horizontally. In addition, in the barcode image in the second frame memory 150 processed by the pinter pattern detection / rotation unit 240, a slope including all the symbols placed at the right vertical position, and all the symbols placed at the center vertical position of the barcode are included. The slope, the slope including all the symbols placed at the top horizontal position of the barcode, and the slope including all the symbols placed at the horizontal center position of the barcode are calculated.

-Finder pattern detection / rotation unit 240

The finder pattern and its position are detected from the barcode image in the second frame memory 150, and based on this, the control unit rewrites the barcode image in the second frame memory 150 into image data obtained by rotating the barcode image by 90 degrees. do. In the barcode image processed by the pinter pattern detection / rotation unit 240, the left finder pattern 103 and the lower finder pattern 104 are located at the left and the bottom of the image, respectively, in the second frame memory 150. And a clock track 102 for displaying the size of the basic symbol for reading the barcode pattern at the top and separated by '0' and '1'.

Symbol calculation unit 250

The symbol operation unit 250 may be configured with respect to the positions of the symbols searched in the barcode image in the second frame memory 150 processed by the pinter pattern detection / rotation unit 240; When you find the right vertical slope, find the array where the retrieved symbol position information is stored, and when you find the center vertical slope, find the linear equation that consists of two points stored at the same index of the array where the retrieved symbol position information is stored. When calculating the array and the center horizontal gradient, obtain the linear equation of the two points stored in the same index of the array where the retrieved symbol position information is. If the pixel value corresponding to the intersection point of the two straight lines is black, '1', white '0' is stored in the 2D matrix memory 260.

Matrix Memory (260)

The 2D matrix memory 260 is a memory device which stores the barcode reading result in the second frame memory 150 as '0' and '1' by the symbol operation unit 250. The 2D matrix barcode of the 2D matrix memory 260 is read by 160a).

Accordingly, the skewed data matrix barcode decoding apparatus of the present invention;

The camera 110 photographing the barcode 100, the first frame memory 120 storing the color image data from the camera 110, and the display 130 outputting the image data of the first frame memory 120. A terminal consisting of:

A controller 160 for controlling data movement between frame memories and individual processing units and finally reading a data matrix barcode; and

A second 2D matrix reading unit for finally reading the 2D matrix barcode of the 2D matrix memory 260 in which the barcode reading result in the second frame memory 150 is stored as '0' and '1'. 160a;

The control unit 160,

A DMA controller 140 for storing image data from the camera 110 stored in the first frame memory 120 described by the transfer command of the controller 160 in the second frame memory 150;

A second frame memory 150 for storing image data of the first frame memory 120;

An arithmetic frame memory 170 for arithmetic processing data of the second frame memory 150;

A gray level converter 180 for converting an image from the camera 110 stored in the second frame memory 150 into a gray level;

A dual linear interpolation unit 190 for performing an operation for reducing and storing the image data of the second frame memory 150 converted to gray level in the operation frame memory 170;

An area filter unit 200 for calculating an edge value with respect to the gray level image in the calculation frame memory 170;

A binary image converter 210 for converting gray level image data of the operation frame memory 170 or the second frame memory 150 into binary images of '1' and '0';

Determining adjacent 'white' pixel groups for the binary image in the operation frame memory 170 and labeling by assigning a unique number for each group, 'white' pixel groups divided by the labeling (labeling) A connection region extraction unit 220 for extracting a 'white' pixel group composed of the most pixels among the pixels;

The candidate region of the barcode extracted by replacing the pixel values of the remaining groups except for the 'white' pixel group composed of the most pixels for the binary image in the operation frame memory 170 with 'black' Find the location of the maximum area for and calculate the slope (vertical gradient: Gradient Vert) between the vertically centered half-boundary coordinates and the half-bottomed boundary coordinates, and horizontally centered 1 Calculate the slope (horizontal slope Gradient Horz) between / 2 right boundary coordinate and 1/2 left boundary coordinate,

A slope including all the symbols placed at the right vertical position in the barcode image in the second frame memory 150 processed by the pinter pattern detection / rotation unit 240, a slope including all the symbols placed at the center vertical position of the barcode, A peripheral gradient extracting unit 230 for calculating a slope including all of the symbols placed at the top horizontal position of the barcode and a slope including all the symbols placed at the horizontal center position of the barcode;

The finder pattern and its position are detected from the barcode image in the second frame memory 150, and the controller rewrites the barcode image in the second frame memory 150 into image data rotated by 90 degrees. In the second frame memory 150, the left finder pattern 103 and the lower finder pattern 104 are located at the left and the bottom of the image, respectively, and the size of the basic symbol for reading the barcode pattern at the rightmost and the topmost is' A pinter pattern detection / rotation unit 240 for calculating the clock track 102 to be divided into 0 'and' 1 ';

When the right vertical slope is obtained with respect to the position of the symbols searched in the barcode image in the second frame memory 150 processed by the pinter pattern detection / rotation unit 240, the array in which the retrieved symbol position information is stored and the center vertical slope are obtained. When you find the linear equation that consists of two points stored in the same index of the array where the retrieved symbol position information is stored, and when you obtain the top horizontal slope, the array where the retrieved symbol position information is stored and the center where the retrieved symbol position information is stored A symbol operation unit 250 for obtaining a linear equation formed by two points stored in an index, and storing '1' if the pixel value corresponding to the intersection point of the two straight lines is black and '0' if it is white in the 2D matrix memory 260;

A 2D matrix memory 260 which stores the barcode reading result in the second frame memory 150 as '0' and '1' by the symbol operator 250;

Characterized in that is configured to be connected.

In addition, the two-dimensional bar code extraction apparatus included in the image of the present invention, the first frame memory 120 and the first frame memory 120 for storing the barcode image 100 and the color image data from the camera 110 For the two-dimensional barcode image included in the image provided from the terminal consisting of the display 130 for outputting the image data of the frame memory 120,

a.When the recorded image does not contain a barcode

b. If the recorded image contains barcode-like image

c.When the barcode is shot too small

d.Low resolution due to poor conditions of the device or for processing efficiency

The present invention provides a two-dimensional bar code extraction apparatus included in an image for extracting a two-dimensional bar code by performing a readable process on the image photographed. To this end, the two-dimensional bar code extraction apparatus included in the image of the present invention in the order of the specific working relationship with respect to the configuration of FIG.

1) (direct memory access) The controller 140 stores the image data from the camera 110 stored in the first frame memory 120 in the second frame memory 150 by a transfer command of the controller 160.

2) The controller 160 converts the color image of the second frame memory 150 to the gray level through the gray level converter 180.

FIG. 7 illustrates a configuration in which the color image of the second frame memory 150 is converted to gray levels by the gray level converter 180 and stored.

The image from the camera 110 is stored in the first frame memory 120, stored in the second frame memory 150 by the DMA controller 140, and stored in the second frame memory 150. green-blue) Converts color images to gray levels.

3) Next, the controller 160 reduces and stores the image data of the second frame memory 150 converted to gray level by the double linear interpolator 190 at the ratio of integer multiples in the operation frame memory 170. 8 illustrates a configuration in which image data of the second frame memory 150 is reduced and stored in the operation frame memory 170.

In the present invention, the image data of the second frame memory 150 is reduced and stored in the operation frame memory 170 at a reduction ratio of 1/4.

4) FIG. 9 illustrates a series of processes for processing gray level image data stored in the operation frame memory 170 by the above process.

The controller 160 calculates an edge value of the image by providing the reduced image data stored in the operation frame memory 170 to the area filter unit 200. The area filter unit 200 allocates a 3 × 3 pixel size memory to be used as a mask for the gray level image in the operation frame memory 170 and arranges three horizontally and vertically around the position of the current pixel to be processed. Read the pixel value of to find the maximum and minimum values and compute the difference as the edge value. When the edge value operation is completed for the entire reduced image data, the image composed of the edge value is stored in the operation frame memory 170.

5) The controller 160 provides the edge value image of the reduced image data stored in the operation frame memory 170 to the binary image converter 210 to convert the image into a binary image.

6) The controller 160 provides the edge region binarized image data of the reduced image data stored in the operation frame memory 170 to the connection region extractor 220 to determine adjacent 'white' pixel groups in the image. Labeling is performed by assigning a unique number to each group, and a 'white' pixel group consisting of the largest number of pixels among the 'white' pixel groups classified by the labeling is extracted.

FIG. 10 is a partial image as an example of edge value binarization image data processing of reduced image data, wherein adjacent 'white' pixel groups in the image are determined for edge value binarization image data of the reduced image data, and a unique number is obtained for each group. As a result of performing the labeling by attaching, an example of the image is shown. The adjacent 'white' pixel groups shown in FIG. 10 are labeled '10' from Group 1 labeled as '1'. It is shown that a 'white' pixel group up to group 10 and consisting of the most pixels can be extracted into group '9'.

Subsequently, the connection area extractor 220 replaces the group consisting of the largest number of extracted pixels with 'white', and replaces the pixel values of the remaining groups except the pixel group with 'black' to candidate for the barcode of 'white'. Extract the area.

7) the controller 160 provides the image data of the operation frame memory 170 including the candidate region of the extracted 'white' barcode to the peripheral gradient extracting unit 230, so that the operation frame memory 170 Calculate the upper and lower boundary coordinates, the vertical height and the slope (vertical gradient: Gradient_Vert) about the center in the vertical direction of the image data, and the 1/2 right boundary coordinates and 1 about the center in the horizontal direction. / 2 Calculate left boundary coordinate, horizontal length and slope (horizontal slope: Gradient_Horz).

FIG. 11 illustrates a configuration of extracting candidate regions of a barcode by providing image data of the operation frame memory 170 to the peripheral gradient extracting unit 230. The barcode candidate region is treated as 'white' and the remaining portion is treated as 'black' by the above-described labeling process. However, in the drawing, portions other than the barcode candidate region are shown as 'white' for clarity of representation. The peripheral gradient extractor 230 obtains (Xa, Ya) and (Xb, Yb), which are boundary coordinates spaced from the center in the vertical direction (Y-axis) of the image data of the operation frame memory 170, and connects them. Find the equation of a straight line.

In addition, we obtain (Xc, Yc) and (Xd, Yd), which are the boundary coordinates spaced from the center of the bottom, and find the straight line equation connecting them. In the horizontal direction (x-axis direction), the boundary coordinates (Xb, Yb) and (Xc, Yc) spaced from the center of the right side are obtained, and the equation of a straight line connecting them is obtained. In addition, the equations of a straight line connecting (Xa, Ya) and (Xd, Yd), which are spaced coordinates centered on the left center, are obtained. Using the extracted four linear equations, the peripheral length and the slope of the candidate region of the barcode can be calculated, respectively. In an embodiment, when the size of the second frame memory 150 is 640 X 480 and the size of the operation frame memory 170 is 160 X 120, the separation distance is set to 5 to 10 pixels to increase the vertical height and the slope. The transverse length and the slope can be calculated.

FIG. 12 illustrates a configuration for obtaining the length and the slope of the candidate region of the barcode by using the candidate region of the barcode extracted by the peripheral slope extracting unit 230 from the image data of the operation frame memory 170.

