JP5145833B2 - Two-dimensional code reading device, two-dimensional code reading method, two-dimensional code reading program, and recording medium - Google Patents

Description

  The present invention relates to a two-dimensional code reading device, a two-dimensional code reading method, a two-dimensional code reading program, and a recording medium. In particular, when a multi-valued input image is binarized and decoded, The present invention relates to a two-dimensional code reading apparatus, a two-dimensional code reading method, a two-dimensional code reading program, and a recording medium that perform decoding after changing binarization parameters according to the ratio of monochrome pixels and binarizing an appropriate image again .

  Conventionally, a document processing system, a document processing method, and a document processing program for decoding a binary image input by a scanner have been proposed (see, for example, Patent Document 1). The printed document input by the scanner is binarized using the binary binarization function of the scanner, and the obtained binary image is decoded only once. .

  Further, as a technique related to binarization, an image binarization apparatus, an image imaging apparatus, and the like have been proposed (see, for example, Patent Document 2). The binarization threshold value is locally changed according to the color and brightness of the character background to properly binarize the document. This allows various input conditions such as a light source as well as various scanners. Even if a document is photographed using a digital camera in a non-constant photographing environment under a light source, characters, barcodes, and two-dimensional codes on the document can be correctly binarized.

By using the technique as described above, it becomes easy to binarize an image captured by a digital camera and decode a two-dimensional code.
JP 2005-122682 A JP 2001-008032 A

  However, the above techniques have the following problems. In order to print a two-dimensional code on a paper document, the two-dimensional code is usually printed by using a general-purpose printer such as an electrophotographic printer or an inkjet printer, instead of printing by a bar code printer such as a label printer. become. Development of electrophotographic printers has been focused on beautifully printing document text using toner using carbon black. As a result, in order to dislike the fading of characters, etc., we have devised measures such as printing thin lines thickly. For this reason, there is a disadvantage that the black portion is expanded when a two-dimensional code or barcode is printed.

  In addition, in an ink jet printer, the ink does not go to the target location and the ink scatters near the black-and-white boundary, or the ink penetrates into the paper and the ink spreads, or the concentration of pigment ink, dye ink, etc. differs. Ink is used, and in the high-quality mode and draft mode, the amount of ink used is different and there is a considerable difference in print density between modes. In other words, even if the two-dimensional code or bar code is printed on the same paper under different printing conditions, the image is input under the same conditions, and the binary image that is output even if the image binarization is performed with the same parameters is expanded in black As a result, white pixels are mixed in the black portion of the code, or the area of the black region is reduced. For this reason, even when a two-dimensional code or barcode is decoded, it cannot always be correctly decoded.

  The present invention has been made in view of such a situation. In the case of decoding failure of a two-dimensional code, the ratio of black pixels to white pixels is calculated, and binarization parameters are adjusted based on the ratio, and the image is again displayed. An object of the present invention is to generate a binary image that can be appropriately decoded by performing binarization, and to execute a decoding process.

In order to achieve the above object, the present invention provides a first aspect as follows:
The multi-valued image including the region of the two-dimensional code having a portion of the cell Le plurality of white and black squares as a unique pattern, the following equation
Cb = x / 16
Where Cb: multiplication factor
x: integer, 1 ≦ x ≦ 15
An image binarization means binarizes and outputs the binary image based on,
Two-dimensional code detection means for detecting a region of the two-dimensional code from the binary image;
A unique pattern detecting means for detecting a unique pattern from the region of the two-dimensional code;
A decoding means for decoding the pre-SL 2-dimensional code, a,
The unique pattern detection means, among the unique pattern of black or white constituting the pattern, the cells constituting one pattern, with the proviso that match all the cells constituting the predetermined pattern Determining means for determining a condition of 1 and a second condition on the condition that a matching degree of the other pattern satisfies a predetermined matching degree;
The binarization means includes
Meet black pattern matching score of the first condition, if the match degree of the white pattern does not satisfy the second condition, the degree of matching of the white pattern to the second condition is satisfied Re-binarize multi-valued image,
Coincidence of the white pattern satisfies the first condition, when the degree of matching of the black pattern does not satisfy the second condition, the degree of matching of the black pattern to the second condition is satisfied Binarize the multi-valued image again,
It detects the leading SL again binarizing the binary image or al the unique pattern, to provide a two-dimensional code reading apparatus characterized by decoding the two-dimensional code in the decoding means.

