JP2002329163A - Device method and program for multi-valued image binarization, and device, method and program for reading two-dimensional code - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a device, a method, and a program for two-dimensional code reading which accurately read a two-dimensional code out of a multi-valued image. SOLUTION: The reading device is equipped with a multi-valued image binarization/quiet zone generation part 21 which generates a quiet zone by binarizing the multi-valued image, a two-dimensional code presence area approximately determining part 22 which approximately determining a two-dimensional code presence area by detecting guide cell peaks, a two-dimensional code rotational angle determination part 23 which determines the angle of rotation of the two-dimensional code, a guide cell circumscribing line detection part 24 which detects a guide cell circumscribing line, a timing cell circumscribing line detection part 25 which detects a timing cell circumscribing line, a timing cell detection part 26 which detects a timing cell, a lost timing cell complementing part 27 which complements a lost timing cell, a data cell area setting part 28 which sets a data cell area, and a data cell logical state identification part 29 which identifies the logical state of a data cell.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus and method for binarizing a multi-valued image and a program for executing the method. More particularly, the present invention relates to a character or graphic represented by a combination of a plurality of dots or a combination thereof. (Hereinafter, collectively referred to as a symbol), a multi-value image binarization device for binarizing a multi-value image obtained by imaging a two-dimensional code or the like included in the symbol, and a multi-value image binarization method And a multi-valued image binarization program. Further, the present invention relates to a two-dimensional code reader, a method, and a program for executing the method,
The present invention relates to a two-dimensional code reader, a two-dimensional code reading method, and a two-dimensional code reading program for reading a two-dimensional code from a multivalued image obtained by imaging a two-dimensional code.

[0002]

2. Description of the Related Art For a long time, a symbol such as a character or a figure is given to an article to identify and manage the article. One specific example is that information to be given (recorded) is represented by a binary number. It is known that the binary number is represented in binary by a plurality of dots.

As another method, a barcode for managing information by a combination of a plurality of line segments which are parallel to each other and have different widths has been widely used. In recent years, a barcode for recording a larger amount of information than a barcode has been used. A two-dimensional code is used as a means.

The two-dimensional code includes a data matrix (registered trademark of International Data Matrix Inc.), a Vericode (registered trademark of Mitsubishi Corporation), a QR code (registered trademark of Denso Corporation), a C code (CP code) The information to be recorded is a binary number such as Teishi Corporation (registered trademark), and white (bright) and black (dark)
A matrix-type two-dimensional code (hereinafter, referred to as a two-dimensional code) that expresses the information by arranging cells (lattices) that represent binary values two-dimensionally in a vertical direction and a horizontal direction like this is known.

The method of expressing a binary value is not limited to the method using cells colored white (bright) and black (dark) as described above. If one cell and the other cell have different characteristics or properties, such as a method of expressing a cell or a method of displaying two types of cells with different colors on an information display device such as a liquid crystal display or a cathode ray tube Good. The dot is not limited to a round shape, but may be any shape that can represent a small point, such as a polygon or an ellipse.

[0006] Symbols and two-dimensional codes (hereinafter, collectively referred to as symbols and two-dimensional codes) recorded on articles in the above-described manner are given to various articles, and the codes are transferred to a CCD camera ( The image is captured by a charge coupled device camera) or a scanner device, captured as a multi-valued image, and the recorded information is read by processing and processing the multi-valued image. Here, in order to correctly read the information, after accurately reading the position and shape of the code existing in the multi-valued image, a plurality of dots and cells represented by a plurality of dots constituting the code are used.
It is necessary to correctly determine the value information.

A method of reading information of a two-dimensional code from a captured multi-valued image will be described below by taking a data matrix code as an example. The data matrix code includes a guide cell, a timing cell, and a data cell. Specifically, as shown in FIG. 1, the guide cells are arranged in a substantially L-shape, form a guide cell row, and form two rows in the image. This is a reference for obtaining the position and shape of the dimensional code.

The timing cells are intermittently and equidistantly arranged at positions parallel to the guide cells to form a timing cell row and serve as a reference for determining the positions of the data cells. Represents the recorded binary information which is arranged inside a rectangle surrounded by the guide cells and the timing cells. It should be noted that the term “cell” is used to refer to a general term of a guide cell, a timing cell, and a data cell.

Conventionally, in order to read information recorded in a two-dimensional code from a multi-valued image obtained by imaging an area including such a two-dimensional code, a multi-valued image having a multi-valued image as a binary image is used. Binarizing means (A), two-dimensional code detecting means (B) for detecting a two-dimensional code from a binary image, timing cell detecting means (C) for detecting a two-dimensional code timing cell, and two-dimensional code A data determining means (D) for detecting a data cell position and determining a logical value of the data cell;
Dimensional code readers are commonly used.

More specifically, the multi-value image binarization means (A) includes a median (maximum value) of the maximum luminance value and the minimum luminance value in the entire multi-value image, Means for binarizing the entire multi-valued image including the two-dimensional code using a fixed threshold value such as the arithmetic mean of the luminance values of the pixels, or shading correction of the entire multi-valued image, and then fixing the entire corrected image. Means for binarizing the multivalued image, dividing the entire multivalued image into regions smaller than the multivalued image to generate a large number of small regions, and fixing a fixed threshold such as an arithmetic average of luminance values in each small region. Means for obtaining a value and binarizing the value are known.

The two-dimensional code detecting means (B) includes a contour tracing process for tracing a contour edge of a figure composed of a group of pixels in a binary image, and a method of labeling a group of connected pixels as one figure. The positions and shapes of a plurality of figures included in the binary image are detected by a labeling process or the like provided and identified, and two or more of the figures are similar to and orthogonal to the cell design shape in the two-dimensional code. There is known a means for detecting a row of graphics forming a straight line of a book and obtaining the position and inclination angle of a two-dimensional code having the guide cell row as a part of the outer periphery by detecting the graphic row as a guide cell row. I have.

The timing cell detecting means (C) detects a row of graphics opposite to the guide cell row detected by the two-dimensional code detection means, and detects the center of the graphics arranged at a predetermined interval in the graphics row. Is known as a timing cell position.

Further, the data determination means (D) determines a point where a normal line of a guide cell row passing through the center of each timing cell intersects another normal line as a data cell position, and determines the position of the data cell. When there is a label (a figure recognized by the labeling process) or a contour (a contour of the figure recognized by the contour processing) included therein, there is known means for determining a logical value of the data cell.

[0014]

However, the two-dimensional code reading apparatus comprising the above means has little unevenness in brightness and noise in the multi-valued image, has no data cell displacement, and has a blurred or missing data cell shape. It is effective when the deformation is stable without such deformation, but when these conditions are not satisfied, it causes various problems.

For example, if the multi-valued image has luminance unevenness, there is a problem that the binarization cannot be performed accurately. If the multi-valued image has noise, the noise becomes an obstacle and causes a reading error of the two-dimensional code. If there is a problem and the cell is misaligned or deformed, it will be difficult to detect the guide cell and the timing cell, so that the data cell determination will be inaccurate and the reliability of the read information will be reduced. was there.

The above problem is not limited to a two-dimensional code, but after binarizing a multi-valued image obtained by imaging a symbol represented by a plurality of dots, the symbol is detected from the binary image. The same applies to the case, and there is a problem that it is difficult to detect a symbol due to noise or uneven brightness.

As means for solving such a problem, in particular, the problem of binarizing a multi-valued image obtained by capturing a two-dimensional code, Japanese Patent Laid-Open No. 10-63771 discloses a method for determining a plurality of small areas. The average value and the amplitude of the maximum value and the minimum value of the density value of the predetermined pixel of the cell in each of the small areas are obtained, and the amplitudes of the adjacent small areas are compared with each other. In the case of a small area having an amplitude, a two-dimensional code reading device having a local threshold value determining means for determining a threshold value based on a small area adjacent to the small area has a short processing time and a small unevenness in luminance. It is disclosed that cells can be binarized without being affected.

However, after the position and shape of the two-dimensional code are accurately detected from the multi-valued image, the local threshold value obtained by the local threshold value determining means is not affected by luminance unevenness. Of the two-dimensional code detected using
If the multi-valued image is binarized using the threshold value obtained by the local threshold value determining unit before detecting the two-dimensional code from the binary image, noise may be generated. And the two-dimensional code cannot be read accurately.

SUMMARY OF THE INVENTION The present invention has been made in view of the above-described problems. In converting a multi-valued image into a binary image, the present invention eliminates luminance unevenness of the multi-valued image and noise around symbols and two-dimensional codes. Multi-value image binarization apparatus, multi-value image binarization method, multi-value image binarization method, multi-value image binarization apparatus, multi-value image binarization device, multi-value image binarization device It is intended to provide a value image binarization program.

In addition, even if there are uneven brightness, noise, deformation of a two-dimensional code, deformation such as bleeding or chipping of a cell, loss of a cell, displacement of a cell, etc., a two-dimensional image can be accurately obtained from a multi-valued image. It is another object of the present invention to provide a two-dimensional code reading device capable of reading a code, a two-dimensional code reading method, and a two-dimensional code reading program.