8) Next, the control unit 160 is horizontal with the vertical height Labeling_height and the horizontal length Labeling_width of the barcode candidate area provided by the connection area extraction unit 220, and the vertical gradient Gradient_Vert calculated by the peripheral slope extraction unit 230. The gradient Gradient_Horz is used to determine if the barcode reading is a valid image. The above determination is performed in two ways. If the ratio of the vertical height Labeling_height and the horizontal length Labeling_width of the barcode candidate area stored in the first frame memory 120 is within a predetermined ratio, it is determined that the reading is possible and the difference is excessive. In this case, the new image data from the camera 110 is stored in the second frame memory 150 from the first frame memory 120 by the transmission command of the controller 160, and the processing is performed again from the beginning. do. The following expression shows an example in which a bar code reading process is performed when the ratio of the vertical height Labeling_height and the horizontal length Labeling_width is within a predetermined value, and otherwise, an error (recognition failure) is processed.

intbig, small;

big = Labeling_height;

small = Labeling_width;

If (Labeling_height <Labeling_width)

{

big = Labeling_width;

small = Labeling_height;

}

if (big / 3 * 2> small)

error;

The second judgment is

The vertical gradient Gradient_Vert and horizontal gradient Gradient_Vert values calculated by the peripheral gradient extractor 230 are close to vertical, the horizontal gradient Gradient_Vorz value is close to horizontal, or the vertical gradient Gradient_Vert and horizontal gradient Gradient_Horz values When it is close to each other, that is, when the product of the vertical gradient Gradient_Vert and the horizontal gradient Gradient_Horz is less than 0, it is determined that the reading can be performed. The new image data from) is stored in the second frame memory 150 from the first frame memory 120 and processed again from the beginning. In the above determination, the vertical gradient Gradient_Vert and the horizontal gradient Gradient_Horz values may be obtained by using boundary coordinates of candidate regions of the barcode extracted by the peripheral gradient extractor 230 from the image data of the operation frame memory 170. The vertical gradient Gradient_Vert is preferably calculated by the inverse and the same condition as the Gradient_Horz value. For example, referring to FIG. 11, the drawing is as follows. In FIG. 11, using the upper left boundary coordinates (Xa, Ya) and (Xb, Yb) and the lower boundary coordinates (Xc, Yc) and (Xd, Yd),

Absolute value [Gradient_Vert] = | (Xa-Xd) / (Ya-Yd) |

Absolute value [Gradient_Horz] = | (Yd-Yc) / (Xd-Xc) | .

At this time, the absolute value [Gradient_Vert] has a small value close to 0 as the vertical value is close, and the absolute value [Gradient_Horz] is close to the horizontal value.

Therefore, when the vertical gradient Gradient_Vert is close to vertical, the horizontal gradient Gradient_Horz is close to horizontal, or the vertical gradient Gradient_Vert and the horizontal gradient Gradient_Horz are close to vertical, i.e., the product of the vertical gradient Gradient_Vert and the horizontal gradient Gradient_Horz is less than 0. In this case, the condition for determining that the reading is possible may be any of the following.

(1) When absolute value [Gradient_Vert] <r,

(2) When absolute value [Gradient_Hor] <r

(3) (Gradient_Vert * Gradient_Horz) <0

As an example, when the horizontal and vertical resolution of the image data of the second frame memory 150 is 640 × 480 and the image data of the operation frame memory 170 is 160 × 120,

(1) When absolute value [Gradient_Vert] <0.000001,

(2) When absolute value [Gradient_Hor] <0.000001

(3) (Gradient_Vert * Gradient_Horz) <0

It is judged that the reading is possible.

9) The control unit 160 measures the vertical height Labeling_height and the horizontal length Labeling_width of the barcode candidate area provided by the connection area extraction unit 220, and the vertical gradient Gradient_Vert and the horizontal gradient Gradient_Horz calculated by the peripheral gradient extraction unit 230. If it is determined that the barcode can be read by using the control unit, the controller 160 selects a pixel value of an area excluding the barcode candidate region from an image of the second frame memory 150 corresponding to the barcode candidate region of the operation frame memory 170. Replace with 'white'.

As described above, the image data of the second frame memory 150 is reduced and stored in the operation frame memory 170 by the dual linear interpolation unit 190 by the control unit 160. An area of the corresponding second frame memory 150 may be extracted by expanding the relative address of the barcode candidate area by a reduced ratio.

FIG. 13 illustrates a configuration of extracting an image of the second frame memory 150 corresponding to the barcode candidate region of the operation frame memory 170.

10) only the portion corresponding to the barcode candidate region remains in the second frame memory 150 and the remaining space is converted into an image filled with 'white'.

FIG. 14 illustrates an image conversion example in which only a portion corresponding to a barcode candidate region remains in the second frame memory 150 and the remaining space is filled with 'white'.

At this time, if the size of the barcode candidate region of the second frame memory 150 is read to a size smaller than or equal to a predetermined size, the controller 160 considers the recognition failure to be performed. The image data is stored in the second frame memory 150 from the first frame memory 120 and processed again from the beginning. In the present invention, for example, when the number of pixels in the barcode candidate region is equal to or less than the total number of symbols in the second frame memory 150, the recognition failure is regarded, and new image data from the camera 110 is provided by the transmission command of the controller 160. The first frame memory 120 is stored in the second frame memory 150 and the process is performed again from the beginning.

11) When the binary barcode image of the second frame memory 150 is prepared, the controller 160 provides the binary barcode image of the second frame memory 150 to the symbol operator 250 to extract the coordinates of the symbol. . First, the symbol operator 250 obtains a slope including all the symbols placed in the right vertical direction of the barcode.

FIG. 15 illustrates a configuration for obtaining a slope including all the symbols placed in the vertical right direction with respect to the binary barcode image of the second frame memory 150. The slope that includes all the symbols placed in the right vertical direction of the barcode is based on the end coordinates (vert_right (x, y)) 152 of the lower right side of the binary barcode image of the second frame memory 150 including the image. Coordinates of the upper right corner (vert_top (x, y)) in the origin direction of the second frame memory 150 starting with a straight line of the slope formed with the coordinates of the upper right corner (vert_top (x, y)) 151. The gradient including the maximum number of symbols is obtained by moving 151 by one pixel in the origin direction.

In this step, the symbol size of the barcode is determined, and the time point at which the symbol size is determined is the upper right endpoint from the slope with the coordinate (vert_top (x, y)) 151 of the upper right corner of the second frame memory 150. When the number of symbols retrieved on the straight line of the slope calculated while moving to the left decreases again from the even maximum value, it stops, and the slope where the maximum number of symbols is detected, the number of symbols and the coordinates of the symbols are stored in the array.

12) The controller calculates a slope including all the symbols placed at the center vertical position of the barcode through the symbol calculator 250. The vertical center symbol inclination initial value Cv is determined according to five cases based on the inclination value of all the symbols placed in the right vertical direction of the barcode and the inclination value of the left finder pattern 103. At this time, each slope is calculated by converting the length in the x-axis direction with respect to the length in the y-axis in consideration of the significant number of binary operations. As a method of determination, suppose that the vertical center symbol slope initial value is Cv, and the slope including all the symbols placed in the right vertical direction of the barcode is R1 and the slope value of the left finder pattern 103 is L,

When the absolute magnitude of -L is greater than the absolute magnitude of R1 and L is close to vertical; Cv = L

The absolute magnitude of L is greater than the absolute magnitude of R1, L is not vertical and the absolute magnitude of L is within a certain amount relative to the absolute magnitude of R1; Cv = L * 0.8 + R1 * 0.2

The absolute magnitude of L is greater than the absolute magnitude of R1, L is not vertical and the absolute magnitude of L and the absolute magnitude of R1 are above a certain amount; Cv = L * 0.5 + R1 * 0.5

The absolute value magnitude of -L is not large relative to the absolute value magnitude of R1, and L is close to vertical; Cv = R1

When the absolute value magnitude of L is not large relative to the absolute value magnitude of R1, L is not vertical and the difference between the absolute value magnitude of L and the absolute value magnitude of R1 is within a certain amount; Cv = L * 0.2 + R1 * 0.8

The absolute value magnitude of L is not large relative to the absolute value magnitude of R1, L is not vertical and the difference between the absolute value magnitude of L and the absolute value magnitude of R1 is greater than or equal to a certain amount; Cv = L * 0.5 + R1 * 0.5

FIG. 16 illustrates an embodiment of a flowchart for obtaining the vertical center symbol gradient initial value Cv when the resolution of the second frame memory is 640 × 480. An example of extracting the vertical center symbol inclination initial value Cv based on the above embodiment will be described as the actual value.

When the magnitude of the absolute value of -L is greater than the magnitude of the absolute value of R1 and L is less than 0.011 (near vertical); Cv = L

The absolute value magnitude of -L is greater than the absolute value magnitude of R1, L is not less than 0.011 (not vertical), and the difference between the absolute value magnitude of L and the absolute value magnitude of R1 is within 0.017; Cv = L * 0.8 + R1 * 0.2

The absolute magnitude of L is greater than the absolute magnitude of R1, L is not less than 0.011 (not vertical), and the absolute magnitude of L and the absolute magnitude of R1 are greater than or equal to 0.017; Cv = L * 0.5 + R1 * 0.5

When the magnitude of the absolute value of -L is not greater than the magnitude of the absolute value of R1 and the absolute value of L is less than 0.009 (near vertical); Cv = R1

When the absolute magnitude of L is not greater than the absolute magnitude of R1, the absolute magnitude of L is not less than 0.009, and the difference between the absolute magnitude of L and the absolute magnitude of R1 is less than 0.017; Cv = L * 0.2 + R1 * 0.8

When the absolute magnitude of L is not greater than the absolute magnitude of R1, the absolute magnitude of L is not less than 0.009, and the difference between the absolute magnitude of L and the absolute magnitude of R1 is not less than 0.017; Cv = L * 0.5 + R1 * 0.5

FIG. 17 shows a general flowchart for obtaining the vertical center symbol gradient initial value Cv.

When the magnitude of the absolute value of -L is greater than the magnitude of the absolute value of R1, and the absolute value of L is smaller than the first vertical allowable value Gv1 (close to the vertical); Cv = L

The absolute magnitude of -L is greater than the absolute magnitude of R1, the absolute value of L is not less than the vertical tolerance Gv1 (not vertical), and the difference between the absolute magnitude of L and the absolute magnitude of R1 is within Gv3. ; Cv = L * 0.8 + R1 * 0.2

The absolute value magnitude of -L is greater than the absolute value magnitude of R1, the absolute value of L is not less than the vertical tolerance Gv1 (not vertical), and the absolute value magnitude of L and the absolute value magnitude of R1 are greater than or equal to Gv3; Cv = L * 0.5 + R1 * 0.5

When the absolute value magnitude of -L is not large compared to the absolute value magnitude of R1, and the absolute value of L is smaller than the second vertical allowable value Gv2 (close to the vertical); Cv = R1

When the absolute value magnitude of -L is not larger than the absolute value magnitude of R1, the absolute value of L is not less than the second vertical allowable value Gv2, and the difference between the absolute value magnitude of L and the absolute value magnitude of R1 is smaller than Gv3; Cv = L * 0.2 + R1 * 0.8

When the absolute value magnitude of -L is not larger than the absolute value magnitude of R1, the absolute value of L is not smaller than the second vertical allowable value Gv2, and the difference between the absolute value magnitude of L and the absolute value magnitude of R1 is not smaller than Gv3. ; Cv = L * 0.5 + R1 * 0.5

When the vertical center symbol gradient initial value Cv is obtained by the above process, the controller 160 extracts coordinates of the symbols from the bar code image in the vertical center direction.

13) FIG. 18 illustrates a configuration for obtaining coordinates of symbols in a vertical center direction from a barcode image of the second frame memory 150. As described above, when the vertical center symbol slope initial value Cv is obtained, the controller 160 divides the lengths of both end coordinates of the lower finder pattern 104 by the number of symbols, and the vertical center symbol slope initial value passing through the coordinates of the symbol located at the center. The linear equation of Cv is obtained, the number of symbols is retrieved by moving the straight line in the right direction, and the coordinate information of the vertical center symbols is stored in the array when the number of symbols coincides with the number of symbols in the right vertical direction of the barcode.