In order to achieve the above object, the present invention provides a second aspect,
A multi-valued image including a two-dimensional code region having a part of a plurality of white and black rectangular cells as a unique pattern is expressed by the following formula:
Cb = x / 16
Where Cb: multiplication factor
x: integer, 1 ≦ x ≦ 15
An image binarization step for binarizing based on the output and outputting a binary image;
A two-dimensional code detection step of detecting a region of the two-dimensional code from the binary image;
A unique pattern detecting step for detecting a unique pattern from the region of the two-dimensional code;
And a decoding step for decoding the two-dimensional code,
The unique pattern detecting step is performed on the condition that, of the black or white patterns constituting the unique pattern, the cells constituting one pattern coincide with all the cells constituting the predetermined pattern. A determination step of determining a condition of 1 and a second condition on the condition that a matching degree of the other pattern satisfies a predetermined matching degree;
The binarization step includes
When the matching degree of the black pattern satisfies the first condition and the matching degree of the white pattern does not satisfy the second condition, the matching degree of the white pattern satisfies the second condition. Re-binarize multi-valued image,
When the matching degree of the white pattern satisfies the first condition and the matching degree of the black pattern does not satisfy the second condition, the matching degree of the black pattern satisfies the second condition. Binarize the multi-valued image again,
A two-dimensional code reading method is provided, wherein the unique pattern is detected from the binarized binary image again, and the two-dimensional code is decoded in the decoding step .

In order to achieve the above object, the present invention provides, as a third aspect, a two-dimensional code reading program characterized by causing a computer to execute the two-dimensional code reading method according to the second aspect .
As a fourth aspect, there is provided a computer-readable recording medium in which the program according to the third aspect is recorded.

  According to the present invention, when the decoding of the two-dimensional code fails, the ratio of the black pixel and the white pixel is calculated, the binarization parameter is adjusted based on the ratio, and the image binarization is performed again, so that it can be decoded appropriately. A binary image can be generated and a decoding process can be executed.

  Embodiments of the present invention will be described below in detail with reference to the drawings. The embodiments described below are preferred embodiments of the present invention, and thus various technically preferable limitations are given. However, the scope of the present invention is particularly limited in the following description. As long as there is no description of the effect, it is not restricted to these aspects.

  First, the “two-dimensional code” used in the present embodiment will be described below. The basic of this two-dimensional code is a rectangle, but in order to maintain high reading performance even when reading using a camera from diagonally forward, in such a case, the two-dimensional code is transformed into a trapezoid (coordinate conversion described later). It is also possible.

[Appearance of 2D code]
A rectangular code as a prototype of the two-dimensional code is shown in FIG. In order to prevent deterioration of the code image due to reading from an oblique direction, the code height is kept low and the number of code cell arrangements is set to 12 × 7 horizontally long. A cell is the smallest unit of a black and white rectangle, and one cell represents one bit. White corresponds to 0 and black corresponds to 1. In this 12 × 7 cell, 7 bytes (56 bits) of data, 2 bytes (16 bits) of error correction data, and 12 bits of code identification pattern are embedded. A 12 × 7 cell is surrounded by a black frame having the same size as the cell.

  If this code is transformed into a trapezoid by performing coordinate transformation, it becomes as shown in FIG. When this code is read with a camera obliquely, a substantially rectangular code image close to the code image shown in FIG. 1 can be obtained.

[Quiet Zone]
The quiet zone is a blank area provided outside the circumscribed rectangle surrounding the black frame of the two-dimensional code, and is an area necessary for reliable reading. It is desirable that the quiet zone is not less than the maximum cell length and not less than 1 mm.

[Code identification pattern]
The code identification pattern (also referred to as “unique pattern”) is determined by the arrangement of 12 cells arranged in the uppermost row of the cell array. The cells are arranged from left to right, “white black-and-white black white black-and-white black-white black-and-white”. Adopting a pattern of two white and black lines so that the code can be determined even if the print quality and reading quality are low. (See FIGS. 3 and 4).

[Cord size]
The code size is determined by determining the cell size. In a general application, the minimum size of a code that does not include a quiet zone is that the resolution for reading the code is 300 dpi, and each read cell contains 4 pixels on each side. It becomes.