[0021]

In order to solve the above-mentioned problems, a multi-valued image binarizing device according to the present invention converts a multi-valued image including a symbol represented by a combination of a plurality of dots into a two-valued image.
A multi-valued image binarizing device, comprising: a multi-valued image dividing means for dividing a multi-valued image into a plurality of small regions; and a binarized small region to be binarized among the small regions. Threshold value calculating means for calculating a threshold value of the binarized small region from a luminance value of a threshold value calculation region having an area larger than the binarized small region. The binarized small area is binarized by the threshold value calculated by the threshold value calculating means, and at the same time, if a symbol exists in the binarized small area, a quiet zone is generated outside the symbol. And small area binarization means.

A multi-valued image binarizing device according to the present invention is a multi-valued image binarizing device for binarizing a multi-valued image including a two-dimensional code. A multi-valued image dividing means for dividing into small areas; and a threshold value calculation area which includes a binarized small area which is a binarization target among the small areas and is larger than the binarized small area. From the luminance value,
Threshold value calculating means for calculating a threshold value of the binarized small area;
The binarized small area is binarized by the threshold value, and at the same time, when a two-dimensional code exists in the binarized small area, a small area 2 for generating a quiet zone outside the two-dimensional code. Value conversion means.

Further, in the multi-value image binarizing device according to the present invention, the threshold value calculating means includes an arithmetic mean of a maximum luminance value and a minimum luminance value in the threshold value calculation region. Value.

Further, the multi-value image binarization method according to the present invention is a multi-value image binarization method for binarizing a multi-value image including a symbol represented by a combination of a plurality of dots. A multi-valued image dividing step of dividing the multi-valued image into a plurality of small regions, including a binarized small region to be binarized among the small regions, and A threshold value calculating step of calculating a threshold value of the binarized small region from a luminance value of a threshold value calculation region having a large area;
The binarized small area is binarized by the threshold value calculated by the threshold value calculating means, and at the same time, if a symbol exists in the binarized small area, a quiet zone is set outside the symbol. Generating a small area binarizing step.

Further, the multi-value image binarization method according to the present invention is a multi-value image binarization method for binarizing a multi-value image including a two-dimensional code. A multi-valued image dividing step of dividing into small regions; and a threshold calculation region including a binarized small region to be binarized among the small regions and being larger than the binarized small region. A threshold value calculating step of calculating a threshold value of the binarized small area from a luminance value; and binarizing the binarized small area with the threshold value, And a small area binarizing step of generating a quiet zone outside the two-dimensional code when the two-dimensional code exists.

Further, in the multi-value image binarization method according to the present invention, the threshold value calculating step includes a step of calculating an arithmetic average of a maximum luminance value and a minimum luminance value in the threshold value calculation region. Value.

A multi-value image binarization program according to the present invention is for causing a computer to execute a multi-value image binarization method.

Further, a two-dimensional code reader according to the present invention is a two-dimensional code reader for reading the two-dimensional code from a multi-valued image including the two-dimensional code. Binarization for generating a quiet zone around a two-dimensional code included in the multi-valued image /
A quiet zone generating means, a timing cell detecting means for detecting a timing cell in a two-dimensional code surrounded by the quiet zone, and a straight line passing through a center or a midpoint of each timing cell detected by the timing cell detecting means. Data cell area setting means for setting a predetermined area centered on another intersection with the straight line as a data cell area used for identifying a logical state of each data cell in the two-dimensional code; Data cell logical state identification means for identifying a logical state of each data cell area based on an overlap between a label obtained by labeling a code area and a data cell area set by the data cell area setting means And with.

Further, the two-dimensional code reader according to the present invention provides the two-dimensional code reader obtained by the binarization / quiet zone generation means.
A two-dimensional code approximate existence area determining unit that determines a rectangular area including the entire area surrounded by the quiet zone in the binarized image as a two-dimensional code approximate existence area; 2 in the two-dimensional code approximate existence area for each of the two sides
Two-dimensional code rotation angle determining means for determining the rotation angle of the two-dimensional code, the rotation angle determined by the two-dimensional code rotation angle determining means, and a guide cell tangent line circumscribing the guide cells arranged in a line in the two-dimensional code A guide cell outer tangent detecting means for detecting the guide cell outer tangent based on the passing point, a rotation angle obtained by the two-dimensional code rotation angle determining means, and timing cells arranged in a line in the two-dimensional code. A timing cell outer tangent detecting means for detecting the timing cell outer tangent based on a passing point of the timing cell outer tangent circumscribing the guide cell outer tangent detected by the guide cell outer tangent detecting means; The area surrounded by the timing cell external tangent detected by the timing cell external tangent detecting means is the two-dimensional code.

Further, the two-dimensional code reader according to the present invention calculates the position where the timing cell has disappeared when a part of the timing cell in the two-dimensional code has disappeared, and calculates the position where the timing cell has disappeared. It is provided with an erasure timing cell complementing means for complementing the timing cell.

Also, in the two-dimensional code reading apparatus according to the present invention, the binarization / quiet zone generating means divides a multi-valued image into a plurality of small areas, and Including the binarized small area that is the target of the binarization,
A threshold calculation region having an area larger than the binarized small region is set, and a threshold value for the binarized small region is set based on a luminance value in the threshold calculation region. Calculate,
The binarized small area is binarized by comparing the threshold value and the luminance value of each pixel in the binarized small area.

Further, in the two-dimensional code reading device according to the present invention, when a guide cell is arranged at an inner edge of the two-dimensional code, the two-dimensional code outline area determining means may include the binarization / quiet zone generation. Performing a pattern matching process on the binarized image obtained by the means using a mask to obtain a pattern matching value at each coordinate of the binarized image, and determining a point at which the pattern matching value is maximum as a guide cell. A quadrangular area that is detected as a vertex, includes the guide cell vertex, and includes the entire area surrounded by the quiet zone is determined as the two-dimensional code general existence area.

Further, in the two-dimensional code reader according to the present invention, when the guide cells are arranged in a substantially L-shape,
A substantially L-shaped mask is used as the mask.

Further, in the two-dimensional code reading device according to the present invention, the two-dimensional code rotation angle determining means may include an angle within a predetermined range with respect to each of two mutually intersecting sides of the two-dimensional code existence area. By rotating the projection axis by a predetermined angle and performing projection processing on the two-dimensional code approximate existence area, a projection intensity on the projection axis is obtained for each rotation angle, and the projection intensity is obtained for each rotation angle of the projection axis. Of the two-dimensional code are obtained, and the rotation angle at which the highest maximum value among the maximum values is obtained is obtained, and the rotation angle is defined as the rotation angle of the two-dimensional code with respect to each side of the two-dimensional code roughly existing area. Things.

Further, in the two-dimensional code reader according to the present invention, the guide cell outer tangent detecting means sets the projection axis as an object with respect to each of two mutually intersecting sides of the two-dimensional code existence area. By performing projection processing by inclining by the rotation angle of the two-dimensional code corresponding to the side being determined, a projection intensity on the projection axis is determined, and a passing point of a guide cell external tangent line is determined based on the projection intensity, A guide cell outer tangent is generated from the passing point and the rotation angle of the two-dimensional code.

Further, in the two-dimensional code reader according to the present invention, the timing cell outer tangent detecting means may detect the projection axis with respect to each of two mutually intersecting sides of the two-dimensional code existence area. By performing the projection process by inclining by the rotation angle of the two-dimensional code corresponding to the target side obtained by the code rotation angle determination means,
A projection intensity on the projection axis is obtained, a passing point of a timing cell outer tangent is obtained based on the projection intensity, and a timing cell outer tangent is generated from the passing point and a rotation angle of the two-dimensional code.

Further, in the two-dimensional code reader according to the present invention, when the timing cell external tangent detection means generates a plurality of timing cell external tangents, the timing cell external tangent detection means outputs the plurality of generated timing tangents. Each of the cell outer tangents is a temporary timing cell outer tangent, and each side of a quadrangle formed by the intersecting temporary timing cell outer tangents and the mutually intersecting guide cell outer tangents detected by the guide cell outer tangent detecting means. Is the closest to the length of one side of the two-dimensional code as a design value, and the second condition is that the number of timing cells existing inside the tentative timing cell outer tangent is the largest. A temporary timing cell external tangent line that satisfies one or both of the above is selected as a normal timing cell external tangent line. Than it is.

Further, in the two-dimensional code reader according to the present invention, the timing cell detecting means may detect the timing cell external tangent detected by the timing cell external tangent detecting means and the timing cell external tangent from the one timing cell external tangent. It is surrounded by a straight line translated by one cell to the guide cell outer tangent side facing one timing cell outer tangent line, the other timing cell outer tangent line, and the guide cell outer tangent line facing the other timing cell outer tangent line. Projecting the binarized image onto the one timing cell outer tangent line to obtain the projection intensity on the one timing cell outer tangent line, calculating a threshold based on the projection intensity, Is used to binarize the projection intensity to obtain a provisional timing cell, which is opposed to the other timing cell external tangent and the other timing cell external tangent. The distance between the guide cell circumscribed line,
A calculated timing cell is obtained from the number of guide cells and the number of timing cells included in one side of the two-dimensional code as a design value, and the provisional timing cell and the calculated timing cell are compared to correspond one-to-one with the calculated timing cells. The temporary timing cell to be used is a normal timing cell.