14) When the start position and the end point coordinate information of the vertical center symbols are extracted, the controller 160 provides a binary barcode image of the second frame memory 150 to the symbol operator 250 so that the upper horizontal symbol of the barcode image is provided. Extract the coordinates of. First, the symbol operation unit 250 obtains a slope including all of the symbols placed in the upper horizontal direction of the barcode.

FIG. 19 illustrates a configuration for obtaining a slope including all the symbols placed in the upper horizontal direction with respect to the binary barcode image of the second frame memory 150. The slope including all the symbols placed in the upper horizontal direction of the barcode is based on the upper left coordinate (vert_left (x, y)) 153 of the binary barcode image, and includes the second frame memory 150 including the image. Coordinates of the upper right corner of the second frame memory 150 in the origin direction of the second frame memory 150 starting with a straight line of the slope formed with the coordinates of the upper right corner of the upper right corner (vert_top (x, y)) 151. The gradient including the maximum number of symbols is obtained by moving 151 by one pixel in the y-axis direction.

In this step, the symbol size of the barcode is determined, and the time point at which the symbol size is determined is the upper right endpoint from the slope with the coordinate (vert_top (x, y)) 151 of the upper right corner of the second frame memory 150. Stops when the number of symbols retrieved on the straight line of the slope calculated while moving in the y-axis direction decreases from the even maximum again, and stores the slope where the maximum number of symbols is detected, the number of symbols and the coordinates of the symbols in the array. do.

15) Then, the control unit obtains a slope including all the symbols placed in the center horizontal position of the barcode through the symbol operation unit 250. The horizontal center symbol inclination initial value Ch is determined according to five cases based on the inclination value of all the symbols placed in the upper direction of the barcode and the inclination value of the lower finder pattern 104. At this time, each slope is calculated by converting the length in the y-axis direction with respect to the length in the x-axis direction in consideration of the significant number of the binary operation. As a method of determination, if the initial value of the vertical center symbol slope is Ch, the slope including all the symbols placed in the right vertical direction of the barcode is R2, and the slope value of the lower finder pattern 104 is B,

When the magnitude of absolute value of B is greater than the magnitude of absolute value of R2, and B is close to horizontal; Ch = B

The absolute magnitude of B is greater than the absolute magnitude of R2, B is not horizontal and the absolute magnitude of B is within a certain amount relative to the absolute magnitude of R2; Ch = B * 0.8 + R2 * 0.2

The absolute magnitude of B is greater than the absolute magnitude of R2, B is not horizontal and the absolute magnitude of B and the absolute magnitude of R2 are above a certain amount; Ch = B * 0.5 + R2 * 0.5

When the absolute magnitude of B is not large relative to the absolute magnitude of R2 and B is close to horizontal; Ch = R2

When the absolute magnitude of B is not large relative to the absolute magnitude of R2, B is not horizontal and the difference between the absolute magnitude of B and the absolute magnitude of R2 is within a certain amount; Ch = B * 0.2 + R2 * 0.8

When the absolute magnitude of B is not large relative to the absolute magnitude of R2, B is not horizontal and the difference between the absolute magnitude of BL and the absolute magnitude of R2 is above a certain amount; Ch = B * 0.5 + R2 * 0.5

FIG. 20 illustrates an embodiment of a flowchart for obtaining the horizontal center symbol gradient initial value Ch when the resolution of the second frame memory is 640 × 480. An example of extracting the horizontal center symbol inclination initial value Ch based on the above embodiment will be described as the actual value.

When the absolute magnitude of -B is greater than the absolute magnitude of R2 and B is less than 0.011 (close to horizontal); Ch = B

When the absolute magnitude of B is greater than the absolute magnitude of R1, B is not less than 0.011 (not horizontal), and the difference between the absolute magnitude of B and the absolute magnitude of R2 is within 0.017; Ch = B * 0.8 + R2 * 0.2

When the absolute magnitude of B is greater than the absolute magnitude of R2, B is not less than 0.011 (not horizontal), and the absolute magnitude of B and the absolute magnitude of R2 are greater than or equal to 0.017; Ch = B * 0.5 + R2 * 0.5

When the magnitude of the absolute value of B is not large relative to the magnitude of the absolute value of R2, and the absolute value of B is less than 0.009 (close to horizontal); Ch = R2

When the absolute magnitude of B is not larger than the absolute magnitude of R2, the absolute value of B is not less than 0.009, and the difference between the absolute magnitude of B and the absolute magnitude of R2 is less than 0.017; Ch = B * 0.2 + R2 * 0.8

When the absolute magnitude of B is not greater than the absolute magnitude of R2, the absolute value of B is not less than 0.009, and the difference between the absolute magnitude of B and the absolute magnitude of R2 is not less than 0.017; Ch = B * 0.5 + R2 * 0.5

21 shows a general flowchart for obtaining the above-described horizontal center symbol gradient initial value Ch.

When the magnitude of the absolute value of B is greater than the magnitude of the absolute value of R2 and the absolute value of B is less than the first horizontal tolerance Gh1 (close to vertical); Ch = B

The absolute magnitude of -B is greater than the absolute magnitude of R2, the absolute value of B is not less than the horizontal tolerance Gh1 (not vertical), and the difference between the absolute magnitude of B and the absolute magnitude of R1 is within Gh3. ; Ch = B * 0.8 + R2 * 0.2

When the absolute magnitude of B is greater than the absolute magnitude of R2, the absolute value of B is not less than the horizontal tolerance Gh1 (not vertical), and the absolute magnitude of B and the absolute magnitude of R2 are greater than or equal to Gh3; Ch = B * 0.5 + R1 * 0.5

When the absolute value magnitude of -B is not large relative to the absolute value magnitude of R2, and the absolute value of B is smaller than the second horizontal tolerance value Gh2 (close to vertical); Ch = R2

When the absolute magnitude of B is not greater than the absolute magnitude of R2, the absolute value of B is not less than the second vertical tolerance Gh2, and the difference between the absolute magnitude of B and the absolute magnitude of R2 is less than Gh3; Ch = B * 0.2 + R2 * 0.8

When the absolute magnitude of -B is not greater than the absolute magnitude of R2, the absolute value of B is not less than the second vertical tolerance Gh2, and the difference between the absolute magnitude of B and the absolute magnitude of R2 is not less than Gh3. ; Ch = B * 0.5 + R2 * 0.5

When the horizontal center symbol inclination initial value Ch is obtained by the above process, the controller 160 extracts coordinates of the symbols from the bar code image in the horizontal center direction.

16) FIG. 22 illustrates a configuration for obtaining coordinates of symbols in a horizontal center direction from a barcode image of the second frame memory 150. Referring to FIG. As described above, when the horizontal center symbol slope initial value Ch is obtained, the controller 160 divides the lengths of both end coordinates of the right finder pattern 103 by the number of symbols, and the vertical center symbol slope initial value passing through the coordinates of the symbol located at the center. Find the linear equation of Ch, search the number of symbols while moving the straight line in the upper direction (-y direction), and store the coordinate information of the horizontal center symbols in the array when it matches the number of symbols placed in the upper horizontal direction of the barcode. do

17) by the above processing, the control unit 160, through the symbol operation unit 250, from the binary bar code image of the second frame memory 150, the symbols placed in the right vertical direction of the bar code; Symbols placed at the center vertical position of the barcode; Symbols placed in the top direction of the barcode; Since the coordinates of the symbols placed at the center horizontal position of the barcode are extracted, linear equations for barcode reading are extracted therefrom. The controller 160 extracts horizontal linear equations that sequentially connect the coordinates of the symbols placed in the right vertical direction of the barcode of the second frame memory 150 with the coordinates of the symbols placed at the center vertical position of the barcode, and then the barcode. Extract the straight-line equations in the vertical direction connecting the symbols placed in the upper direction of and the coordinates in the center horizontal position of the barcode.

FIG. 23 illustrates an arrangement relationship of the linear equations in the horizontal direction and the linear equations in the vertical direction of the barcode of the second frame memory 150. FIG. Since the straight line equations in the horizontal direction and the straight line equations in the vertical direction are straight lines connecting the symbols in the horizontal-vertical direction, barcode dots are formed on each straight line.

18) FIG. 24 shows the intersection of a single horizontal straight line equation and a vertical straight line equation among the straight lines connecting the symbols in the horizontal-vertical direction. The controller 160 calculates coordinates of the second frame memory 150 that intersect in the horizontal-vertical direction using the linear equations for each of the straight lines and uses the image data of the corresponding coordinates to generate the 2D matrix memory 260. Complete the barcode. In this case, the 2D matrix memory 260 is configured to correspond to the horizontal and vertical number of the symbol of the barcode to be read, and is configured such that one symbol is arranged at one address.

FIG. 25 illustrates a conversion relationship with respect to the coordinates of the second frame memory 150 in which the controller 160 intersects in the horizontal-vertical direction using the above linear equations. When the binary image of the coordinates of the second frame memory 150 intersecting in the horizontal-vertical direction is read as black using the above linear equations, the controller 160 sets '1' to the corresponding 2D matrix memory 260 address. When the binary image of the coordinates of the second frame memory 150 that intersects in the horizontal-vertical direction is read as white, the controller 160 records a series of '0's at the corresponding 2D matrix memory 260 addresses. The process is performed on all straight lines to complete the 2D barcode in the 2D matrix memory 260.

FIG. 26 illustrates an example in which the controller 16 generates a 2D barcode in the 2D matrix memory 260 by using barcode data read from the second frame memory 150. Since the data in the 2D matrix memory 260 generated as described above is a barcode composed of '1' and '0', the controller 160 can decode the barcode through the 2D matrix reading unit 160a having the barcode decoding means. Will be. FIG. 27 illustrates a configuration relationship for decoding data in the 2D matrix memory 260 through the 2D matrix reading unit 160a.

Hereinafter, the method of extracting and decoding the data matrix barcode included in the image of the present invention will be described by referring to the flowcharts illustrated in FIGS. 28 to 30 based on the configuration of the two-dimensional barcode extraction apparatus included in the image of the present invention. Explain as.

1) copying input image (S1010)

When the two-dimensional bar code extraction starts (S1000), in step S1010, the controller 160 transmits a command to the direct memory access (DMA) controller 140 to store the camera 110 stored in the one-frame memory 120. Image data from the second frame memory 150 is stored.

2) gray level conversion step (S1020)

Following the input image copying step (S1010), the controller 160 converts the color image of the second frame memory 150 to the gray level through the gray level converter 180.

3) Dual linear interpolation step (S1030)

Following the gray level conversion step (S1020), the controller 160 reduces and stores the image data of the second frame memory 150 converted to gray level by the dual linear interpolator 190 to the operation frame memory 170. . In the present invention, the image data of the second frame memory 150 is reduced and stored in the operation frame memory 170 at a reduction ratio of 1/4.

4) Storage step of the operation frame memory 170 (S1040)

Following the double linear interpolation step (S1030), the controller 160 stores the image data of the reduced second frame memory 150 in the operation frame memory 170.

5) Edge extraction step (S1050)

Subsequent to the operation frame memory 170 storing step (S1040), the controller 160 provides the reduced image data stored in the operation frame memory 170 to the area filter unit 200 to calculate edge values of the image. do. The area filter unit 200 allocates a 3 × 3 pixel size memory to be used as a mask for the gray level image in the operation frame memory 170, and three horizontally and vertically around the position of the current target pixel. Read the pixel value of to find the maximum and minimum values and compute the difference as the edge value. When the edge value calculation is completed for the entire reduced image data, the image composed of the edge value is stored in the operation frame memory 170.