[Error correction]
The error correction method used for the two-dimensional code is the Reed-Solomon method. 2-byte error correction data is added to 7 bytes of data. The number of bytes that can be error-corrected is one.

[Data arrangement]
As described above, data is 7 bytes, error correction data is 2 bytes, and in addition to a total of 9 bytes of data, a 12-bit code identification pattern is arranged in the code. Since the code is read obliquely from the front, the height of the code is kept low, and the number of code cell arrangements is set to 12 × 7 horizontally long. The arrangement position of data in each cell is shown in FIG. FIG. 3 shows the arrangement of data in bytes in each cell. Each byte of data is (s ₁ to s ₇ ), and each byte of the error correction code is (r ₁ to r ₂ ). The arrangement of white 2 cells and black 2 cells in the top row is an identification pattern (also referred to as a unique pattern) for identifying a code. FIG. 4 shows the arrangement of each data in bit units. The number 0 represents MSB and 7 represents LSB.

Next, an encoding (Encode) procedure for creating a trapezoid will be described.
[Outline of encoding procedure]
An outline of procedures required when converting input data into a two-dimensional code is as follows.
Procedure 1: Input data masking Input data is converted into a 7-byte binary format, and an exclusive OR operation is performed on each byte and 0x96 (hexadecimal notation).
Procedure 2 ... Generation of error correction code An error correction code is added to the masked data.
Step 3 ... Generation of rectangular code Input data to which an error correction code is added is arranged in the rectangular code.
Step 4 ... Converting the rectangular code to a trapezoid Transform the rectangular code into a trapezoid.

[Masking of input data] (step S1 in FIG. 5)
When the data format is an integer value character string, it is converted to 7-byte integer type data. If the data format is character or binary, use it as is. After the data format conversion, an exclusive OR operation is performed on each byte of the data to be encoded with a value of 0x96 (hexadecimal notation).

[Generation of Error Correcting Code] (Steps S2 and S3)
Next, an error correction code is generated using a Reed-Solomon code. Polynomial S (x) = s ₁ x ⁸ + s ₂ x ⁷ + s ₃ x ⁶ + s ₄ x ⁵ with each byte value of 7-byte data (s _{1 to} s ₇ ) subjected to exclusive OR operation as a coefficient + S ₅ x ⁴ + s ₆ x ³ + s ₇ x ² is generated and divided by a generator polynomial G (x) = (x−α) (x−α ² ). The generator polynomial actually used is G (x) = x ² + 6x +8. The coefficients (r ₁ , r ₂ ) of the remainder polynomial R (x) = r ₁ x + r ₂ obtained by the division are error correction codes. Here, α is the root of the primitive element 2 on GF (2 ⁸ ), and all operations are performed on GF (2 ⁸ ). The error correction capability is 1 byte.

[Generation of Rectangle Code] (Step S4)
As shown in FIG. 3, each bit of s1 to s7, r1, and r2 is mapped to a code. FIG. 4 shows the arrangement of each data in bit units, where the number 0 represents the MSB and 7 represents the LSB. Bit 1 of each data is assigned to a black cell and bit 0 is assigned to a white cell.

[Rectangle Code Trapezoidalization] (Step S5)
The optical system parameters necessary for calculating the trapezoid of the rectangular code are shown in FIGS. The meanings of the symbols shown are as follows.
O: spatial coordinate origin, S: imaging center of imaging element, SC: optical center, CO: paper surface (z = 0), A: image end to be imaged, B: lower end of code, C: center to be imaged, L: edge of image to be captured (straight line), T: trapezoidal code on z = 0, R: rectangular code on z = 0 that is the basis of trapezoidal code T, SO: distance between imaging element and paper surface, ∠CSO: tilt angle when shooting

  Coordinate transformation (two coordinate transformation) for trapezoidalization will be described. FIG. 8 is a flowchart showing an operation process for coordinate conversion of a two-dimensional code. First, the rectangular code R that is the basis of the two-dimensional code is created on the plane z = 0 (see FIG. 9). Next, the point P in the rectangular code R is rotated by ∠CSO about the straight line L as an axis. This point is set as Pr (step S6). Next, an intersection point Pt between the straight line connecting the centers S and Pr of the imaging surface of the image sensor and the plane z = 0 is obtained (step S7). In this way, the rectangular code R is converted into a trapezoid code T. When all the points in the code have been processed (step S8 / Yes), the coordinate conversion process ends.