Further, in the two-dimensional code reader according to the present invention, the lost timing cell complementing means sets a position of the calculated timing cell corresponding to the lost timing cell as a temporary position of the lost timing cell, Two regular timing cells adjacent to the same temporary position of the cell are obtained, and the timing cell is complemented to a position indicated by an internal dividing point or an external dividing point of the two adjacent normal timing cells.

Further, in the two-dimensional code reader according to the present invention, the data cell area setting means may be configured to determine whether a straight line passing through a center or a midpoint of each timing cell detected by the timing cell detecting means and another straight line. An intersection is obtained, and each region having a predetermined area centered on each intersection is generated as the data cell region.

Further, in the two-dimensional code reader according to the present invention, the data cell logical state identifying means performs a labeling process on an area surrounded by the guide cell outer tangent and the timing cell outer tangent, and Is generated, among the generated labels, the area of one label is calculated the average area of the labels within a predetermined range, of the generated labels, the area is centered on the average area A predetermined range of labels is extracted, and a logical state of each data cell area is determined based on an overlap between the extracted label and the data cell area set by the data cell area setting means.

Further, in the two-dimensional code reader according to the present invention, the data cell logical state identifying means may be arranged such that when the generated label extends over a plurality of data cell regions by itself, the label is most widely overlapped. The logic state of the data cell area is determined to be on, and the ratio of the area occupied by a label other than the single label to the data cell area where the logic state is determined to be off is a predetermined value or more,
In addition, when the label overlaps the center of the data cell area at a ratio of a predetermined area or more, the logic state of the data cell area is determined to be on.

A two-dimensional code reading method according to the present invention is a two-dimensional code reading method for reading the two-dimensional code from a multi-valued image including the two-dimensional code, wherein the two-dimensional code is binarized. Binarization for generating a quiet zone around a two-dimensional code included in the multi-valued image /
A quiet zone generating step, a timing cell detecting step of detecting a timing cell in the two-dimensional code surrounded by the quiet zone, and a straight line passing through the center or midpoint of each timing cell detected by the timing cell detecting step. A data cell area setting step of setting a predetermined area centered on an intersection with the other straight line as a data cell area used for identifying a logical state of each data cell in the two-dimensional code; A data cell logical state identifying step for identifying a logical state of each data cell area based on an overlap between a label obtained by labeling a code area and the data cell area set in the data cell area setting step; And

Further, in the two-dimensional code reading method according to the present invention, in the binarized image obtained in the binarizing / quiet zone generating step, a rectangular area including the entire area surrounded by the quiet zone is provided. Determining the approximate two-dimensional code existence area as a two-dimensional code general existence area; and determining the two-dimensional code in the two-dimensional code general existence area for each of two sides of the two-dimensional code general existence area that intersect each other. A two-dimensional code rotation angle determining step for determining a rotation angle, the rotation angle determined in the two-dimensional code rotation angle determining step, and passage of a guide cell tangent tangent to a guide cell arranged in a line in the two-dimensional code A guide cell outer tangent detecting step of detecting the guide cell outer tangent based on a point; Rotation angle obtained in step, and based on the passing point of the timing cell circumscribed line circumscribing the timing cells arranged in a row in said two-dimensional code,
A timing cell outer tangent detecting step of detecting the timing cell outer tangent, and a guide cell outer tangent detected in the guide cell outer tangent detecting step; and a timing cell detected in the timing cell outer tangent detecting step. An area surrounded by the external tangent is defined as a two-dimensional code in the timing cell detection step.

Further, in the two-dimensional code reading method according to the present invention, when a part of the timing cell in the two-dimensional code is lost, the position where the timing cell is lost is calculated, and the calculated position is determined. There is an erasure timing cell complementation step for complementing the timing cell.

Further, in the two-dimensional code reading method according to the present invention, the binarizing / quiet zone generating step may include:
A multi-valued image dividing step of dividing the multi-valued image into a plurality of small regions; and a small region obtained in the multi-valued image dividing step, including a binarized small region to be binarized. A threshold calculation area setting step of setting a threshold calculation area having a larger area than the binarized small area; and the binarization based on a luminance value in the threshold calculation area. A threshold value calculating step of calculating a threshold value for the small area; and comparing the threshold value with the luminance value of each pixel in the binarized small area to determine the binarized small area. And binarizing the small area.

Further, in the two-dimensional code reading method according to the present invention, when a guide cell is arranged at an inner edge of the two-dimensional code, the two-dimensional code general area determining step includes:
Performing a pattern matching process using a mask on the binarized image obtained in the binarization / quiet zone generation step, thereby obtaining a pattern matching value for each coordinate of the binarized image. And a guide cell vertex detection step of detecting, as a guide cell vertex, a point at which the pattern matching value obtained in the pattern matching processing step is maximum,
A rectangular area including the guide cell vertex and including the entire area surrounded by the quiet zone is defined as the two-dimensional code general existence area.

In the two-dimensional code reading method according to the present invention, when the guide cells are arranged in a substantially L-shape,
A substantially L-shaped mask is used as the mask.

Further, in the two-dimensional code reading method according to the present invention, the two-dimensional code rotation angle determining step may include:
For each of the two sides intersecting each other in the two-dimensional code general existence area, the projection axis is rotated by a predetermined angle at a predetermined angle, and the projection processing is performed on the two-dimensional code general existence area. A projection processing step of obtaining a projection intensity on the projection axis, a maximum projection intensity detection step of obtaining a maximum value of the projection intensity for each rotation angle of the projection axis, and a maximum value obtained in the maximum projection intensity detection step A rotation angle detection step of obtaining a rotation angle at which a maximum value having the highest value is obtained, and converting the rotation angle obtained in the rotation angle detection step to each side of the two-dimensional code roughly existing area. To the rotation angle of the two-dimensional code.

In the two-dimensional code reading method according to the present invention, in the step of detecting a tangent line outside the guide cell, the projecting axis is set to each of two mutually intersecting sides of the two-dimensional code roughly existing area. A projection processing step of obtaining a projection intensity on the projection axis by performing projection processing by inclining by a rotation angle of the two-dimensional code corresponding to the side that has been determined in the dimension code rotation angle determination step, A guide cell outer tangent passing point detecting step of obtaining a guide cell outer tangent passing point based on the projection intensity obtained in the projection processing step;
Generating a guide cell outer tangent from the rotation angle of the dimensional code.

Also, in the two-dimensional code reading method according to the present invention, the timing cell external tangent detecting step may include:
For each of the two sides of the two-dimensional code approximate existence area that intersect each other, the projection axis is determined by the rotation angle of the two-dimensional code corresponding to the target side obtained in the two-dimensional code rotation angle determination step. By performing the projection process by tilting, the projection processing step for obtaining the projection intensity on the projection axis, and the timing cell external tangent passing to obtain the passing point of the timing cell external tangent based on the projection intensity obtained in the projection processing step A point detection step, a timing cell outer tangent generating step of generating a timing cell outer tangent from the passing point and the rotation angle of the two-dimensional code,
Having.

In the two-dimensional code reading method according to the present invention, when a plurality of external tangents to the timing cell are generated in the step of generating an external tangent to the timing cell, each of the plurality of external tangents to the generated timing cell is replaced with a temporary timing cell. As the external tangent, the length of each side of the quadrangle formed by the temporary timing cell external tangents that intersect each other and the guide cell external tangents that intersect each other detected in the guide cell external tangent detection step is a design value. And / or both of the second condition that the number of timing cells existing inside the temporary timing cell outer tangent is the largest, and the second condition that is closest to the length of one side of the two-dimensional code. Normal timing cell external tangent selection that selects temporary timing cell external tangents that intersect each other as regular timing cell external tangents It has a step.

Further, in the two-dimensional code reading method according to the present invention, in the timing cell detecting step, the timing cell detecting step includes detecting one of the timing cell external tangents detected in the timing cell external tangent detecting step and the one timing cell external tangent. A straight line translated by one cell toward the guide cell outer tangent opposite to the one timing cell outer tangent, the other timing cell outer tangent, and the guide cell outer tangent facing the other timing cell outer tangent. Projecting the binarized image on the one tangent line outside the timing cell to project the projection intensity on the tangent line outside the one timing cell, and based on the projection intensity obtained in the projection processing step. A threshold calculating step of calculating a threshold by using the threshold, A provisional timing cell detection step for obtaining a cell, a distance between the other timing cell external tangent and a guide cell external tangent facing the other timing cell external tangent, and one side of the two-dimensional code as a design value A calculation timing cell deriving step of obtaining a calculation timing cell from the number of guide cells and the number of timing cells; comparing the calculation timing cell with the calculation timing cell; A normal timing cell setting step as a normal timing cell.

Further, in the two-dimensional code reading method according to the present invention, in the erasure timing cell complementing step, the erasure timing cell corresponding to the erasure timing cell is set as a temporary location of the erasure timing cell. A provisional position determination step, a proximity timing cell determination step for obtaining two regular timing cells adjacent to the same position as the provisional position of the lost timing cell, and an internal dividing point or an external dividing point of the two adjacent regular timing cells. A timing cell complementing step of complementing the timing cell at the position indicated by the dot.