6) Binarization image extraction step (S1060)

Following the edge extraction step (S1050), the controller 160 provides the edge value image data of the reduced gray level image data stored in the operation frame memory 170 to the binary image converter 210 to convert the image into a binarized image.

7) Maximum connection area extraction step (S1070)

After extracting the binarized image (S1060), the controller 160 provides the edge region binarized image data of the reduced image data stored in the operation frame memory 170 to the connection region extractor 220, so that the adjacent ' White 'pixel groups are judged, each group is assigned a unique number for labeling, and the' white 'pixel group is composed of the largest number of pixels among the' white 'pixel groups separated by the labeling. Performing the maximum connection area extraction step (S1070) for extracting.

In this case, the connection area extractor 220 replaces the group consisting of the most pixels among the 'white' pixel groups classified by the labeling with 'white', and pixel values of the remaining groups except the pixel group. Substituting for 'black' extracts the candidate region of the barcode of 'white'.

8) boundary coordinates extraction step (S1080)

Following the maximum connection region extraction step (S1070), the controller 160 provides the peripheral gradient extracting unit 230 with image data of the operation frame memory 170 including the candidate region of the extracted 'white' barcode. In addition, the vertical height Labeling_height and the slope (vertical gradient: Gradient_Vert) are calculated from the boundary coordinates of the top and bottom of the bar code candidate area spaced by d about the center in the vertical direction of the image data of the operation frame memory 170, and horizontally. The horizontal length Labeling_width and the slope (horizontal gradient: Gradient_Horz) are calculated from the right boundary coordinates and the left boundary coordinates of the barcode candidate region spaced by d from the center in the direction. In an embodiment, when the size of the second frame memory 150 is 640 X 480 and the size of the operation frame memory 170 is 160 X 120, the separation distance is set to 5 to 10 pixels, so that the vertical height Labeling_height and the slope ( Vertical slope: Gradient_Vert) and horizontal length Labeling_width and slope (horizontal slope: Gradient_Horz) can be calculated.

9) Horizontal Vertical Length Effective Determination Step (S1090)

Subsequent to the boundary coordinate tilt extraction step (S1080), the controller 160 determines whether the barcode is valid by using the vertical height Labeling_height and the horizontal length Labeling_width of the barcode candidate region provided by the connection region extraction unit 220. To judge. The above determination is considered to be possible to read when the ratio of the vertical height Labeling_height and the horizontal length Labeling_width of the barcode candidate region stored in the first frame memory 120 is within a predetermined ratio, and is considered to be a recognition failure when there is an excessive difference. In response to the transfer command of the controller 160, the process is repeated from the input image copying step (S1010) of storing new image data from the camera 110 from the first frame memory 120 to the second frame memory 150. To perform.

10) vertical slope valid determination step (S1091)

Following the horizontal vertical length valid determination step (S1090), the controller 160 determines whether the vertical gradient Gradient_Vert calculated by the peripheral gradient extractor 230 is a valid value of the barcode reading. When the vertical gradient Gradient_Vert value calculated by the peripheral gradient extractor 230 is close to the vertical or the horizontal gradient Gradient_Horz value is close to the horizontal, it is determined that the reading is possible. The process is performed again from the input image copying step (S1010) of storing new image data from the camera 110 from the first frame memory 120 to the second frame memory 150 by a transmission command of. In the above determination, the vertical gradient Gradient_Vert value may be obtained by using boundary coordinates of candidate regions of the barcode extracted by the peripheral gradient extractor 230 from the image data of the operation frame memory 170. At this time, it is preferable to calculate the vertical gradient Gradient_Vert under the same condition as the Gradient_Horz value. In this case, the absolute value [Gradient_Vert] may be expressed as a small value close to 0 as the vertical value gets closer.

Therefore, when the vertical gradient Gradient_Vert is close to vertical and the allowable slope is r, it is determined that the reading is possible when the absolute value [Gradient_Vert]) <r, otherwise it is regarded as a recognition failure.

As an example, when the horizontal and vertical resolution of the image data of the second frame memory 150 is 640 × 480 and the image data of the operation frame memory 170 is 160 × 120, as an example, (Absolute value [ Gradient_Vert]) <0.000001, it is determined that reading is possible.

11) Horizontal slope valid determination step (S1092)

Following the vertical slope valid determination step (S1091), the controller 160 determines whether the horizontal slope Gradient_Horz calculated by the peripheral slope extractor 230 is a valid value. When the horizontal gradient Gradient_Horz value calculated by the peripheral gradient extractor 230 is close to the horizontal, it is determined that the barcode can be read. The process is performed again from the input image copying step (S1010) of storing new image data from the first frame memory 120 to the second frame memory 150. In the above determination, the horizontal gradient Gradient_Horz value may be obtained using boundary coordinates of candidate regions of the barcode extracted by the peripheral gradient extractor 230 from the image data of the operation frame memory 170.

At this time, the absolute value [Gradient_Horz] has a small value close to 0 as it is horizontal. Therefore, when the horizontal gradient Gradient_Horz value is close to the horizontal, a condition for determining that the reading is possible is that if the allowable slope is r, the barcode can be read when the absolute value [Gradient_Hor] <r.

As an example, when the horizontal and vertical resolution of the image data of the second frame memory 150 is 640 × 480 and the image data of the operation frame memory 170 is 160 × 120, the absolute value [Gradient_Hor] When it is <0.000001, it is determined that the barcode can be read.

12) calculating the angle of the horizontal vertical tilt (S1100)

Following the horizontal slope valid determination step (S1092), the controller 160 determines the gradient Gradient_Vert and the horizontal gradient Gradient_Horz to determine whether the vertical gradient Gradient_Vert and the horizontal gradient Gradient_Horz values calculated by the peripheral gradient extractor 230 are perpendicular to each other. Multiply the values by each other. In the above calculation, the vertical gradient Gradient_Vert and the horizontal gradient Gradient_Horz values may be obtained by using boundary coordinates of candidate regions of the barcode extracted by the peripheral gradient extractor 230 from image data of the operation frame memory 170. At this time, it is preferable to calculate the vertical gradient Gradient_Vert under the same condition as the Gradient_Horz value. For example, referring to FIG. 17, the drawing is as follows. In FIG. 17, when the left boundary coordinates (Xh, Yh) and the lower boundary coordinates (Xc, Yc) and the left boundary coordinates (Xd, Yd) are used,

[Gradient_Vert] = (Yh-Yg) / (Xh-Xg)

[Gradient_Horz] = (Yc-Yd) / (Xc-Xd).

At this time, the absolute value [Gradient_Vert] is close to vertical and the absolute value [Gradient_Horz] is close to 0, and has a small value close to zero.

13) Angle determination step of the horizontal vertical tilt (S1110)

Following the step of calculating the angle of horizontal vertical tilt (S1100), the controller 160 determines that the product can be read when the product of the vertical gradient Gradient_Vert and the horizontal gradient Gradient_Horz is less than 0. The process is performed again from the input image copying step (S1010) of storing new image data from the camera 110 from the first frame memory 120 to the second frame memory 150 by the transfer command of 160.

14) Barcode Area Extraction Step (S1120)

The boundary coordinate inclination extraction step (S1080), the horizontal vertical length valid determination step (S1090), the vertical slope valid determination step (S1091), the horizontal slope valid determination step (S1092), the horizontal vertical slope angle calculation step (S1100), horizontal In step S1110 of determining the angle of the vertical tilt, the controller 160 calculates the vertical height Labeling_height and the horizontal length Labeling_width of the barcode candidate region provided by the connection region extraction unit 220 and the peripheral slope extraction unit 230. When it is determined that the barcode can be read using the vertical gradient Gradient_Vert and the horizontal gradient Gradient_Horz, the controller 160 determines the barcode candidate in the image of the second frame memory 150 corresponding to the barcode candidate region of the operation frame memory 170. Replace the pixel value of the area except the area with 'white'.

15) Extraction area valid determination step (S1130)

Following the bar code area extraction step (S1120), the controller 160 determines that the bar code candidate area of the second frame memory 150 can be read if the size is greater than or equal to a predetermined value. In response, the new image data from the camera 110 is stored in the second frame memory 150 from the first frame memory 120 by the transfer command of the controller 160 and the process is performed again from the beginning. In the present invention, in the second frame memory 150, for example, when the number of pixels in the bar code candidate area is less than or equal to the number of symbols, the recognition failure is regarded as a new image from the camera 110 by the transmission command of the controller 160. The process is performed again from the input image copying step (S1010) of storing data from the first frame memory 120 to the second frame memory 150.

16) Binarization step of the extraction region image (S1140)

If it is determined that the size of the barcode candidate area of the second frame memory 150 is readable in the extraction area valid determination step (S1130), the controller 160 controls the gray corresponding to the barcode candidate area of the second frame memory 150. By supplying a level image to the binary image converter 210 to perform binary image conversion, the binary image corresponding to the barcode candidate region is stored in the second frame memory 150.

17) Finder pattern search step (S1150)

Following the binarization step S1140 of the extracted region image, the controller 160 detects the finder pattern by providing the binary image to the finder pattern detection / rotation unit 240.

In order to detect the position and inclination of the lower finder pattern 104, the finder pattern detection / rotation unit 240 uses the upper left side of the screen as the origin of the second frame memory 150 as the x axis and the vertical direction as the y axis. When the x coordinate axis is roughly divided into three parts, the integer coordinate corresponding to one third of the bottom x coordinate of the screen and the integer coordinate corresponding to two thirds of the bottom x coordinate of the screen are the directions of the screen's upper direction, which is the origin direction of the y coordinate. Move to to find the 'black' point of the binary image and search for each to get the slope and position between the two coordinates.

In order to detect the position and inclination of the left finder pattern 103, the finder pattern detection / rotation unit 240 uses the upper left side of the screen corresponding to the second frame memory 150 as the origin, and the horizontal direction is the x axis and the vertical direction is y. In the case of an axis, the y coordinate axis is roughly divided into three parts, and the integer coordinates corresponding to one third of the left y-coordinate on the screen and the integer coordinates corresponding to two thirds of the y-coordinate on the left side of the screen are moved from the integer coordinates to the right side of the screen in the x coordinate direction. As we move, we search each one until we find 'black' point of binary image and find slope and position between two coordinates.

18) Image Rotation Step (S1160)

Following the finder pattern search operation (S1150), the controller 160 rotates the binary image in units of 90 degrees according to the detection result of the finder pattern to position the binary data matrix barcode image in the normal direction.

At this time, the controller 160 determines the rotation by checking the authenticity of the left finder pattern 103 and the lower finder pattern 104 detected by the finder pattern detection / rotation unit 240. The characteristic of the data matrix bar code is determined so that the angle between the left finder pattern 103 which is the lower solid line and the lower finder pattern 104 which is the solid left line in the rectangular bar code is 90 degrees. Therefore, it is necessary to determine whether the searched left finder pattern 103 and the lower finder pattern 104 are normal solid lines. The judging method corresponds to one-third of the y-coordinate by roughly dividing the y-coordinate axis when the upper-left corner of the screen of the second frame memory 150 is the origin as the x-axis and the vertical-direction y-axis. Pixels of y coordinate ± 1, moving from the coordinates corresponding to the two coordinates and the two thirds to the right, and moving the points on the slope between the two coordinates obtained by searching each one until the 'black' point of the binary image is found. Check that the value is 'white'. If any one of the pixel values of the three coordinates is 'white', since it is not a solid line, it is determined that the rotation is necessary so that the left finder pattern 103 is in the correct position. The determination of the rotation by the lower finder pattern 104 is also determined in the same manner as in the left finder pattern 103.