  The two-dimensional code created as described above is printed on paper with a printer.

  Next, the two-dimensional code reading device of this embodiment will be described. FIG. 10 is a block diagram of the two-dimensional code reading apparatus according to this embodiment. The two-dimensional code reader is input as a multi-valued image (gray scale, which extracts a luminance signal in the case of color) from image input means such as a digital camera or a scanner, and is temporarily stored in the memory 1. After being stored in the memory 1, binarization is performed using the binarization parameter set in advance by the binarization parameter setting unit 4 (binarization unit 2 in FIG. 10). Using the binarized image, the decoding unit 3 decodes a two-dimensional code existing in the binarized image.

  When the decoding of the two-dimensional code fails, information indicating the failure is output to the binarization parameter setting unit 4. The binarization parameter setting unit 4 corrects the binarization parameter based on the input information of the decoding failure. Decoding failure information is, for example, information indicating that the binarized image has too many black pixels or too many white pixels. Then, image binarization is performed again using the corrected binarization parameter, and the binary image is decoded.

  The configuration of the binarization unit 2 can be considered to use Japanese Patent Laid-Open No. 2001-008032. In this case, the binarization parameter set by the binarization parameter setting unit 4 corresponds to the multiplication coefficient Cb used in the binarization threshold setting circuit disclosed in Japanese Patent Laid-Open No. 2001-008032 (for binarization of multi-valued images). (When the binarization threshold TH to be used is set, the average luminance value ave is multiplied by a coefficient Cb). When Cb increases, black pixels increase in the binarized image, and when Cb decreases, white pixels increase in the binarized image.

  At this time, if Cb = x / 16 (1 <= x <= 15; x is an integer), the binarization parameter setting unit 4 changes the value of x in the range of 1-15. In this case, normally, x = 8 is set. When information indicating that there are too many black pixels is input from the decoding unit 3, (x + 1) / 2 is newly set to x. When information indicating that there are too many white pixels is input from the decoding unit 3, (x + 15) / 2 is newly set to x. Thereby, even when the first decoding fails, a binarized image more suitable for decoding the two-dimensional code can be obtained when the second decoding is executed.

  Next, the decoding unit 3 will be described in detail. FIG. 11 is a block diagram of the decoding unit 3. A portion surrounded by a dotted line in FIG.

  The decoding unit 3 includes a vertex candidate detection unit 5, a black frame determination unit 6, a projective transformation coefficient calculation unit 7, a cell position acquisition unit 8, a code determination unit 9, a bit string acquisition unit 10, an error correction unit 11, a data restoration unit 12, a black The expansion rate calculation unit 13 is configured. Each configuration will be described in detail below.

[Vertex candidate detection unit 5]
The vertex candidate detection unit 5 detects each vertex candidate of the two-dimensional code. Since the vertex candidate is determined to be a vertex of the two-dimensional code only after it is recognized as a black frame (= code frame) by the next black frame determination unit 6, it is referred to as a vertex “candidate”.

  An operation process for detecting a vertex candidate will be described. First, as shown in FIG. 12, diagonal scanning is performed from the four corners of the image to detect black pixels. Let them be A, B, C, D. Further, in FIG. 13, the pixels are tracked by a minimum of three or more pixels such as A, B, C, and D in the direction of the arrow, for example, 45 degrees on the lower right for A and 45 degrees on the lower left for B. Determine if it is a black pixel. If they are all black pixels, A, B, C, and D are detected as vertex candidates, and the process proceeds to the black frame determination unit 6.

[Black frame determination unit 6]
The black frame determination unit 6 determines whether there is a black frame of a two-dimensional code that connects the detected vertex candidates. When there is almost a white pixel area outside the straight line connecting the two adjacent vertices and almost black pixel area inside, it is determined that the frame is a black frame. In the black frame detection, as shown in FIG. 14, the ratio of black pixels among pixels passing through four straight lines (black frame determination lines) connecting the end pixels tracked by the vertex candidate detection unit 5 from the vertex candidates A, B, C, and D. Is 90% or more for every straight line, it is determined that it is a temporary black frame.

  Next, pixels on the straight line connecting adjacent quiet zone determination pixels are defined as pixels having a length of about 0.5 mm or about 5 pixels outward from the vertex candidates A, B, C, and D, respectively. If it is 80% or more of white pixels for each straight line, it is determined as a quiet zone, and is determined to be a true black frame together with the previously determined temporary black frame.