Further, in the two-dimensional code reading method according to the present invention, in the data cell area setting step, the straight line passing through the center or the midpoint of each timing cell detected by the timing cell detecting step and the other straight lines may be used. And a data cell region generating step of generating, as the data cell region, each region having a predetermined area centered on each intersection obtained by the intersection setting step.

Further, in the two-dimensional code reading method according to the present invention, in the data cell logical state identifying step, a labeling process is performed on an area surrounded by the guide cell external tangent and the timing cell external tangent, and the label is processed. A labeling processing step of generating, and among all the labels generated in the labeling processing step, an area of one label is an averaged area calculation step of calculating an averaged area of labels within a predetermined range; and the labeling processing step A label extracting step of extracting a label of a predetermined range centered on the averaging area from the labels generated in the step, a label extracted by the label extracting step, and a data cell area setting step. Data cells that determine the logical state of each data cell area based on the overlap with the data cell area A logical state determination step.

Further, in the two-dimensional code reading method according to the present invention, in the data cell logical state determining step, when the label generated in the labeling processing step covers a plurality of data cell regions by itself, the label A first data cell logical state determining step of determining the logical state of the data cell area where the data area overlaps the widest, and a data cell area of which the logical state is determined to be off in the first data cell logical state determining step On the other hand, when the ratio of the area occupied by the labels other than the single label is equal to or more than a predetermined value, and the label overlaps the center of the data cell region with the ratio of the predetermined area or more, A second data cell logic state determination step of determining the logic state to be on.

Further, a two-dimensional code reading program according to the present invention is for causing a computer to execute a two-dimensional code reading method.

Multivalued Image Binarization Apparatus of the Present Invention, Multivalued Image 2
In the binarization method and the binarization program, a multi-level image dividing means (multi-level image division step) divides a multi-level image including a symbol represented by a combination of a plurality of dots into a plurality of small areas. Then, the threshold value calculating means (threshold value calculating step) includes a binarized small region to be binarized among the small regions and an area larger than the binarized small region. The threshold value of the binarized small area is calculated from the luminance value of the threshold value calculation area having the threshold value, and the small area binarization means (small area binarization step) calculates the threshold value by the threshold value calculation means. The binarized small area is binarized by a threshold value, and at the same time, when a symbol exists in the binarized small area, a quiet zone is generated outside the symbol.

In the multi-value image binarizing device, the multi-value image binarization method and the multi-value image binarization program, the multi-value image dividing means (multi-value image dividing step) includes a two-dimensional code. The multi-valued image is divided into a plurality of small areas, and a threshold value calculating means (threshold value calculating step) includes a binarized small area to be binarized among the small areas, and The threshold value of the binarized small area is calculated from the luminance value of the threshold value calculation area that is larger than the binarized small area. , The binarized small area is binarized, and at the same time, when a two-dimensional code exists in the binarized small area, a quiet zone is generated outside the two-dimensional code.

As described above, the threshold value is calculated based on the luminance value of the threshold value calculation region having a larger area than the binarized small region, and the threshold value is calculated by using the threshold value. By binarizing, a symbol or 2 appears in the binarized small area.
When a dimensional code is present, a quiet zone is generated outside the symbol or the two-dimensional code, so that the outline of the symbol or the two-dimensional code is extremely clear. Therefore, even if the multi-valued image contains noise,
2 such as detection of position, inclination, shape, etc. of dimensional code
There is an effect that image processing after the binarization can be performed accurately. Further, since a threshold value is calculated for each of the divided small areas and the threshold value is used to binarize the binarized small area, it is necessary to accurately binarize even if the multi-valued image has uneven brightness. It has the effect of being able to

As a result, the luminance unevenness and noise of the multi-valued image, the deformation of the two-dimensional code, the deformation such as bleeding or chipping of the cell, the disappearance of the cell, the displacement of the cell, etc. By performing the processing, the two-dimensional code in the multi-valued image can be accurately read.

In the multi-value image binarizing device, the multi-value image binarization method and the multi-value image binarization program, the threshold value calculating means (threshold value calculating step) The arithmetic average of the maximum luminance value and the minimum luminance value is used as the threshold. Since the threshold value calculation region has a larger area than the binarized small region and the threshold value can be obtained by a simple calculation such as averaging the maximum luminance value and the minimum luminance value, A quiet zone can be generated almost certainly and at high speed. Therefore,
The noise of the multi-value image can be removed with high processing efficiency.

In the two-dimensional code reading device, the two-dimensional code reading method and the two-dimensional code reading program,
The binarization / quiet zone generation means (binarization / quiet zone generation step) binarizes the multi-valued image and generates a quiet zone around the two-dimensional code included in the multi-valued image, The cell detecting means (timing cell detecting step) detects a timing cell in the two-dimensional code surrounded by the quiet zone, and the data cell area setting means (data cell area setting step) detects the timing cell in the timing cell detecting section. A predetermined area centered on the intersection of a straight line passing through the center or midpoint of each timing cell and another straight line is set as a data cell area used to identify the logical state of each data cell in the two-dimensional code. In the data cell logic state identification means (data cell logic state identification step), the two-dimensional code By identifying the logical state of each data cell area based on the overlap between the label obtained by performing the labeling process on the area and the data cell area set in the data cell area setting step, a multi-dimensional code including a two-dimensional code is identified. The two-dimensional code is read from the value image.

The binarization / quiet zone generation means (2
In the binarization / quiet zone generation step), a quiet zone is generated around the two-dimensional code.
The contour of the dimensional code becomes very clear, and even if noise is included in the multi-valued image, it is not so affected. Also,
In the data cell area setting means (data cell area setting step), a predetermined area centered on the intersection of a straight line passing through the center or midpoint of the timing cell detected by the timing cell detecting means (timing cell detecting step) and another straight line Is set as a data cell area, and in the data cell logic state identification means (data cell logic state identification step), each data cell area is determined based on the overlap between the label obtained by the labeling process and the set data cell area. , The two-dimensional code can be accurately read from the multi-valued image including the two-dimensional code.

In the two-dimensional code reading device, the two-dimensional code reading method and the two-dimensional code reading program,
In the two-dimensional code approximate existence area determination means (two-dimensional code approximate existence area determination step), a quadrangular area including the entire area surrounded by the quiet zone in the binarized image obtained by the binarization / quiet zone generation means Is defined as a two-dimensional code rotation existence area (two-dimensional code rotation angle determination step), and the two-dimensional code rotation existence determining area (two-dimensional code rotation angle determination step) determines the two-dimensional code rotation existence The rotation angle of the two-dimensional code in the area is determined, and the rotation angle determined by the two-dimensional code rotation angle determination means in the guide cell outer tangent detection means (guide cell outer tangent detection step) and the rotation angle determined by the two-dimensional code are aligned. The tangent to the guide cell is detected based on the passing point of the tangent to the guide cell. In the external tangent detecting means (timing cell external tangent detecting step), the rotation angle obtained by the two-dimensional code rotation angle determining means and the passing point of the timing cell external tangent circumscribing the timing cells arranged in a line in the two-dimensional code are determined. A region surrounded by the guide cell outer tangent detected by the guide cell outer tangent detector and the timing cell outer tangent detected by the timing cell outer tangent detector is detected as a two-dimensional code. I have.

The two-dimensional code approximate existence area determining means (two-dimensional code approximate existence area determination step) extracts a part of the binarized image as a two-dimensional code approximate existence area, and reads the two-dimensional code except for the area. Therefore, it is not affected by noise mixed in an area other than the area. The guide cell outer tangent detecting means (guide cell outer tangent detecting step) detects the guide cell outer tangent based on the rotation angle and the passing point obtained by the two-dimensional code rotation angle determining means (two-dimensional code rotation angle determining step). And timing cell external tangent detection means (timing cell external tangent detection step)
Since the tangent line outside the timing cell is detected based on the rotation angle and the passing point obtained by the two-dimensional code rotation angle determining means (two-dimensional code rotation angle determination step), the position, shape and area of the two-dimensional code can be accurately determined. Can be specified.

In the two-dimensional code reading device, the two-dimensional code reading method and the two-dimensional code reading program,
When a part of the timing cell in the dimensional code has disappeared, a position where the timing cell has disappeared is calculated in the disappearance timing cell complementing step (erasure timing cell complementing means), and the timing cell is complemented to the calculated position. are doing. Therefore, even if the timing cells disappear, the positions of all the timing cells can be accurately obtained.

In the two-dimensional code reading device, the two-dimensional code reading method and the two-dimensional code reading program,
The binarization / quiet zone generation means (binarization / quiet zone generation step) divides the multi-valued image into a plurality of small areas (multi-valued image division step), and A threshold calculation area including a binarized small area to be converted and having a larger area than the binarized small area (threshold calculation area setting step); Based on the luminance value in the area,
Calculating a threshold value for the binarized small area (threshold value calculating step); comparing the threshold value with the luminance value of each pixel in the binarized small area; Is binarized (small area binarization step).

In particular, it is desirable that the binarization / quiet zone generating means (threshold value calculating step) calculates an arithmetic average of the maximum luminance value and the minimum luminance value in the threshold value calculation area as a threshold value. .