When the upper left corner of the screen of the second frame memory 150 is the origin, the x-axis and the y-axis are divided into approximately three equal parts and the integer coordinates corresponding to one third of the x coordinate Pixels of x coordinate ± 1, moving from the integer coordinates corresponding to two thirds upward in the -y coordinate direction, moving the points on the slope between the two coordinates retrieved, respectively, until the 'black' point of the binary image is found. Check that the value is 'white'. If any one of the pixel values of the x coordinate ± 1 coordinate is 'white', since it is not a solid line, it is determined that the rotation is necessary so that the lower finder pattern 104 is in the correct position.

The rotation of the binary image is performed only in 90-degree increments to prevent pixel value loss. Therefore, if the lower finder pattern 104 is a solid line and the left finder pattern 103 is a solid line, rotation is unnecessary, and if the lower finder pattern 104 is a solid line and the left finder pattern 103 is not a solid line, it is rotated 90 degrees clockwise. If the lower finder pattern 104 is not a solid line and the left finder pattern 103 is a solid line, it is rotated 90 degrees counter clockwise, and if it is not a solid line, it is rotated clockwise 180 degrees. Through this process, the binary barcode image of the second frame memory 150 has the left finder pattern 103 and the lower finder pattern 104 positioned on the left side and the lower side, respectively, and the finder pattern is positioned in an 'L' shape. .

19) Detecting the finder pattern tilt and coordinates (S1170)

Following the image rotation step (S1160), the control unit 160 has a left finder pattern 103 and a lower finder pattern 104 positioned in an 'L' shape on the left and lower sides of the binary barcode image of the second frame memory 150. Calculate its slope and the coordinates of both ends of each pattern.

20) extracting the right symbol coordinate (S1180)

Following the finder pattern tilt and coordinate extraction step (S1170), the controller 160 provides a binary barcode image of the second frame memory 150 to the symbol calculator 250 to extract the coordinates of the symbol. First, the symbol operator 250 obtains a slope including all the symbols placed in the right vertical direction of the barcode. The slope that includes all the symbols placed in the right vertical direction of the barcode is based on the end coordinates (vert_right (x, y)) 152 of the lower right of the binary barcode image. Coordinates of the upper right corner (vert_top (x, y)) in the origin direction of the second frame memory 150 starting from a straight line of the slope formed with the coordinates of the upper right corner (vert_top (x, y)) 151. A slope including the maximum number of symbols is obtained while moving 151 by one pixel in the origin direction.

In this step, the symbol size of the barcode is determined, and the time at which the symbol size is determined is the upper right endpoint from the slope with the coordinate (vert_top (x, y)) 151 of the upper right end corner of the second frame memory 150. When the number of symbols retrieved on the straight line of the slope calculated while moving to the left decreases again from the even maximum value, it stops, and the slope where the maximum number of symbols is detected, the number of symbols and the coordinates of the symbols are stored in the array.

21) central vertical symbol coordinate extraction step (S1190)

Following the step of extracting the right symbol coordinate (S1180), the controller 160 obtains a slope including all the symbols placed at the center vertical position of the barcode through the symbol calculator 250. The vertical center symbol inclination initial value Cv is determined according to five cases based on the inclination value of all the symbols placed in the right vertical direction of the barcode and the inclination value of the left finder pattern 103. At this time, each slope is calculated by converting the length in the x-axis direction with respect to the length in the y-axis in consideration of the significant number of binary operations. As a method of judgment, if the initial value of the vertical center symbol slope is Cv, the slope including all the symbols placed in the right vertical direction of the barcode is R1, and the slope value of the left finder pattern 103 is L,

When the absolute magnitude of -L is greater than the absolute magnitude of R1 and L is close to vertical; Cv = L

The absolute magnitude of L is greater than the absolute magnitude of R1, L is not vertical and the absolute magnitude of L is within a certain amount relative to the absolute magnitude of R1; Cv = L * 0.8 + R1 * 0.2

The absolute magnitude of L is greater than the absolute magnitude of R1, L is not vertical and the absolute magnitude of L and the absolute magnitude of R1 are above a certain amount; Cv = L * 0.5 + R1 * 0.5

The absolute value magnitude of -L is not large relative to the absolute value magnitude of R1, and L is close to vertical; Cv = R1

When the absolute value magnitude of L is not large relative to the absolute value magnitude of R1, L is not vertical and the difference between the absolute value magnitude of L and the absolute value magnitude of R1 is within a certain amount; Cv = L * 0.2 + R1 * 0.8

The absolute value magnitude of L is not large relative to the absolute value magnitude of R1, L is not vertical and the difference between the absolute value magnitude of L and the absolute value magnitude of R1 is greater than or equal to a certain amount; Cv = L * 0.5 + R1 * 0.5

As an embodiment of the present invention, an example of extracting the vertical center symbol gradient initial value Cv when the resolution of the second frame memory is 640 × 480 will be described as follows.

When the magnitude of the absolute value of -L is greater than the magnitude of the absolute value of R1 and L is less than 0.011 (near vertical); Cv = L

The absolute value magnitude of -L is greater than the absolute value magnitude of R1, L is not less than 0.011 (not vertical) and the difference between the absolute value magnitude of L and the absolute value magnitude of R1 is within 0.017; Cv = L * 0.8 + R1 * 0.2

The absolute magnitude of L is greater than the absolute magnitude of R1, L is not less than 0.011 (not vertical), and the absolute magnitude of L and the absolute magnitude of R1 are greater than or equal to 0.017; Cv = L * 0.5 + R1 * 0.5

When the magnitude of the absolute value of -L is not greater than the magnitude of the absolute value of R1 and the absolute value of L is less than 0.009 (near vertical); Cv = R1

When the absolute value magnitude of L is not greater than the absolute value magnitude of R1, the absolute value of L is not less than 0.009, and the difference between the absolute value magnitude of L and the absolute value magnitude of R1 is less than 0.017; Cv = L * 0.2 + R1 * 0.8

When the absolute magnitude of L is not greater than the absolute magnitude of R1, the absolute magnitude of L is not less than 0.009, and the difference between the absolute magnitude of L and the absolute magnitude of R1 is not less than 0.017; Cv = L * 0.5 + R1 * 0.5

Calculate

The above-described steps for obtaining the vertical center symbol gradient initial value Cv are as follows.

When the magnitude of the absolute value of -L is greater than the magnitude of the absolute value of R1 and the absolute value of L is less than the first vertical tolerance value Gv1 (close to the vertical); Cv = L

The absolute magnitude of -L is greater than the absolute magnitude of R1, the absolute value of L is not less than the vertical tolerance Gv1 (not vertical), and the difference between the absolute magnitude of L and the absolute magnitude of R1 is within Gv3. ; Cv = L * 0.8 + R1 * 0.2

The absolute value magnitude of -L is greater than the absolute value magnitude of R1, the absolute value of L is not less than the vertical tolerance Gv1 (not vertical), and the absolute value magnitude of L and the absolute value magnitude of R1 are greater than or equal to Gv3; Cv = L * 0.5 + R1 * 0.5

When the absolute value magnitude of -L is not large compared to the absolute value magnitude of R1, and the absolute value of L is smaller than the second vertical allowable value Gv2 (close to the vertical); Cv = R1

When the absolute value magnitude of -L is not larger than the absolute value magnitude of R1, the absolute value of L is not less than the second vertical allowable value Gv2, and the difference between the absolute value magnitude of L and the absolute value magnitude of R1 is smaller than Gv3; Cv = L * 0.2 + R1 * 0.8

When the absolute value magnitude of -L is not larger than the absolute value magnitude of R1, the absolute value of L is not smaller than the second vertical allowable value Gv2, and the difference between the absolute value magnitude of L and the absolute value magnitude of R1 is not smaller than Gv3. ; Cv = L * 0.5 + R1 * 0.5

Calculate

When the vertical center symbol gradient initial value Cv is obtained by the above process, the controller 160 extracts coordinates of the symbols from the bar code image in the vertical center direction.

As described above, when the vertical center symbol slope initial value Cv is obtained, the controller 160 divides the length of both end coordinates of the lower finder pattern 104 by the number of symbols, and the vertical center symbol slope initial value Cv that passes through the coordinates of the symbol located at the center. The equation of a straight line is obtained, and the coordinates of the vertical center symbols are stored in the array when the number of symbols is matched with the number of symbols placed in the right vertical direction of the barcode by moving the straight line in the right direction.

22) the upper horizontal symbol coordinate extraction step (S1200)

After the central vertical symbol coordinate extraction step (S1190), the controller 160 provides the binary barcode image of the second frame memory 150 to the symbol operator 250 to extract the coordinates of the upper horizontal symbol of the barcode image. . First, the symbol operation unit 250 obtains a slope including all of the symbols placed in the top horizontal direction of the barcode.

The second frame memory 150 including the image may include a slope including all the symbols placed in the top horizontal direction of the barcode based on the upper left coordinate (vert_left (x, y)) 153 of the binary barcode image. The coordinates of the upper right end corner in the origin direction of the second frame memory 150 starting with the straight line of the slope formed with the coordinates (vert_top (x, y)) 151 of the upper right end corner of the second frame memory (vert_top (x, y)). The gradient including the maximum number of symbols is obtained by moving 151 by one pixel in the y-axis direction.

In this step, the symbol size of the barcode is determined, and the time point at which the symbol size is determined is the upper right endpoint from the slope with the coordinate (vert_top (x, y)) 151 of the upper right corner of the second frame memory 150. Stops when the number of symbols retrieved on the straight line of the slope calculated while moving in the y-axis direction decreases from the even maximum again, and stores the slope where the maximum number of symbols is detected, the number of symbols and the coordinates of the symbols in the array. do.

23) central horizontal symbol coordinate extraction step (S1210)

Subsequent to the above-described upper horizontal symbol coordinate extraction step (S1200), the controller 160 obtains a slope including all the symbols placed at the center horizontal position of the barcode through the symbol calculator 250. The horizontal center symbol inclination initial value Ch is determined according to five cases based on the inclination value of all the symbols placed in the upper direction of the barcode and the inclination value of the lower finder pattern 104. At this time, each slope is calculated by converting the length in the y-axis direction with respect to the length in the x-axis direction in consideration of the significant number of the binary operation. As a method of judgment, if the initial value of the horizontal center symbol slope is Ch, the slope including all the symbols placed in the upper horizontal direction of the barcode is R2, and the slope value of the lower finder pattern 104 is B,

When the magnitude of absolute value of B is greater than the magnitude of absolute value of R2, and B is close to horizontal; Ch = B

The absolute magnitude of B is greater than the absolute magnitude of R2, B is not horizontal and the absolute magnitude of B is within a certain amount relative to the absolute magnitude of R2; Ch = B * 0.8 + R2 * 0.2

The absolute magnitude of B is greater than the absolute magnitude of R2, B is not horizontal and the absolute magnitude of B and the absolute magnitude of R2 are above a certain amount; Ch = B * 0.5 + R2 * 0.5

When the absolute magnitude of B is not large relative to the absolute magnitude of R2 and B is close to horizontal; Ch = R2

When the absolute magnitude of B is not large relative to the absolute magnitude of R2, B is not horizontal and the difference between the absolute magnitude of B and the absolute magnitude of R2 is within a certain amount; Ch = B * 0.2 + R2 * 0.8

When the absolute magnitude of B is not large relative to the absolute magnitude of R2, B is not horizontal and the difference between the absolute magnitude of BL and the absolute magnitude of R2 is above a certain amount; Ch = B * 0.5 + R2 * 0.5

Calculate

As an embodiment of the present invention, an example of extracting the horizontal center symbol gradient initial value Ch when the resolution of the second frame memory is 640 × 480 will be described as follows.