  In this way, the vertex candidates A, B, C, and D are determined as the vertices A, B, C, and D of the two-dimensional code. At this time, the coordinates of the vertices A, B, C, and D are also detected. Here, if the black frame of the code cannot be detected, reading ends. If a black frame is detected, the process proceeds to projective transformation coefficient calculation processing.

[Projection conversion coefficient calculation unit 7]
The projection conversion coefficient calculation unit 7, in the projection conversion coefficient calculation unit 7, virtually defines the coordinates of each vertex of the code frame detected by the black frame detection and each vertex of the code frame when the code is rectangular. From the coordinates, a coefficient (projective transformation coefficient) for mapping the specified center coordinates of each cell and the center coordinates of each cell of the read two-dimensional code image is obtained. The prescribed coordinates are coordinates when a two-dimensional code is created as a square having a length of 1 for each cell, and the code becomes a rectangular shape as a result. The specified center coordinates are the center coordinates of each cell in the specified coordinate system as described above. Details of the projective transformation will be described later.

[Cell position acquisition unit 8]
The cell position acquisition unit 8 calculates the coordinates of each cell of the two-dimensional code read from the specified center coordinates of each cell using the obtained projective transformation coefficient, and acquires the position of the cell to be read.

[Code determination unit 9]
The code determination unit 9 has a function as a unique pattern detection unit. The code determination unit 9 detects a pattern specific to the two-dimensional code (code identification pattern) from the acquired cell position, determines whether the target two-dimensional code is the target, and the detection result of the two-dimensional code specific pattern is When a predetermined condition is satisfied, the result is output to the binarization parameter setting unit 4.

  For code determination, the binary image and the cell position of the identification pattern are required. The cell position of the identification pattern is set as the sampling center coordinate of the image data, and the data is read as ‘1’ if the number of black pixels of 3 × 3 pixels centering on the coordinate exceeds the number of white pixels, and ‘0’ otherwise. The read data is compared with the true data of the identification pattern. If there are two or less errors, it is determined that the currently handled graphic is a two-dimensional code.

  If the code determination is successful, the process proceeds to the next bit string acquisition unit 10.

  If the code determination fails, the number of black patterns in which the true pattern data matches the read data is compared with the number of white patterns, and the black patterns all match (all match: first match Condition), the number of matched white patterns has a predetermined degree of coincidence (satisfies a predetermined degree of coincidence: second condition). A signal for increasing white pixels in the image after binarization is output to the binarization parameter setting unit 4. On the other hand, when all white patterns match and the number of matching black patterns is 4 or less (when the white pattern satisfies the first condition and the black pattern does not satisfy the second condition), the image in the image after binarization A signal for increasing the number of black pixels is output to the binarization parameter setting unit 4. Receiving this signal, the binarization parameter setting unit 4 resets the binarization parameters as described above, and redoes the decoding process from the image binarization.

[Bit string acquisition unit 10]
The bit string acquisition unit 10 reads the black and white information of each cell of the two-dimensional code determined as the target two-dimensional code and acquires a bit string. Therefore, to acquire a bit string, a two-dimensional code image and a cell position of the read two-dimensional code image are necessary. If the position of each cell of the read two-dimensional code image is the sampling center coordinate of the image data and the number of black pixels of 3 × 3 pixels centering on that coordinate exceeds the number of white pixels (there are many black pixels), “1”, so Otherwise, the data is read as “0” (there are many white pixels) and the data is arranged.

[Error correction unit 11]
The error correction unit 11 performs error correction from the bit string, performs a predetermined operation when error correction cannot be performed, and outputs the result to the binarization parameter setting unit 4. The read 72-bit data is input to the error correction unit 11 and judgment of Reed-Solomon error correction is performed. If there is no error or error correction is possible, data of 56 bits (7 bytes) is output to the data restoration unit 12. If error correction is impossible, the black expansion rate is calculated from the ratio of the number of black and white pixels and the 72-bit data. When the black expansion rate is greater than 1, a signal for increasing white pixels in the image after binarization is output to the binarization parameter setting unit 4. When the black expansion rate is smaller than 1, a signal for increasing black pixels in the image after binarization is output to the binarization parameter setting unit 4. Details of the black expansion rate will be described later.