As described above, the binarization / quiet zone generation means (binarization / quiet zone generation step)
Since the threshold value is calculated based on the luminance value in the threshold value calculation region having an area larger than that of the binarized small region, and the binarized small region is binarized using the calculated threshold value, A quiet zone is generated around the two-dimensional code. Therefore, even if noise is included in the multi-valued image, the outline of the two-dimensional code becomes extremely clear. Further, since a threshold value is calculated for each of the divided small areas and the threshold value is used to binarize the binarized small area, it is necessary to accurately binarize even if the multi-valued image has uneven brightness. Can be.

In the two-dimensional code reading device, the two-dimensional code reading method and the two-dimensional code reading program,
When a guide cell is placed at the inner edge of the dimension code,
In the two-dimensional code outline area determination means (two-dimensional code outline area determination step), binarization / quiet zone generation means (binarization / quiet zone generation step)
A pattern matching process is performed on the binarized image obtained in step 2 using a mask to obtain a pattern matching value at each coordinate of the binarized image, and a point having the maximum pattern matching value is detected as a guide cell vertex. A quadrangular area including the guide cell vertices and including the entire area surrounded by the quiet zone is defined as the two-dimensional code general existence area.

In particular, in the two-dimensional code reading device, the two-dimensional code reading method and the two-dimensional code reading program, when the guide cells are arranged in a substantially L-shape, a substantially L-shaped mask is used as the mask. desirable.

As described above, since the guide cell vertices are detected by performing the pattern matching process using the mask, the approximate position of the two-dimensional code can be specified even if the two-dimensional code is rotated. In addition, a part of the area in the binarized image is determined as a general existence area of the two-dimensional code, and the area other than the area is not used for reading the two-dimensional code. It is not affected by noise mixed in In other words, this is equivalent to eliminating noise outside the two-dimensional code roughly existing area.

In the two-dimensional code reading device, the two-dimensional code reading method and the two-dimensional code reading program,
In the two-dimensional code rotation angle determining means (two-dimensional code rotation angle determination step), the projection axis is rotated by a predetermined angle by a predetermined angle with respect to each of two mutually intersecting sides of the two-dimensional code roughly existing area, and By performing projection processing on the approximately existing three-dimensional code area, a projection intensity on a projection axis is obtained for each rotation angle (projection processing step),
The maximum value of the projection intensity is obtained for each rotation angle of the projection axis (maximum projection intensity detection step), and the rotation angle at which the highest maximum value among the obtained maximum values is obtained (rotation angle detection step) The rotation angle is defined as the rotation angle of the two-dimensional code with respect to each side of the two-dimensional code approximate existence area.

As described above, in the two-dimensional code rotation angle determining means (two-dimensional code rotation angle determining step), the projection axis is rotated by a predetermined angle, and the projection intensity is obtained by performing the projection processing each time. , The rotation angle having the highest value of the projection intensity is defined as the rotation angle of the two-dimensional code. Therefore, even if noise is included in the two-dimensional code approximate existence region, or individual cells have deformation such as bleeding or sprinkling, or even if individual cells have disappeared, the optimal result is obtained from among a plurality of results. Since the object is selected, the rotation angle can be accurately determined. Also, since the rotation angle of the two-dimensional code is determined independently for each of the two sides of the two-dimensional code approximate existence area that intersect with each other,
Even if the two-dimensional code is deformed due to the influence of printing or distortion of the imaging optical system, the rotation angle of the two-dimensional code can be accurately obtained.

In the two-dimensional code reading device, the two-dimensional code reading method and the two-dimensional code reading program, the guide cell outer tangent detecting means (guide cell outer tangent detecting step) intersects the two-dimensional code roughly existing areas. Projection processing is performed on each of the two sides by tilting the projection axis by the rotation angle of the two-dimensional code corresponding to the target side, thereby obtaining a projection intensity on the projection axis (projection processing step). A passing point of the guide cell outer tangent is obtained based on the obtained projection intensity (guide cell outer tangent passing point detecting step), and a guide cell outer tangent is generated from the passing point and the rotation angle of the two-dimensional code (guide cell). External tangent line generation step).

In particular, the projection intensity is examined in order from the point where the projection intensity is maximum toward the vertex side of the guide cell, and the projection intensity is lower than the first predetermined ratio of the maximum projection intensity for the first time and at a predetermined distance in the same direction. When the projection intensity does not exceed, it is desirable to determine a point that first falls below the first predetermined ratio of the maximum projection intensity as a passing point.

As described above, in the guide cell outer tangent detection means (guide cell outer tangent detection step), the projection axis is inclined by the rotation angle obtained by the two-dimensional code rotation angle determination means (two-dimensional code rotation angle determination step). A point satisfying the condition is determined as a passing point of the guide cell outer tangent based on the projection intensity obtained as a result.
Therefore, even if each guide cell has a deformation such as bleeding or chipping or disappears, it is necessary to obtain the external tangent line of the guide cell independently for each of the two intersecting sides of the two-dimensional code roughly existing area. And, as a result, 2
Even if the dimensional code is deformed due to the influence of printing, distortion of the imaging optical system, or the like, the outer tangent of each guide cell can be accurately obtained.

In the two-dimensional code reading device, the two-dimensional code reading method and the two-dimensional code reading program, the timing cell outside tangent detection means (timing cell outside tangent detection step) intersects the two-dimensional code roughly existing areas. Projection processing is performed on each of the two sides by tilting the projection axis by the rotation angle of the two-dimensional code corresponding to the target side, thereby obtaining a projection intensity on the projection axis (projection processing step). The passing point of the timing cell outer tangent is determined based on the obtained projection intensity (timing cell outer tangent passing point detecting step), and the timing cell outer tangent is generated from the passing point and the rotation angle of the two-dimensional code (timing cell). External tangent line generation step).

In particular, the projection intensity is checked in order from the point where the projection intensity is maximum toward the direction opposite to the vertex of the guide cell, and falls below a second predetermined ratio of the maximum projection intensity for the first time, and the predetermined distance in the same direction is reduced. When the projection intensity does not exceed the maximum projection intensity, it is desirable to determine a point that first falls below the second predetermined ratio of the maximum projection intensity as a passing point.

As described above, in the timing cell outer tangent detecting means (timing cell outer tangent detecting step), the projection axis is inclined by the rotation angle obtained by the two-dimensional code rotation angle determining means (two-dimensional code rotation angle determining step). A projection process is performed, and based on the projection intensity obtained as a result, a point satisfying the condition is determined as a passing point of the timing cell external tangent. Therefore, even if each of the timing cells has a deformation such as bleeding or chipping or disappears, it is necessary to obtain the tangent line of the timing cell independently for each of the two sides of the two-dimensional code roughly existing area that intersect each other. As a result, even if the two-dimensional code is deformed due to the influence of printing, distortion of the imaging optical system, or the like, it is possible to accurately determine the external tangent of each timing cell.

In the two-dimensional code reader, the two-dimensional code reading method and the two-dimensional code reading program, when a plurality of timing cell outer tangents are generated by the timing cell outer tangent detecting means (timing cell outer tangent detecting step), Each of the generated plurality of timing cell external tangents is a temporary timing cell external tangent, and the length of each side of the quadrangle formed by the temporary timing cell external tangent intersecting with each other and the guide cell external tangent intersecting with each other, Either one or both of a first condition that is closest to the length of one side of a two-dimensional code as a design value and a second condition that the number of timing cells existing inside a temporary timing cell outer tangent is the largest Tentative timing cell external tangents that meet each other are selected as regular timing cell external tangents. Therefore, even if timing cells are missing or noise is mixed in the timing cell row,
The external tangent of the timing cell can be accurately detected.

In the two-dimensional code reader, the two-dimensional code reading method and the two-dimensional code reading program, the timing cell detecting means (timing cell detecting step)
, One of the timing cell external tangents detected by the timing cell external tangent detection means (timing cell external tangent detecting step), and the guide cell external tangent facing the one timing cell external tangent from the one timing cell external tangent. The binarized image surrounded by a straight line parallelly moved by one cell to the side, the other timing cell outer tangent, and the guide cell outer tangent facing the other timing cell outer tangent is the one timing cell outer tangent. The projection intensity is calculated on the outer tangent line of the one timing cell (projection processing step), and a threshold value is calculated based on the obtained projection intensity (threshold value calculation step). Is used to binarize the projection intensity to determine a tentative timing cell (a tentative timing cell detection step). Calculating a calculated timing cell from a distance between the outer tangent to the guide cell facing the outer tangent to the imaging cell and the number of guide cells and the number of timing cells included in one side of the two-dimensional code as a design value (calculated timing cell deriving step); The tentative timing cell is compared with the calculated timing cell, and the tentative timing cell corresponding to the calculated timing cell on a one-to-one basis is set as a regular timing cell (regular timing cell setting step).

In particular, it is desirable that the timing cell detecting means (threshold value calculating step) calculate the arithmetic mean value of the maximum luminance value and the minimum luminance value of the projection intensity as a threshold value.

The timing cell detecting means (calculated timing cell deriving step) calculates the position of the calculated timing cell by dividing the distance by the number of timing cells included in one side of the two-dimensional code, and calculates the distance by the two-dimensional code. It is desirable to calculate the width of the calculated timing cell by dividing by the number of guide cells included in one side.