When the absolute magnitude of -B is greater than the absolute magnitude of R2 and B is less than 0.011 (close to horizontal); Ch = B

When the absolute magnitude of B is greater than the absolute magnitude of R1, B is not less than 0.011 (not horizontal), and the difference between the absolute magnitude of B and the absolute magnitude of R2 is within 0.017; Ch = B * 0.8 + R2 * 0.2

When the absolute magnitude of B is greater than the absolute magnitude of R2, B is not less than 0.011 (not horizontal), and the absolute magnitude of B and the absolute magnitude of R2 are greater than or equal to 0.017; Ch = B * 0.5 + R2 * 0.5

When the magnitude of the absolute value of B is not large relative to the magnitude of the absolute value of R2, and the absolute value of B is less than 0.009 (close to horizontal); Ch = R2

When the absolute magnitude of B is not larger than the absolute magnitude of R2, the absolute value of B is not less than 0.009, and the difference between the absolute magnitude of B and the absolute magnitude of R2 is less than 0.017; Ch = B * 0.2 + R2 * 0.8

When the absolute magnitude of B is not greater than the absolute magnitude of R2, the absolute value of B is not less than 0.009, and the difference between the absolute magnitude of B and the absolute magnitude of R2 is not less than 0.017; Ch = B * 0.5 + R2 * 0.5

Calculate

The steps for obtaining the horizontal center symbol gradient initial value Ch for the above steps are generally described as follows.

When the magnitude of the absolute value of B is greater than the magnitude of the absolute value of R2 and the absolute value of B is less than the first horizontal tolerance Gh1 (close to vertical); Ch = B

The absolute magnitude of -B is greater than the absolute magnitude of R2, the absolute value of B is not less than the horizontal tolerance Gh1 (not vertical), and the difference between the absolute magnitude of B and the absolute magnitude of R1 is within Gh3. ; Ch = B * 0.8 + R2 * 0.2

When the absolute magnitude of B is greater than the absolute magnitude of R2, the absolute value of B is not less than the horizontal tolerance Gh1 (not vertical), and the absolute magnitude of B and the absolute magnitude of R2 are greater than or equal to Gh3; Ch = B * 0.5 + R1 * 0.5

When the absolute value magnitude of -B is not large relative to the absolute value magnitude of R2, and the absolute value of B is smaller than the second horizontal tolerance value Gh2 (close to vertical); Ch = R2

When the absolute magnitude of B is not greater than the absolute magnitude of R2, the absolute value of B is not less than the second vertical tolerance Gh2, and the difference between the absolute magnitude of B and the absolute magnitude of R2 is less than Gh3; Ch = B * 0.2 + R2 * 0.8

When the absolute magnitude of -B is not larger than the absolute magnitude of R2, the absolute value of B is not less than the second vertical tolerance Gh2, and the difference between the absolute magnitude of B and the absolute magnitude of R2 is not less than Gh3. ; Ch = B * 0.5 + R2 * 0.5

Calculate

When the horizontal center symbol inclination initial value Ch is obtained by the above process, the controller 160 extracts coordinates of the symbols from the bar code image in the horizontal center direction. When the horizontal center symbol gradient initial value Ch is obtained, the controller 160 divides the lengths of both end coordinates of the right finder pattern 103 by the number of symbols and linear equation of the vertical center symbol slope initial value Ch that passes through the coordinates of the symbol located at the center. And retrieve the number of symbols while moving the straight line in the upper direction (-y direction) and store the coordinate information of the horizontal center symbols in the array when the number of symbols coincides with the number of symbols placed in the upper horizontal direction of the barcode.

24) barcode reading initialization step (S1220)

When the central horizontal symbol coordinate extraction step (S1210) is completed, the controller 160 includes symbols placed in a right vertical direction of a barcode from a binary barcode image of the second frame memory 150 through a symbol calculator 250; Symbols placed at the center vertical position of the barcode; Since the coordinates of the symbols placed in the upper direction of the barcode; the symbols placed in the center horizontal position of the barcode; are extracted, and the initialization step S1220 for reading the barcode is performed therefrom. In the barcode reading initialization step (S1220), the coordinates of the first symbol placed in the right vertical direction of the barcode among the coordinates of the extracted symbols are V1, and the coordinates of the first symbol placed in the center vertical position of the barcode are placed in the horizontal direction of the top of the barcode. The reading step is started by setting an initial value in which the coordinate of the first symbol is H1 and the coordinate of the first symbol placed at the center horizontal position of the barcode is H2.

25) step of creating a linear equation between vertical symbols (S1230)

Subsequently, the controller 160 obtains an equation F1 (y) of a straight line connecting the vertical symbol coordinates V1 and V2 in the second frame memory 150.

26) Step of creating a linear equation between horizontal symbols (S1240)

Subsequently, the control unit 160 obtains the equation F2 (y) of the straight line connecting the horizontal symbol coordinates H1 and H2 in the second frame memory 150.

27) barcode pixel reading step (S1250)

Subsequently, the control unit 160 calculates coordinates of the second frame memory 150 where F1 (y) and F2 (y) intersect and read pixel values of the coordinates. do.

28) 2D matrix memory storage step (S1260)

The pixel value of the corresponding coordinate read in the bar code pixel reading step S1250 is stored in the 2D matrix memory 260 by black = '1' and white = '0' by the controller 160. In this case, the 2D matrix memory 260 is configured to correspond to the horizontal and vertical number of the symbol of the barcode to be read, and is configured such that one symbol is arranged at one address.

At this time, if the binary image of the coordinates of the second frame memory 150 in which the linear equations of F1 (y) and F2 (y) intersect in the horizontal-vertical direction is read as black, the controller 160 corresponds to the corresponding 2D matrix. If the binary image of the coordinates of the second frame memory 150 that intersects in the horizontal-vertical direction is read out as white in the memory 260, the controller 160 controls the corresponding 2D matrix memory 260. Record '0' at the address.

29) Horizontal Symbol End Check Step (S1270)

Following the above 2D matrix memory storing step (S1260), the controller 160 controls the horizontal symbol coordinates H1 in the second frame memory 150 used for reading as the linear equations of F1 (y) and F2 (y). Determine whether H2 is the last symbol.

30) horizontal symbol coordinate update step (S1280)

If it is determined that the horizontal symbol coordinates H1 and H2 in the second frame memory 150 are not the last symbols in the horizontal symbol end check step (S1270), the controller 160 adjusts the horizontal symbol coordinates H1 and H2 to the upper horizontal level. Refer to the horizontal symbol coordinates extracted in the symbol coordinate extraction step (S1200) and the central horizontal symbol coordinate extraction step (S1210), the next time that is placed in the top horizontal direction of the barcode H1, the next time that is placed in the center horizontal position of the barcode After updating the coordinates of the symbol to H2, the next straight line equation F2 of the horizontal symbols in the second frame memory 150 is connected to the next straight line equation H2 and H2 in the second frame memory 150. Repeat from step (y). When the horizontal symbol coordinates H1 and H2 in the second frame memory 150 are the last symbols in the horizontal symbol end check step S1270, the next vertical symbol end check step S1310 is performed.

31) Vertical Symbol End Check Step (S1310)

If it is determined that the horizontal symbol coordinates H1 and H2 in the second frame memory 150 are the last symbols in the horizontal symbol end check step (S1270), the control unit 160 controls F1 (y) and F2 (y). It is determined whether the vertical symbol coordinates V1 and V2 in the second frame memory 150 used for reading are the last symbols by the straight line equation.

32) Vertical symbol coordinate update step (S1280)

If it is determined that the vertical symbol coordinates V1 and V2 in the second frame memory 150 are not the last symbols in the vertical symbol end check step (S1310), the controller 160 controls the vertical symbol coordinates V1 and V2 to the right symbol. Coordinates of the next symbol placed in the right vertical direction of the barcode with reference to the vertical symbol coordinates extracted in the coordinate extracting step (S1180) and the center vertical symbol coordinate extracting step (S1190) of V1, the next symbol placed at the center vertical position of the barcode Update the coordinates of V2 to V2.

33) Horizontal symbol coordinate initialization step (S1300)

Following the vertical symbol coordinate updating step (S1280), the control unit 1160 and the horizontal symbol coordinate H1 in the second frame memory 150 used for reading by the linear equation of F1 (y) and F2 (y) and Update H2 to the coordinates of the first symbol and jump to the step of creating a linear equation between vertical symbols (S1230), and the next equation F1 (y of the straight line connecting the vertical symbol coordinates V1 and V2 in the second frame memory 150) Repeat from step to find).

34) Barcode Decoding Step (S1320)

If it is determined that the vertical symbol coordinates V1 and V2 are the last symbols in the vertical symbol end check step (S1310), all barcode pixels in the second frame memory 150 are read and stored in the 2D matrix memory 260. The controller 160 provides barcode data consisting of '1' and '0' of the 2D matrix memory 260 to the 2D matrix reading unit 160a to perform decoding and ends (S1330).

As described above, although the present invention has been described by way of limited embodiments and drawings, the present invention is not limited thereto and will be described by the person skilled in the art to which the present invention pertains. Various modifications and variations are possible without departing from the scope of the appended claims.

100: barcode 101: 2X2 data matrix
103: Left finder pattern 104: Lower finder pattern
110: camera 120: first frame memory
130: display 140: DMA controller
150: second frame memory 160: control unit
160a: two-dimensional matrix reading unit 170: arithmetic frame memory
180: gray level conversion unit 190: interpolation reduction unit
220: connection area extraction unit 230: coordinate extraction unit
240: finder pattern detection / rotation unit 250: symbol operation unit
260: 2D matrix memory

Claims (23)