[Data restoration unit 12]
The data restoration unit 12 restores the original data after error correction. Specifically, the original data is restored by calculating an exclusive OR with 0x96 (hexadecimal notation) for each byte of the 7-byte data obtained by the error correction unit 11.

[Black expansion rate part 13]
The black expansion rate is a measure indicating how much the black area of the two-dimensional code to be decoded is expanded. Since the input image is a grayscale image, binarization processing is executed to binarize the image. When the code is converted into a rectangle such that each cell is square in the two-dimensional code area, it defines how much the ratio of black and white changes from the ideal black and white ratio state.

First, the number of pixels of one line of “white monochrome black white white” of the code identification pattern “white monochrome black white monochrome black white monochrome black” is counted, and the number of white pixels is set to w and the number of black pixels is set to b. The average expansion pixel number exp in the black and white adjacent portion is expressed by the following equation.
exp = (b / 2−w / 4) / 2 = (2b−w) / 8

Therefore, the black expansion rate Bexp is as follows.
Bexp = exp / (2b + w) * 8 + 1 = 4b / (2b + w)
This shows the protrusion ratio of black cells adjacent to white on the basis of the ideal cell length. If there is no black expansion, the value of Bexp is 1.

  When error correction fails in the error correction unit 11 and the black expansion rate is greater than 1, a signal for increasing white pixels in the image after binarization is output to the binarization parameter setting unit 4 . When the black expansion rate is smaller than 1, a signal for increasing black pixels in the image after binarization is output to the binarization parameter setting unit 4.

  The operation process of the two-dimensional barcode reader is shown in the flowchart of FIG. An image is input (a grayscale image is input) (step S10), and binarization parameters are set (step S11). As described above, the binarization parameter set in the binarization parameter setting step corresponds to the multiplication coefficient Cb used in the binarization threshold setting circuit disclosed in Japanese Patent Laid-Open No. 2001-008032. When Cb increases, black pixels increase in the binarized image, and when Cb decreases, white pixels increase in the binarized image. At this time, if Cb = x / 16 (1 <= x <= 15; x is an integer), the binarization parameter setting step changes the value of x in the range of 1-15. Normally, x = 8 is set. When information indicating that there are too many black pixels is input from the decoding step, (x + 1) / 2 is newly set to x. When information indicating that there are too many white pixels is input from the decoding step, (x + 15) / 2 is newly set to x. As a result, even if the first decoding fails, the binarization parameter is reset in the binarization parameter setting step depending on the situation, and a binary image suitable for decoding the two-dimensional code can be obtained using the parameter. It is possible to redo the decoding of the two-dimensional code.

  Next, image binarization is performed (step S12). Image binarization binarizes an image using binarization parameters. Thereafter, each vertex candidate of the two-dimensional code in the binarized image is detected (step S13), and the black frame of the two-dimensional code is determined using the detected vertex candidate (step S14). A projective transformation coefficient for correcting the distortion of the two-dimensional code is calculated based on the coordinates of each vertex of the area determined to be a black frame (step S15). Each cell position in the two-dimensional code is acquired by projective transformation (step S16).

  Next, code determination processing is performed (step S17). A pattern specific to the two-dimensional code is detected from the acquired cell position to determine whether or not the target two-dimensional code is the target (step S18). When the detection result of the two-dimensional code specific pattern satisfies a predetermined condition The result is output to the binarization parameter setting unit 4.

  When the target two-dimensional code is not determined (step S18 / No), whether the black pattern is completely matched: first condition (step S19), whether the white pattern is completely matched: second condition (Step S26) is determined, and the result is output to the binarization parameter setting step (Step S11).

  If it is determined that the target two-dimensional code is determined (step S18 / Yes), a read bit string is acquired for the black and white information of each cell of the two-dimensional code determined to be the target two-dimensional code (step S20).

  Error correction is performed from the bit string (step S21). If error correction cannot be performed (step S22 / No), the black expansion rate is calculated (step S24). It is determined whether or not the black expansion rate exceeds 1.0 (step S25), and the result is output to the binarization parameter setting step (step S11).

  If the error correction is successful (step S23), the original data is restored (step S23).

  Note that the operation processing shown in FIG. 15 corresponds to a procedure for decoding a two-dimensional code by software in a computer or the like, and falls within the scope of the present invention.