Further, the timing cell detecting means (regular timing cell setting step) determines that the position difference between the calculated timing cell and the provisional timing cell corresponding to the calculated timing cell is smaller than a first predetermined value. conditions,
And a temporary timing cell satisfying one or both of the second condition that a width difference between the width of the calculated timing cell and the width of the temporary timing cell corresponding to the calculated timing cell is smaller than a second predetermined value. Is desirably determined to be a regular timing cell.

When a plurality of timing cell external tangents are detected by the timing cell external tangent detecting means (timing cell external tangent detecting step), each timing cell detected by the timing cell detecting means (normal timing cell detecting step) is detected. The number of regular timing cells in contact with the external tangent is counted (timing cell counting step).

As described above, the timing cell detecting means (timing cell detecting step) binarizes the obtained projection intensity using the threshold value, so that even if small noise is mixed in the timing cell row. , Canceled by the average effect of the projection process. Timing cell external tangent detection means (timing cell external tangent detection step)
By comparing the provisional timing cell obtained based on the external tangent of the timing cell and the calculated timing cell, the provisional timing cell corresponding to the calculation timing on a one-to-one basis is detected as a regular timing cell. . Therefore, even if the timing cell is deformed such as bleeding or chipping, the timing cell can be obtained accurately.

In the two-dimensional code reading device, the two-dimensional code reading method and the two-dimensional code reading program, the position of the calculated timing cell corresponding to the lost timing cell is determined by the lost timing cell complementing means (erasing timing cell complementing step). Is determined as the temporary position of the lost timing cell (temporary position determination step of lost timing cell), and two regular timing cells adjacent to the same position as the temporary position of the lost timing cell are determined (nearly determined timing cell determination step). The timing cell is complemented at the position indicated by the internal dividing point or the external dividing point of the two regular timing cells (timing cell complementing step).

In particular, when two adjacent normal timing cells are located on both sides of the temporary position of the lost timing cell, the lost timing cell complementing means (timing cell complementing step) performs the internal division of the two adjacent normal timing cells. A point is supplemented with a timing cell, and when two adjacent regular timing cells are on one side of the temporary position of the lost timing cell, the timing cell is supplemented at the outside division point of the two adjacent regular timing cells.

Therefore, when there is no regular timing cell corresponding to the calculated timing cell, a timing cell is generated at the position where the lost timing cell was located. Can be determined accurately.

In the two-dimensional code reader, the two-dimensional code reading method and the two-dimensional code reading program, the data cell area setting means (data cell area setting step)
In the above, the intersection of a straight line passing through the center or the midpoint of each timing cell detected by the timing cell detecting means (timing cell detecting step) and another straight line is determined (intersection setting step), and each determined intersection is defined as the center. Each area having the predetermined area is generated as a data cell area (data cell area generating step).

In particular, when the timing cell is complemented by the lost timing cell complementing means (erasing timing cell complementing step), the data cell area setting means (intersection setting step) sets the center of the complemented timing cell or the complemented timing. It is also desirable to determine the intersection of the cell or the timing cell detected by the timing cell detection means (timing cell detection step) with a straight line passing through the middle point.

As described above, since the data cell region is generated around the point where the straight line passing through the center or the middle point of each timing cell intersects, it is necessary to set the region for determining the logical state of the data cell. Can be.

In the two-dimensional code reading device, the two-dimensional code reading method and the two-dimensional code reading program, the data cell logical state identifying means (data cell logical state identifying step) uses the external tangent to the guide cell and the external tangent to the timing cell. A labeling process is performed on the area surrounded by the to generate a label (labeling process step), and, among all the generated labels, an average area of one label having an area within a predetermined range is calculated ( Averaging area calculating step), among the generated labels, extracting a label whose area is within a predetermined range centered on the averaging area (label extracting step), and extracting the extracted label and a data cell area setting means (data cell area The logical state of each data cell area is determined based on the overlap with the data cell area set in the area setting step). Data cell logic state determination step).

In particular, in the two-dimensional code reading device, the two-dimensional code reading method and the two-dimensional code reading program, the data cell logical state identifying means (data cell logical state determination step) is a method in which the extracted label is a single data item. When straddling the cell area, the logic state of the data cell area where the label overlaps the widest is determined to be ON (first data cell logical state determination step). As described above, since the threshold value is not used to determine the logical state of the data cell area, the logical state can be accurately obtained even if there is no label at the center of the data cell area due to lack of data cells, shrinkage, displacement, etc. be able to.

The ratio of the area occupied by the label other than the single label to the data cell region in which the logical state is off is not less than a predetermined value, and the area of the data cell region is larger than the predetermined area at the center of the data cell region. When the labels overlap at the rate of
The logic state of the data cell area is determined to be on (second
Data cell logic state determination step). Therefore, even if adjacent data cells overhang due to bleeding or displacement of the data cells, the logical state can be accurately obtained.

[0099]

Embodiments of the present invention will be described below. In the present invention, a dot is a pigment, a dye or a paint expressed on the surface of an article by a paint or the like,
In such a way that the projections or depressions are provided on the surface of the article, such as a method of displaying dots on an information display device such as a liquid crystal display or a cathode ray tube, etc. What was formed is mentioned. The shape of the dot is not particularly limited.

A symbol is an object of perception used to indicate a certain thing (Iwanami Shoten, Kojien Fifth Edition)
And is a character for expressing various languages, a graphic constituted by a geometric pattern or the like, or a combination thereof.

The multi-valued image is an image in which the luminance value is represented by three or more gradations, and is represented by, for example, a gray level or the like in addition to the binary values of white and black. It is an image.

Next, a data matrix code which is one of the two-dimensional codes will be described as an example. First, FIG.
The configuration of the data matrix code shown in FIG. Data matrix code 1 is data cell 2
And a timing cell 3 and a guide cell 4. The two-dimensional array includes 10 rows × 10 columns to 144 rows × 144 columns. Each cell is represented by a binary value of ON or OFF. In FIG. 1, ON cells are described in black (filled out), and OFF cells are described in white (unfilled).

The guide cell 4 has a plurality of ON cells connected in series and has two substantially linear shapes which are substantially orthogonal to each other to form a substantially L-shape, and detects the position and shape of a two-dimensional code (data matrix code). It is a standard to do. Further, the timing cells 3 are arranged so as to be parallel and opposed to the guide cells 4 and so that ON cells and OFF cells alternately appear, and form two substantially L-shaped dotted lines. The guide cell 4 and the timing cell 3 form four sides of a substantially square, and form the outer edge of the two-dimensional code.

The data cell is a cell arranged in a substantially square area surrounded by the guide cell 4 and the timing cell 3, and the information to be recorded is represented by an ON or OFF cell in binary.

[Multi-Image Binarization] Next, an embodiment of a multi-value image binarization apparatus, a multi-value image binarization method, and a multi-value image binarization program according to the present invention will be described. The first multi-level image binarization program according to the present embodiment causes a computing unit such as a computer to function as a multi-level image binarization apparatus, and includes a symbol represented by a combination of a plurality of dots. A multi-valued image is input and the image is processed.

Further, the second multi-valued image binarization program of the present embodiment comprises a multi-valued image reading device (not shown) provided with an image pickup means, a storage means and a calculation means. It functions as an image binarization device. After imaging a two-dimensional code given to the surface of various articles by an imaging means such as a CCD camera (charge coupled device camera), a storage means such as a magnetic disk or a semiconductor memory Is stored as a multi-valued image, and then the multi-valued image is processed by arithmetic means.

The operation means operated by the first multi-valued image binarization program and the second multi-valued image binarization program of the present embodiment is configured to function as a multi-value image binarization device. The multi-valued image binarizing device includes a multi-valued image dividing means 21, a threshold value calculating means 22, and a small area binarizing means 23, as shown in the flowchart of FIG.
And

Next, the processing (multi-value image binarization method) by the above-described multi-value image binarization apparatus will be described with reference to FIG.
The multi-valued image dividing means 21 divides the multi-valued image 31 into a plurality of small areas 32. Further, the threshold value calculating means 22
Represents a plurality of small areas 32 obtained by dividing the multi-valued image 31
Of the threshold value calculation region 34 that includes the binarized small region 33 to be binarized and has a larger area than the binarized small region, Small area 33
Means for calculating the threshold value. At this time, if the length of one side of the binarized small region 33 is less than three times, the area of the threshold value calculation region 34 may not be appropriately generated as a quiet zone, which will be described later. Small area 3
If the length of one side of 3 is more than 5 times, it will be greatly affected by luminance unevenness and noise around the two-dimensional code.
As for the area of the threshold value calculation obtaining area 34, it is preferable that the length of one side of the binarized small area 33 is 3 to 5 times.

The small area binarizing means 23 compares the threshold value obtained by the threshold value calculating means 22 with the luminance value of each pixel in the binarized small area 33, and performs binarization. It is a means to do. At this time, when a part or the whole of the symbol or the two-dimensional code exists in the binarized small area 33, a quiet zone is generated outside the symbol or the two-dimensional code. The threshold value calculating means 22 and the small area binarizing means 23 are applied to all the small areas obtained by the multi-value image dividing means 21.