A skewed data matrix barcode decoding apparatus from a camera for capturing an image, the apparatus comprising:
The camera 110 photographing the barcode 100, the first frame memory 120 storing the color image data from the camera 110, and the display 130 outputting the image data of the first frame memory 120. A terminal consisting of:
A controller 160 for controlling data movement between frame memories and individual processing units and finally reading a data matrix bar code; and
A second 2D matrix reading unit for finally reading a 2D matrix barcode of the 2D matrix memory 260 in which the barcode reading result in the second frame memory 150 is stored as '0' and '1'. 160a);
The control unit 160,
A DMA controller (140) for storing image data from the camera (110) stored in the first frame memory (120) described by the transfer command of the controller (160) in the second frame memory (150);
A second frame memory 150 for storing image data of the first frame memory 120;
An arithmetic frame memory 170 for arithmetic processing data of the second frame memory 150;
A gray level converter 180 for converting an image from the camera 110 stored in the second frame memory 150 to a gray level;
A dual linear interpolation unit 190 performing an operation for reducing and storing the image data of the second frame memory 150 converted to gray level in the operation frame memory 170;
An area filter unit 200 for calculating an edge value with respect to the gray level image in the calculation frame memory 170;
A binary image converter 210 for converting gray level image data of the operation frame memory 170 or the second frame memory 150 into binary images of '1' and '0';
Determining adjacent 'white' pixel groups with respect to the binary image in the operation frame memory 170 and labeling by assigning a unique number for each group, 'white' pixel groups divided by the labeling (labeling) A connection region extraction unit 220 for extracting a 'white' pixel group composed of the most pixels among the pixels;
The candidate region of the barcode extracted by replacing the pixel values of the remaining groups except for the 'white' pixel group composed of the most pixels for the binary image in the operation frame memory 170 with 'black' Calculate the gradient (vertical gradient: Gradient Vert) between the vertically spaced center coordinates by determining the location of the largest region with respect to Calculate the horizontal gradient Gradient Horz,
A slope including all the symbols placed at the right vertical position in the barcode image in the second frame memory 150 processed by the pinter pattern detection / rotation unit 240, a slope including all the symbols placed at the center vertical position of the barcode, A peripheral gradient extracting unit 230 for calculating a slope including all the symbols placed at the top horizontal position of the barcode and a slope including all the symbols placed at the horizontal center position of the barcode;
The finder pattern and its position are detected from the barcode image in the second frame memory 150, and based on this, the control unit rewrites the barcode image in the second frame memory 150 into image data obtained by rotating the barcode image by 90 degrees. In the second frame memory 150, the left finder pattern 103 and the lower finder pattern 104 are located at the left and the bottom of the image, respectively, and the size of the basic symbol for reading the barcode pattern at the rightmost and uppermost. A finder pattern detection / rotation unit 240 that calculates a clock track 102 to be divided into '0' and '1';
When the right vertical slope is obtained with respect to the position of the symbols searched in the bar code image in the second frame memory 150 processed by the pinter pattern detection / rotation unit 240, the array in which the retrieved symbol position information is stored and the center vertical slope are obtained. When you find the linear equation that consists of two points stored at the same index of the array where the retrieved symbol position information is stored, and when you obtain the top horizontal slope, the array where the retrieved symbol position information is stored and the same position of the array where the retrieved symbol position information is stored A symbol operation unit 250 for obtaining a straight line equation formed by two points stored in an index, and storing '1' if the pixel value corresponding to the intersection point of the two straight lines is black and '0' if the white color is 2D;
A 2D matrix memory 260 which stores the barcode reading result in the second frame memory 150 as '0' and '1' by the symbol calculating unit 250;
Skew data matrix barcode decoding device characterized in that the configuration is connected
The method of claim 1, wherein the area filter unit 200,
Receive the reduced image data stored in the operation frame memory 170 from the control unit 160 to calculate an edge value of the image,
The area filter unit 200 allocates a memory having a size of 3 X 3 pixels to be used as a mask for the gray level image in the operation frame memory 170,
The maximum and minimum values are obtained by reading three pixel values horizontally and vertically around the current pixel position, and the difference is calculated as an edge value.
Skew data matrix barcode decoding device characterized in that configured to store the image consisting of the edge value in the operation frame memory 170 when the edge value calculation is completed for the entire reduced image data
The peripheral slope extraction unit 230 of claim 1,
Receive image data of the operation frame memory 170 including the candidate region of the 'white' bar code extracted from the controller 160,
Two boundary coordinates spaced apart by d from the center in the vertical direction of the image data of the operation frame memory 170, the vertical height and the slope (vertical slope: Gradient_Vert) are further calculated, and are horizontally spaced about the center. Skew data matrix barcode decoding device, characterized in that configured to further calculate two boundary coordinates, horizontal length and slope (horizontal slope: Gradient_Horz)
The peripheral slope extraction unit 230 of claim 1,
Extract two boundary coordinates spaced d by the center in the vertical direction (Y axis) of the image data of the operation frame memory 170, and two spaced apart by the center d in the horizontal direction (x axis) A skewed data matrix barcode decoding device, configured to extract boundary coordinates and calculate a length and a slope of candidate regions of a barcode using the boundary coordinates, respectively.
The method of claim 1, wherein the controller 160,
The bar code reading is effective by using the vertical height Labeling_height and the horizontal length Labeling_width of the barcode candidate area provided by the connection area extractor 220 and the vertical gradient Gradient_Vert and the horizontal gradient Gradient_Horz calculated by the peripheral gradient extractor 230. Skew data matrix barcode decoding device, characterized in that configured to determine whether the image
The method of claim 5, wherein the control unit 160,
If the ratio of the vertical height Labeling_height and the horizontal length Labeling_width of the barcode candidate region stored in the first frame memory 120 is within a certain ratio, it is determined that the reading is possible. Skew data matrix barcode decoding device characterized in that configured to store the new image data from the camera 110 from the first frame memory 120 to the second frame memory 150 by a transmission command of
The method of claim 5, wherein the control unit 160,
The vertical gradient Gradient_Vert and horizontal gradient Gradient_Vort values calculated by the peripheral gradient extractor 230 are close to vertical, the horizontal gradient Gradient_Horz value is close to horizontal, or the vertical gradient Gradient_Vert and horizontal gradient Gradient_Horz values When it is close to each other, that is, when the product of the vertical gradient Gradient_Vert and the horizontal gradient Gradient_Horz is less than 0, it is determined that the reading is possible. Otherwise, the recognition failure is regarded as a recognition failure. Skew data matrix barcode decoding device, characterized in that it is configured to store new image data from the first frame memory 120 to the second frame memory 150
In claim 1, the symbol operation unit 250,
Coordinates of the upper right end corner of the second frame memory 150 including the image based on the lower right coordinates (vert_right (x, y)) 152 of the binary barcode image of the second frame memory 150. The coordinates (vert_top (x, y)) 151 of the upper right end corner in the origin direction of the second frame memory 150 starting with the straight line of the slope, are obtained in the origin direction. Skew data matrix barcode decoding device characterized in that it is configured to obtain a slope including all the symbols placed in the right vertical direction of the barcode while moving by one pixel
According to claim 8, If the symbol operation unit 250 obtains the slope including all the symbols placed in the right vertical direction of the barcode control unit 160,
Based on the inclination value of the left finder pattern 103 and the slope including all the symbols placed in the right vertical direction of the barcode.
Skew data matrix barcode decoding device characterized in that it is configured to obtain a vertical center symbol gradient initial value (Cv) including all the symbols placed at the center vertical position of the barcode.
The control unit 160 of claim 9,
The slope of the left and right finder patterns 103 and the vertical center symbol tilt initial value (Cv) are taken into consideration for the length of the y-axis direction in consideration of the significant digits of the binary operation. Skew data matrix barcode decoding device, characterized in that configured to calculate in terms of the length in the axial direction
The method of claim 9 or 10, wherein the control unit 160,
If the initial value of the vertical center symbol slope is Cv, the slope including all the symbols placed in the right vertical direction of the barcode is R1, and the slope value of the left finder pattern 103 is L,
When the magnitude of the absolute value of L is larger than the magnitude of the absolute value of R1 and the absolute value of L is smaller than the first vertical allowable value Gv1 (close to the vertical); Cv = L
Cv when the absolute magnitude of L is greater than the absolute magnitude of R1, the absolute value of L is not less than the vertical tolerance Gv1 (not vertical), and the difference between the absolute magnitude of L and the magnitude of R1 is within Gv3. = L * 0.8 + R1 * 0.2
When the absolute magnitude of L is greater than the absolute magnitude of R1, the absolute value of L is not less than the vertical tolerance Gv1 (not vertical), and the absolute magnitude of L and the absolute magnitude of R1 are greater than or equal to Gv3, Cv = L * 0.5 + R1 * 0.5
When the magnitude of the absolute value of L is not large compared to the magnitude of the absolute value of R1, and the absolute value of L is smaller than the second vertical tolerance value Gv2 (close to the vertical); Cv = R1
When the magnitude of the absolute value of L is not larger than the magnitude of the absolute value of R1, the absolute value of L is not less than the second vertical tolerance Gv2, and the difference between the absolute magnitude of L and the magnitude of the absolute value of R1 is less than Gv3, Cv = L * 0.2 + R1 * 0.8
Cv when the magnitude of the absolute value of L is not larger than the magnitude of the absolute value of R1, the absolute value of L is not less than the second vertical tolerance Gv2, and the difference between the absolute value of L and the magnitude of the absolute value of R1 is not less than Gv3. = L * 0.5 + R1 * 0.5
Skew data matrix barcode decoding device, characterized in that configured to be determined
The control unit 160 of claim 9,
Obtain the vertical center symbol slope initial value Cv and divide the length of both end coordinates of the lower finder pattern 104 by the number of symbols to obtain a straight line equation of the vertical center symbol slope initial value Cv passing the coordinates of the symbol located at the center, Squeeze data matrix barcode decoding device, characterized in that to retrieve the number of symbols while moving in the right direction and to store the coordinate information of the vertical center symbols in the array when it matches the number of symbols placed in the right vertical direction of the barcode
In claim 1, the control unit 160,
The skewed data matrix barcode decoding device of claim 2, wherein the binary barcode image of the second frame memory 150 is provided to the symbol operator 250 to obtain a slope including all the symbols placed in the horizontal direction of the barcode.
The symbol operator 250 of claim 13,
The slope including all the symbols placed in the upper horizontal direction of the binary barcode image of the second frame memory 150 is based on the end coordinates (vert_left (x, y)) 153 of the upper left of the binary barcode image. The upper right end of the second frame memory 150 in the origin direction of the second frame memory 150, starting with a straight line of the slope formed with the coordinates (vert_top (x, y)) 151 of the upper right end corner of the second frame memory 150 including. Skew data matrix barcode decoding device characterized in that it is configured to obtain the slope including the maximum number of symbols while moving the coordinates (vert_top (x, y)) 151 of the corner by one pixel in the y-axis direction
The controller 160 of claim 13, wherein when the symbol operation unit 250 obtains an inclination including all of the symbols placed in the horizontal direction of the upper side of the binary barcode image, the controller 160 determines whether the symbol operation unit 250 includes the symbols.
Based on the inclination value of the lower finder pattern 104 and the slope including all the symbols placed in the upper direction of the barcode.
Skew data matrix barcode decoding device, characterized in that configured to obtain a horizontal center symbol gradient initial value (Ch) including all the symbols placed at the center horizontal position of the barcode
The method of claim 15, wherein the control unit 160,
The inclination value of all the symbols placed in the upper direction of the barcode and the inclination value of the lower finder pattern 104 and the initial value of the horizontal center symbol inclination Ch are taken into consideration for the length of the x-axis direction in consideration of the significant digits of the binary operation. Skew data matrix barcode decoding device, characterized in that configured to calculate in terms of the length in the axial direction
The method according to claim 15 or 16, wherein the control unit 160,
If the initial value of the horizontal center symbol slope is Ch, the slope including all the symbols placed in the upper horizontal direction of the barcode is R2, and the slope value of the lower finder pattern 104 is B,
When the magnitude of the absolute value of B is larger than the magnitude of the absolute value of R2, and the absolute value of B is smaller than the first horizontal tolerance Gh1 (close to vertical); Ch = B