  Next, details of the projective transformation will be described. Projective transformation is transformation when a figure / object in a three-dimensional space is displayed on a two-dimensional plane / screen. The coordinates of an object in a three-dimensional space are converted into coordinates on a two-dimensional plane.

  Exact projective transformation will be described with reference to FIG. FIG. 16 is a diagram schematically showing a read code image (left), and a diagram schematically showing an image including an electronically generated code (right). Let each vertex of the code be As to Ds and Ar to Dr. Each relationship is connected by the following two formulas.

  The above formulas 1 and 2 represent conversion formulas from code symbol coordinates to image coordinates, and define the coordinate conversion from Ar to Dr to As to Ds. In this equation, b1 to b8 are conversion parameters, and since these are unknown numbers, the coordinate values of Ar to Dr and As to Ds are substituted into the equation to generate and calculate an eight-way linear simultaneous equation. Using the calculated b1 to b8, the center coordinate Prk (k = 0 to 83) of each cell of the electronically generated code is converted to obtain the sampling center coordinate Psk (k = 0 to 83) of the code image. Ask.

As to Ds: calculated by detecting vertex candidates.
Psk: calculated by the above formula.
Ar to Dr: stored at the time of code generation.
Prk: Stored at the time of code generation.

  According to the invention of the present application, even if the decoding of a two-dimensional code printed with inks of various concentrations or a two-dimensional code printed with various types of printers fails, the cause of the failure is fed back to image binarization. By doing so, an appropriate binary image can be obtained by decoding, and the two-dimensional code can be reliably decoded.

  Note that the program for the CPU to execute the above-described operation processing of the two-dimensional code reader (the processing shown in the flowchart of FIG. 15) constitutes a program according to the present invention. As a recording medium for recording the program, a semiconductor storage unit, an optical and / or magnetic storage unit, or the like can be used. By using such a program and a recording medium in a system having a configuration different from that of each of the above-described embodiments and causing the CPU to execute the program, substantially the same effect as the present invention can be obtained.

  Although the present invention has been specifically described above based on the preferred embodiments, it is needless to say that the present invention is not limited to the above-described ones and can be variously modified without departing from the gist thereof.

It is a figure which shows the two-dimensional code before the coordinate transformation which concerns on embodiment of this invention. It is a figure which shows the trapezoid two-dimensional code after the coordinate transformation which concerns on embodiment of this invention. It is a figure which shows arrangement | positioning of the byte unit in the two-dimensional code which concerns on embodiment of this invention. It is a figure which shows arrangement | positioning of the bit unit in the two-dimensional code which concerns on embodiment of this invention. It is a flowchart which shows the operation | movement process of the encoding which concerns on embodiment of this invention. It is a figure which shows the optical system parameter required for calculation of the trapezoidization of the rectangular code based on embodiment of this invention. It is a figure which shows the optical system parameter required for calculation of the trapezoidization of the rectangular code based on embodiment of this invention. It is a flowchart which shows the operation | movement process of the coordinate transformation which concerns on embodiment of this invention. It is a figure which shows the optical system parameter required for calculation of the coordinate transformation which concerns on embodiment of this invention. It is a block block diagram of the two-dimensional code reader which concerns on embodiment of this invention. It is a block block diagram of the decoding part 3 which concerns on embodiment of this invention. It is a figure for demonstrating the vertex candidate pixel detection which concerns on embodiment of this invention. It is a figure for demonstrating the vertex candidate detection which concerns on embodiment of this invention. It is a figure for demonstrating the black frame determination which concerns on embodiment of this invention. It is a flowchart which shows the operation | movement process of the two-dimensional barcode reader which concerns on embodiment of this invention. It is a figure for demonstrating the projective transformation which concerns on embodiment of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Memory 2 Binarization part 3 Decoding part 4 Binarization parameter setting part 5 Vertex candidate detection part 6 Black frame determination part 7 Projection transformation coefficient calculation part 8 Cell position acquisition part 9 Code determination part 10 Bit sequence acquisition part 11 Error correction part 12 Data restoration unit 13 Black expansion rate calculation unit

Claims (8)