Usually, in order to binarize a multi-valued image and detect the position, shape and / or size of a symbol or a two-dimensional code, a blank portion called a quiet zone outside the symbol or the two-dimensional code is used. However, in the case of a conventional program that calculates a threshold using only the luminance value in the binarized small area 33, noise in the multi-valued image affects the calculation of the threshold, There is a problem that noise remains even after binarization. The quiet zone is an area represented by OFF, that is, an area represented by zero when represented by a binary number. FIG. 4 is an explanatory diagram of two examples of the quiet zone. According to the present invention, the quiet zone 36 (enclosed by a broken line) is outside the two-dimensional code (7) or the letter (7) of (a). Region) is generated.

The above-mentioned problem that a quiet zone is not generated and noise remains will be described with reference to FIG. 5 which is an explanatory diagram of a conventional threshold value calculation. Note that the broken line 41 in FIG. 5 represents a change in the luminance value of the multi-valued image, and the broken line 42 is a threshold value calculated from the luminance value 41 of the multi-valued image. A luminance value of 43 indicates noise, and a luminance value of 44 indicates a cell in an ON state.

As in the prior art, when the threshold value of the binarized small area 45 is calculated by using only the luminance value in the binarized small area 45, the noise 43 is affected. The threshold value becomes 42, and as a result, noise 47 remains in the binary image.

On the other hand, as shown in FIG. 6, which is an explanatory diagram of the threshold value calculation according to the present embodiment, the threshold value calculation region 48
Is larger than the binarized small area 45, the difference in luminance value in the threshold value calculation area 48 becomes remarkable, and as a result, the cell 44 in the ON state and the noise 43
Is a value lower than the luminance value of the noise 43. Therefore, when binarization is performed based on the threshold value 42 ′, the noise 43 is removed,
A blank portion, that is, a quiet zone 49, is generated in the binary image 46. in this way,
By binarizing the multi-valued image by the multi-valued image binarizing means of the present embodiment, the multi-valued image can be converted into a binary image, and at the same time, a quiet zone is generated around a symbol or a two-dimensional code. Can be.

At this time, when the threshold value calculating means 22 calculates the threshold value of the binarized small area 45 based on the luminance value in the threshold value calculation area 48, the threshold value calculation area 48
If an arithmetic average value of the maximum luminance value _Lmax and the minimum luminance value _{Lmin in} , that is, ( _Lmax + _Lmin ) ÷ 2, is used as the threshold value, an appropriate threshold value can be calculated at high speed.

As described above, according to the two-dimensional code reading device, the two-dimensional code reading method and the two-dimensional code reading program of the present embodiment, a quiet zone can be generated outside a code such as a symbol or a two-dimensional code. So
The outline of the code becomes extremely clear, and in the subsequent image processing, the position, inclination,
Also, there is an effect that the shape can be accurately detected.
In particular, by performing binarization using the arithmetic mean value of the maximum luminance value and the minimum luminance value in the threshold value calculation region as a threshold value, noise can be removed almost surely and at high speed. Efficiency is improved.

[Two-dimensional code reading]
The two-dimensional code reading apparatus, the two-dimensional code reading method, and the two-dimensional code reading program are intended to accurately read a matrix-type two-dimensional code from a multi-valued image containing a lot of noise.

First, a data matrix code, one of the matrix type two-dimensional codes, will be described. FIG.
Is a substantially L-shaped guide cell row 11 in which a plurality of cells whose logic state is ON (filled) are arranged on two orthogonal sides, and a guide cell row 11
A timing cell row 12 in which cells whose logic state is on and cells in which the logic state is off (no fill) are alternately arranged on the remaining two sides parallel to and opposed to each other, and are substantially surrounded by a guide cell row 11 and a timing cell row 12. And on / off data cells 13 arranged in a grid in a square area. In FIG. 7, ON cells are indicated by black circles, and OFF cells are indicated by white circles. However, the shape of the cells is not limited to a circle but may be a square. In the following description, each cell constituting the guide cell row 11 is referred to as a guide cell, and a cell constituting the timing cell row 12 whose logic state is ON is referred to as a timing cell.

In the figure, reference numeral 14 denotes a quiet zone (to be described later) generated around the guide cell row 11 and the timing cell row 12, that is, around the data matrix code 10. The quiet zone 14 is a blank area of one cell width or more, and is for detecting the position, shape, size, and the like of the data matrix code. Reference numerals 15 and 16 denote external tangents of the guide cell, reference numerals 17 and 18 denote external tangents of the timing cell, and reference numeral 19 denotes a vertex of the guide cell.

Next, an embodiment of a two-dimensional code reading apparatus and a two-dimensional code reading method according to the present invention will be described. Since the two-dimensional code reading program according to the present invention is a program for executing the two-dimensional code reading method, the description is included in the following description of the two-dimensional code reading method. Further, in the present embodiment relating to two-dimensional code reading, a data matrix code will be described as an example of a two-dimensional code, but this data matrix code is referred to as a two-dimensional code unless otherwise specified.

As shown in FIG. 8, the two-dimensional code reader of this embodiment has a multi-valued image binarization / quiet zone generation unit 121 corresponding to the binarization / quiet zone generation means of the present invention. Two-dimensional code approximate existence area determining unit 1 corresponding to two-dimensional code approximate existence area determination means
22, a two-dimensional code rotation angle determining unit 123 corresponding to the two-dimensional code rotation angle determining unit, a guide cell external tangent detecting unit 124 corresponding to the guide cell external tangent detecting unit, and a timing cell external tangent detecting unit. Timing cell external tangent detection section 125, timing cell detection section 126 corresponding to timing cell detection means, and lost timing cell complementation section 127 corresponding to lost timing cell complementation means
And a data cell area setting unit 128 corresponding to the data cell area setting means, and a data cell logical state identifying unit 129 corresponding to the data cell logical state identifying means.

As shown in FIG. 9, the two-dimensional code reading method according to the present embodiment includes a multi-value image binarization / quiet zone generation step S21, a two-dimensional code general existence area determination step S22, and a two-dimensional code Code rotation angle determination step S23, guide cell outer tangent detection step S24
A timing cell external tangent detection step S25, a timing cell detection step S26, a lost timing cell complementation step S27, and a data cell area setting step S25.
28 and a data cell logic state identification step S29. That is, the flowchart shown in FIG. 9 is a main routine of the two-dimensional code reading method, and each step is a subroutine. Hereinafter, each of these subroutine steps will be described in detail.

<Generation of Multivalued Image Binarization / Quiet Zone> First, the multivalued image binarization / quiet zone generation step S21 (multivalued image binarization / quiet zone generation unit 21) will be described. This step is shown in FIG.
FIG. 10 is a step of binarizing a multi-valued image obtained by imaging the image area including the two-dimensional code shown in FIG. 10 with a CCD camera (charge-coupled device camera) or a scanner device, and simultaneously generating a quiet zone. Is comprised of a plurality of sub-steps shown in the flowchart of FIG.

First, in step S31, as shown in FIG. 3, the multivalued image 31 is divided into a plurality of small areas 32 as binarized sections. Next, in step S32, one of the plurality of small regions 32 obtained in step S31 is set as a binarized small region 33 to be binarized, and the binarized small region 33 is included. In addition, a threshold calculation area 34 having an area larger than the binarized small area 33 is determined. Next, in step S33, the arithmetic average of the maximum luminance value and the minimum luminance value in the threshold value calculation area 34 is calculated as a threshold value for the binarized small area 33. Next, in step S34,
The threshold value calculated in step S33 is compared with the luminance value of each pixel in the binarized small area 33 to binarize the binarized small area 33.

In this embodiment, the width of the binarized small area 33 is about several times the cell width, and the width of the threshold value calculation area 34 is about several times that of the binarized small area 33. did. As a result, it is desirable that the area of the threshold value calculation region 34 be set to be 3 to 5 times the binarized small region 33 as shown in FIG. The reason for setting it to 3 to 5 times is that if it is less than 3 times, a quiet zone described later will not be generated properly. If it exceeds 5 times, it will be greatly affected by luminance unevenness and noise around the two-dimensional code. It is because it becomes like this.

In the multivalued image binarization / quiet zone generation step S21, steps S32 to S34 are performed on each small area 32 obtained in step S31. As a result, a blank area having a width of about one cell called a quiet zone as shown by reference numeral 14 in FIG. 7 is generated around the two-dimensional code.

Hereinafter, a mechanism or a mode of generation of a quiet zone will be described with reference to FIGS. 5 and 6. FIG. 5 shows a conventional binarization method when a threshold value calculation area is set at the same position as a binarized small area, and FIG. 6 shows an area larger than the above-described binarized small area. 5 shows a binarization method according to the present invention in which a threshold value calculation region is set.

In FIGS. 5 and 6, a broken line indicated by reference numeral 41 indicates a change in the luminance value of the multi-valued image.
The dashed line indicated by 42 'indicates a threshold value calculated from the luminance value of the multi-valued image, reference numeral 43 indicates noise, reference numeral 44 indicates an example of a cell whose logical state is turned on, and reference numeral 45.
Indicates a binarized small area. Reference numeral 46 indicates a binary image, reference numeral 47 indicates a binary image as noise, reference numeral 48 indicates a threshold value calculation area according to the present invention, and reference numeral 49
Indicates a quiet zone generated in the multi-level image binarization / quiet zone generating step S21.