When the absolute magnitude of B is greater than the absolute magnitude of R2, the absolute value of B is not less than the horizontal tolerance Gh1 (not vertical), and the difference between the absolute magnitude of B and the magnitude of R1 is within Gh3. = B * 0.8 + R2 * 0.2
When the absolute magnitude of B is greater than the absolute magnitude of R2, the absolute value of B is not less than the horizontal tolerance Gh1 (not vertical), and the absolute magnitude of B and the absolute magnitude of R2 are greater than or equal to Gh3, Ch = B * 0.5 + R1 * 0.5
When the absolute value magnitude of B is not large relative to the absolute value magnitude of R2, and the absolute value of B is smaller than the second horizontal tolerance value Gh2 (close to vertical); Ch = R2
When the absolute magnitude of B is not greater than the absolute magnitude of R2, the absolute value of B is not less than the second vertical tolerance Gh2, and the difference between the absolute magnitude of B and the absolute magnitude of R2 is less than Gh3, Ch = B * 0.2 + R2 * 0.8
When the absolute magnitude of B is not larger than the absolute magnitude of R2, the absolute value of B is not less than the second vertical tolerance Gh2, and the difference between the absolute magnitude of B and the absolute magnitude of R2 is not less than Gh3. = B * 0.5 + R2 * 0.5
Skewed data matrix barcode decoding device characterized in that it is configured to determine the horizontal center symbol gradient initial value Ch
The method of claim 15, wherein the control unit 160,
The horizontal center symbol slope initial value Ch is calculated, and the controller 160 divides the lengths of both end coordinates of the right finder pattern 103 by the number of symbols. The linear equation of the vertical center symbol slope initial value Ch passes through the coordinates of the symbol located at the center. To obtain the number of symbols while moving the straight line in the upper direction (-y direction) and store the coordinate information of the horizontal center symbols in an array when the number of symbols coincides with the number of symbols placed in the upper horizontal direction of the barcode. Skewed data matrix barcode decoding device
The method of claim 1, wherein the controller 160,
From the binary barcode image of the second frame memory 150 through the symbol operation unit 250, the symbols placed in the right vertical direction of the barcodes. The symbols placed in the center vertical position of the barcodes. Obtain the coordinates of the symbols placed in the horizontal position and extract the horizontal linear equations that sequentially connect the coordinates of the symbols placed in the right vertical direction of the barcode of the second frame memory 150 and the coordinates of the symbols placed in the center vertical position of the barcode. Then, skewed data matrix barcode decoding device characterized in that it is configured to extract the linear equations in the vertical direction connecting the symbols placed in the upper direction of the barcode and the coordinates of the symbols placed in the center horizontal position of the barcode in sequence.
The method of claim 19, wherein the controller 160,
Using the above linear equations, the coordinates of the second frame memory 150 intersecting in the horizontal-vertical direction are calculated for the respective straight lines, and the barcode is completed in the 2D matrix memory 260 using the image data of the corresponding coordinates. Skew data matrix barcode decoding device, characterized in that configured to
The method of claim 19, wherein the control unit 160,
When the binary image of the coordinates of the second frame memory 150 intersecting in the horizontal-vertical direction is read as black using the above linear equations, the controller 160 sets '1' to the corresponding 2D matrix memory 260 address. When the binary image of the coordinates of the second frame memory 150 that intersects in the horizontal-vertical direction is read as white, the controller 160 records a series of '0's at the corresponding 2D matrix memory 260 addresses. Skew data matrix barcode decoding device, characterized in that configured to complete the 2D barcode in the 2D matrix memory 260 by performing the process for all straight lines
The method of claim 21, wherein the controller 160,
Skew data matrix barcode decoding device characterized in that the barcode data consisting of '1' and '0' in the 2D matrix memory 260 is configured to decode the barcode through the 2D matrix reading unit 160a having the barcode decoding means.
A method of driving an apparatus for decoding a skewed data matrix barcode from a camera taking an image, the method comprising:
Start step (S1000) that the two-dimensional bar code extraction is started;
After the start step (S100), the controller 160 transmits a command to the direct memory access (DMA) controller 140 to display image data from the camera 110 stored in the one-frame memory 120. An input image copying step (S1010) to be stored at 150;
A gray level converting step (S1020) of the control unit 160 converting the color image of the second frame memory 150 to the gray level through the gray level converting unit 180 after the input image copying step (S1010);
Following the gray level conversion step (S1020), the control unit 160 reduces and stores the image data of the second frame memory 150 converted to gray levels by the dual linear interpolation unit 190 to the operation frame memory 170. Double linear interpolation step S1030;
An operation frame memory storage step (S1040) of storing the image data of the reduced second frame memory 150 in the operation frame memory 170 by the dual linear interpolation step (S1030);
After the operation frame memory 170 storage step (S1040), the controller 160 calculates an edge value of an image by providing the reduced image data stored in the operation frame memory 170 to the area filter unit 200. Edge extraction step (S1050);
After the edge extraction step (S1050), the controller 160 provides an edge value image data of the reduced gray level image data stored in the operation frame memory 170 to the binary image converter 210 to convert the image into a binarized image. Extraction step (S1060);
After the binarization image extraction step (S1060), the controller 160 provides the edge region binarization image data of the reduced image data stored in the operation frame memory 170 to the connection area extraction unit 220, so that the adjacent ' White 'pixel groups are determined and a label is assigned by assigning a unique number to each group, and a' white 'pixel group composed of the most pixels among the' white 'pixel groups classified by the labeling. Extracting the maximum connection area (S1070);
Following the maximum connection region extraction step (S1070), the controller 160 provides the peripheral gradient extracting unit 230 with image data of the operation frame memory 170 including the candidate region of the extracted 'white' barcode. In the vertical direction of the image data of the operation frame memory 170, the vertical height and the slope (vertical slope: Gradient_Vert) are calculated about the center, and the horizontal length and the slope (horizontal slope: Gradient_Horz) about the center in the horizontal direction. A boundary coordinate gradient extraction step (S1080) of calculating);
Following the step of extracting the boundary coordinate gradient (S1080), the controller 160 has a height of the labeling_height and a length of the labeling_width of the barcode candidate region stored in the first frame memory 120 provided by the connection region extracting unit 220. A horizontal vertical length validity determining step (S1090), in which the ratio is determined to be readable when the ratio is within a predetermined ratio;
If it is determined in the horizontal vertical length validity determination step (S1090) that the reading is possible by the control unit 160, the control unit 160 determines whether the vertical gradient Gradient_Vert calculated by the peripheral gradient extraction unit 230 is a valid value for reading the barcode. Determining a vertical slope valid step (S1091);
If the vertical gradient Gradient_Vert calculated by the peripheral gradient extracting unit 230 is determined to be a valid bar code reading in the vertical gradient valid determination step (S1091), the controller 160 determines the peripheral gradient extracting unit 230. A horizontal gradient validity determining step (S1092) for determining whether the horizontal gradient Gradient_Horz calculated by H) is a valid value;
When the horizontal gradient Gradient_Horz is valid in the horizontal slope valid determination step (S1092), the controller 160 determines whether the vertical gradient Gradient_Vert and the horizontal gradient Gradient_Horz values calculated by the peripheral gradient extractor 230 are close to the vertical. Calculating the angle of the vertical slope (S1100); Lt; / RTI &gt;
Following the angle calculating step of the horizontal vertical tilt (S1100), the control unit 160 determines an angle of the horizontal vertical tilt determining that the product can be read when the product of the vertical gradient Gradient_Vert and the horizontal gradient Gradient_Horz is smaller than 0 (S1110);
When the control unit 160 determines that the barcode candidate region provided by the connection region extracting unit 220 is capable of reading the barcode in the step S1110 of determining the angle of horizontal and vertical tilting, the control unit 160 controls the operation frame memory 170. Performing a bar code region extraction step (S1120) of replacing a pixel value of an area excluding the bar code candidate region with white in the image of the second frame memory 150 corresponding to the bar code candidate region of S1120;
Following the bar code area extraction step (S1120), the control unit 160 includes: an extraction area valid determination step (S1130) for determining that the bar code candidate area of the second frame memory 150 can be read if the size is greater than or equal to a predetermined value;
If it is determined that the size of the barcode candidate region of the second frame memory 150 is readable in the extraction region valid determination step (S1130), the controller 160 corresponds to a gray corresponding to the barcode candidate region of the second frame memory 150. Binary step of extracting the extracted region image to store the binary image corresponding to the barcode candidate region in the second frame memory 150 by supplying the level image to the binary image converter 210 to perform binary image conversion (S1140).
Following the binarization of the extracted region image (S1140), the controller 160 detects the finder pattern by providing the binary image to the finder pattern detection / rotation unit 240 (S1150);
An image rotation step of positioning the binary data matrix barcode image in the normal direction by rotating the binary image in units of 90 degrees according to the finder pattern search step (S1150);
Following the image rotation step (S1160), the control unit 160 has a left finder pattern 103 and a lower finder pattern 104 positioned in an 'L' shape on the left and lower sides of the binary barcode image of the second frame memory 150. A finder pattern slope and coordinate extraction step of calculating a slope thereof and both end coordinates of each pattern (S1170);
Following the finder pattern tilt and coordinate extraction step (S1170), the control unit 160 provides a binary barcode image of the second frame memory 150 to the symbol operation unit 250 to extract the coordinates of the symbol. S1180);
Following the right symbol coordinate extracting step (S1180), the controller 160 extracts a center vertical symbol coordinate extracting step (S1190) of obtaining a slope including all the symbols placed at the center vertical position of the barcode through the symbol calculating unit 250; and,
Following the central vertical symbol coordinate extraction step (S1190), the controller 160 provides the binary barcode image of the second frame memory 150 to the symbol operator 250 to extract the coordinates of the upper horizontal symbol of the barcode image. Extracting upper horizontal symbol coordinates (S1200);
A central horizontal symbol coordinate extraction step (S1210) of which the control unit 160 obtains a slope including all the symbols placed at the central horizontal position of the barcode through the symbol operation unit 250 after the upper horizontal symbol coordinate extraction step (S1200);
When the central horizontal symbol coordinate extraction step (S1210) is completed, the controller 160 determines the coordinate of the first symbol placed in the right vertical direction of the barcode V1, the coordinate of the first symbol placed in the center vertical position of the barcode V2, and the top of the barcode. A barcode reading initialization step (S1220) in which the coordinates of the first symbol placed in the horizontal direction are set to H1 and the coordinates of the first symbol placed at the center horizontal position of the barcode are set to H2;
Following the bar code reading initialization step (S1220), the control unit 160 creates a straight line-to-symbol straight line equation for obtaining a straight line equation F1 (y) connecting the vertical direction coordinates V1 and V2 in the second frame memory 150. Step S1230; and
In the step S1230 of creating a vertical equation between the vertical symbols, the control unit 160 calculates a horizontal symbol symbol for obtaining a linear equation F2 (y) connecting the horizontal symbol coordinates H1 and H2 in the second frame memory 150. Line equation creation step (S1240);
Following the straight line inter-symbol creation step (S1240), the controller 160 calculates coordinates of the second frame memory 150 where F1 (y) and F2 (y) intersect, and reads pixel values of the coordinates. A barcode pixel reading step S1250;
The 2D matrix memory storage step of storing the pixel values of the corresponding coordinates read in the barcode pixel reading step S1250 in the 2D matrix memory 260 with black = '1' and white = '0' by the controller 160. (S1260);
Following the above 2D matrix memory storing step (S1260), the controller 160 controls the horizontal symbol coordinates H1 in the second frame memory 150 used for reading as a linear equation of F1 (y) and F2 (y). A horizontal symbol end check step (S1270) to determine whether H2 is the last symbol;
When the horizontal symbol coordinates H1 and H2 are the last symbols in the horizontal symbol end check step (S1270), the control unit 160 uses the second equation used for reading as the linear equation of F1 (y) and F2 (y). A vertical symbol end check step (S1310) of determining whether the vertical symbol coordinates V1 and V2 in the frame memory 150 are the last symbols;
If it is determined that the vertical symbol coordinates V1 and V2 are the last symbols in the vertical symbol end check step (S1310), all barcode pixels in the second frame memory 150 are read and stored in the 2D matrix memory 260. The control unit 160 provides a bar code decoding step (S1320) of performing decoding by providing barcode data composed of '1' and '0' of the 2D matrix memory 260 to the 2D matrix reading unit 160a;
Method of driving an apparatus for decoding a skewed data matrix bar code, characterized in that configured to terminate through (S1330)

Images