The multi-valued image including the region of the two-dimensional code having a portion of the cell Le plurality of white and black squares as a unique pattern, the following equation
Cb = x / 16
Where Cb: multiplication factor
x: integer, 1 ≦ x ≦ 15
An image binarization means binarizes and outputs the binary image based on,
Two-dimensional code detection means for detecting a region of the two-dimensional code from the binary image;
A unique pattern detecting means for detecting a unique pattern from the region of the two-dimensional code;
A decoding means for decoding the pre-SL 2-dimensional code, a,
The unique pattern detection means, among the unique pattern of black or white constituting the pattern, the cells constituting one pattern, with the proviso that match all the cells constituting the predetermined pattern Determining means for determining a condition of 1 and a second condition on the condition that a matching degree of the other pattern satisfies a predetermined matching degree;
The binarization means includes
Meet black pattern matching score of the first condition, if the match degree of the white pattern does not satisfy the second condition, the degree of matching of the white pattern to the second condition is satisfied Re-binarize multi-valued image,
Coincidence of the white pattern satisfies the first condition, when the degree of matching of the black pattern does not satisfy the second condition, the degree of matching of the black pattern to the second condition is satisfied Binarize the multi-valued image again,
Before SL again binarizing detects a binary image or al the unique pattern, the two-dimensional code reading apparatus characterized by decoding the two-dimensional code in the decoding means.   When the binarization means binarizes the multi-valued image again, the value of x in the equation is
  If the black pattern coincidence satisfies the first condition and the white pattern coincidence does not satisfy the second condition, increase
  Decrease if the degree of coincidence of the white pattern satisfies the first condition and the degree of coincidence of the black pattern does not satisfy the second condition
  The two-dimensional code reading device according to claim 1, wherein processing is performed.   The multi-value image has data information and error correction information,
  When the decoding means fails in error correction using the error correction information,
  Calculate the rate of change from the ideal black-and-white ratio to the black-and-white ratio of the image that was binarized,
  When the binarization means binarizes the multi-valued image again, the value of x in the equation is
  If the rate of change indicates that the percentage of black is increasing, reduce it,
  Increase if the rate of change indicates that the percentage of white is high
  The two-dimensional code reading device according to claim 1 or 2, wherein processing is performed. The multi-valued image including the region of the two-dimensional code having a portion of the cell Le plurality of white and black squares as a unique pattern, the following equation
Cb = x / 16
Where Cb: multiplication factor
x: integer, 1 ≦ x ≦ 15
By binarizing based on an image binarization step of outputting the binary image,
A two-dimensional code detection step of detecting a region of the two-dimensional code from the binary image;
A unique pattern detecting step for detecting a unique pattern from the region of the two-dimensional code;
It includes a decoding step for decoding the pre-SL 2-dimensional code, a,
The unique pattern detection step, one of the unique pattern of black or white constituting the pattern, the cells constituting one pattern, with the proviso that match all the cells constituting the predetermined pattern A determination step of determining a condition of 1 and a second condition on the condition that a matching degree of the other pattern satisfies a predetermined matching degree;
The binarization step includes
Meet black pattern matching score of the first condition, if the match degree of the white pattern does not satisfy the second condition, the degree of matching of the white pattern to the second condition is satisfied Re-binarize multi-valued image,
Coincidence of the white pattern satisfies the first condition, when the degree of matching of the black pattern does not satisfy the second condition, the degree of matching of the black pattern to the second condition is satisfied Binarize the multi-valued image again,
It detects the leading SL again binarizing the binary image or al the unique pattern, two-dimensional code reading method characterized by decoding the two-dimensional code in the decoding step.   When the binarization step binarizes the multi-value image again, the value of x in the equation is
  If the black pattern coincidence satisfies the first condition and the white pattern coincidence does not satisfy the second condition, increase
  Decrease if the degree of coincidence of the white pattern satisfies the first condition and the degree of coincidence of the black pattern does not satisfy the second condition
  5. The two-dimensional code reading method according to claim 4, wherein processing is performed.   The multi-value image has data information and error correction information,
  The decoding step, when error correction using the error correction information fails,
  Calculate the rate of change from the ideal black-and-white ratio to the black-and-white ratio of the image that was binarized,
  When the binarization step binarizes the multi-value image again, the value of x in the equation is
  If the rate of change indicates that the percentage of black is increasing, reduce it,
  Increase if the rate of change indicates that the percentage of white is high
  6. The two-dimensional code reading method according to claim 4, wherein processing is performed. A two-dimensional code reading program for causing a computer to execute the two-dimensional code reading method according to any one of claims 4 to 6 . A computer-readable recording medium having recorded thereon the program according to claim 7.

Images