As shown in FIG. 5, the threshold value calculation region is set at the same position as the binarized small region 45, and the arithmetic mean value of the maximum luminance value and the minimum luminance value in this region is set as the threshold value. 2
If the value is converted, it is affected by the noise 43.
The noise 47 is included in the binary image 46. However, as shown in FIG. 6, by setting an area larger than the binarized small area 45 as the threshold calculation area 48, the difference in luminance value in the threshold calculation area 48 becomes significant. , The threshold value 42 ′ is lower than the luminance value of the noise 43. Therefore, when binarization is performed based on the threshold value 42 ′, noise 47 is eliminated from the binary image 46, and a quiet zone 49 is generated.

As described above, in the multi-valued image binarization / quiet zone generation step S21 (multi-valued image binarization / quiet zone generation unit 21), the threshold is larger than the binarized small region. Since the threshold value is calculated based on the brightness value in the value calculation area and the binarized small area is binarized using the calculated threshold value, a quiet zone is generated around the two-dimensional code. Therefore, even if noise is included in the multi-valued image, the outline of the two-dimensional code becomes extremely clear. Further, since a threshold value is calculated for each of the divided small areas and the threshold value is used to binarize the binarized small area, it is necessary to accurately binarize even if the multi-valued image has uneven brightness. Can be.

<Determination of Schematic Two-Dimensional Code Existing Area> Next, a two-dimensional code schematic existing area determination step S22 (2
The dimension code approximate existence area determination unit 22) will be described. This step is a step of detecting a guide cell vertex of the two-dimensional code and determining a part in the binarized image as a general existence region of the two-dimensional code, and includes a plurality of sub-steps shown in the flowchart of FIG. It is composed of

First, in step S61, using the L-shaped mask (hereinafter referred to as an L-shaped mask) 70 shown in FIG. 12, the binary image obtained in the multi-value image binarization / quiet zone generation step S21 is used. Scans the entire area of the structured image. The L-shaped mask used in this step defines the value of each area so that the matching value when the two-dimensional code overlaps the guide cell is maximized. Also, 2 in the binarized image
In order to allow the dimensional code to be inclined several degrees, for example, ± 5 degrees, and to detect the guide cell apex,
The length 71 of one side is about 2/3 of one side of the two-dimensional code, and the width 72 of the side is about one cell. The mask is not limited to the L-shape, but may be an inverted U-shape or the like.

Next, in step S62, a pattern matching process is performed on the L-shaped mask and the binarized image to obtain a pattern matching value. The pattern matching value pm is obtained by setting (x ₀ , y ₀ ) the coordinates of the vertex 73 of the L-shaped mask shown in FIG.
Let the value of each coordinate of the L-shaped mask be mask (x, y),
The image coordinates overlapping the vertex 73 of the character mask are (i ₀ , j ₀ )
And the luminance value at each coordinate of the binarized image is I (i ₀ ,
j ₀ ), it is obtained by the following equation (1).

Pm = {mask (x + n, y + m) × I (i ₀ + n, j ₀ + m)} (1) (where n = 0, 1,..., M = 0, 1,...)

Next, in step S63, step S6
The point at which the pattern matching value obtained in step 2 becomes the maximum value is detected as a guide cell vertex. Next, step S64
Represents a square area including the guide cell vertices and including the entire area (two-dimensional code) surrounded by the quiet zone.
Defined as the approximate existence area of the dimensional code. At this point, since the size of the two-dimensional code is unknown, FIG.
As shown in the figure, several tens%
The large area is defined as the approximate existence area 81 of the two-dimensional code 82.

As described above, in the two-dimensional code approximate existence area determining step S22 (two-dimensional code approximate existence area determination unit 22), the guide cell apex is detected by performing pattern matching using an L-shaped mask. Therefore, even if the two-dimensional code is rotated several degrees, the approximate position of the two-dimensional code can be determined. In addition, a part of the area in the binarized image is determined as a general existence area of the two-dimensional code, and an area other than the area defined in this step is not processed in subsequent steps. Determining the existence area is equivalent to eliminating noise outside the area. In other words, in the subsequent steps, the area ratio of the two-dimensional code in the image area to be processed increases, so that the noise presence ratio in the image can be relatively reduced.

<Determination of Two-Dimensional Code Rotation Angle>
The dimension code rotation angle determination step S23 (two-dimensional code rotation angle determination unit 23) will be described. This step is a step of determining the rotation angle of the two-dimensional code, and is composed of a plurality of sub-steps shown in the flowchart of FIG. Note that the rotation angle of the two-dimensional code refers to the two-dimensional code general existence area determination step S22.
Refers to the angular deviation from the quadratic two-dimensional code approximate existence area defined in. Further, in the present embodiment, one side (preferably, a lower side) of the two-dimensional code approximate existence area is set as the X axis, and one side (preferably, the left side) orthogonal to the X axis is set as the Y axis.

First, in step S91, the projection axis is rotated by θstep degrees within a range of ± θmax degrees with respect to the X axis, and projection processing is performed on the binarized image of the two-dimensional code approximate existence area. And the projection intensity on the projection axis for each rotation angle. Note that the projection processing is to project in a direction perpendicular to the projection axis set at a certain angle θ, and the projection intensity is such that the logical state is ON (filled) in the direction perpendicular to the projection axis. The higher the number of cells, the higher the value. Next, in step S92, a maximum value Max (θ) of the projection intensity is obtained for each rotation angle θ of the projection axis. Next, in step S93, the maximum value Max having the highest value among the maximum values Max (θ) obtained in step S92.
The rotation angle θ _{0 at} which (θ ₀ ) is obtained is defined as the rotation angle of the two-dimensional code with respect to the X axis.

FIG. 15A shows that the rotation angle of the projection axis is θ _0.
The state of the projection processing when “1” is set is shown. As shown in the figure, the projection intensity is highest at a point where the projection cell intersects the guide cell row on the projection axis. FIG. 15B shows the projection intensity on the projection axis when the rotation angle of the projection axis is not θ ₀ .

The method for determining the rotation angle of the two-dimensional code with respect to the X axis has been described above.
By performing 1 to S93, the rotation angle φ ₀ of the two-dimensional code with respect to the Y axis is also determined. That is, the rotation axis φ ₀ of the two-dimensional code with respect to the Y axis is determined by rotating the projection axis with respect to the Y axis and projecting the binarized image of the two-dimensional code approximate existence area at each rotation angle. .

As described above, the two-dimensional code rotation angle determination step S23 (two-dimensional code rotation angle determination unit 2)
In 3), the projection axis is rotated little by little, and the rotation angle having the highest projection intensity among the projection intensities obtained by performing the projection process each time is set as the rotation angle of the two-dimensional code. Therefore, even if noise is included in the two-dimensional code approximate existence region, or individual cells have deformation such as bleeding or sprinkling, or even if individual cells have disappeared, the optimal result is obtained from among a plurality of results. Since the object is selected, the rotation angle can be accurately determined. Further, since the rotation angle of the two-dimensional code is determined independently for the X axis and the Y axis, for example, a two-dimensional code that is originally a square is deformed into a diamond shape due to the influence of printing, distortion of the imaging optical system, and the like. Even if it is, the rotation angle of the two-dimensional code can be obtained accurately.

<Detection of Guide Cell Outer Tangent Line> Next, the guide cell outer tangent detection step S24 (guide cell outer tangent detector 24) will be described. This step is a step of obtaining an outer tangent of the guide cell included in the two-dimensional code, and is composed of a plurality of sub-steps shown in the flowchart of FIG.

First, in step S111, the projection axis is set to 2
Dimension code rotation angle determination step S23 (two-dimensional code
The rotation angle with respect to the X axis determined by the rotation angle determination unit 23)
Degree θ ₀By performing the projection process by tilting only the projection axis
Obtain the above projected intensity. Next, in step S112,
Guide based on the projection intensity obtained in step S111
Determine the passing point of the cell's outer tangent.

For example, in the case of the projection intensity shown in FIG. 15A, the projection intensity is examined for each pixel in order from the point 101 where the projection intensity is maximum toward the guide cell vertex 102 side.
Projection intensity is the maximum value (projection intensity at the point indicated by reference numeral 101)
Falls for the first time below the predetermined ratio Rg% (103), and
Further, each projection intensity of a predetermined number (Ng) cells in the same direction is
The point 10 below the predetermined ratio Rg% of the maximum value for the first time
In the case where the projection intensity does not exceed the projection intensity of No. 3, a point 103 at which the projection intensity first falls below the predetermined ratio Rg% of the maximum value is determined as a passing point Pxg of the tangent to the guide cell.

Similarly, the projection axis is set to the two-dimensional code rotation angle determination step S23 (two-dimensional code rotation angle determination unit 23).
The projection process is performed by inclining by the rotation angle φ _{0 with} respect to the Y axis determined in the above, thereby obtaining the passing point Pyg of the tangent to the guide cell that intersects the passing point Pxg. From the passing points Pxg and Pyg obtained in this manner, step S11 is performed.
In step 3, an expression representing each tangent to the guide cell is determined.

Y = tan φ ₀ × (x−Pxg) (2) x = tan θ ₀ × (y−Pyg) (3) (where Pxg represents only the x component,
Pxy represents only the y component)

Next, in step S114, the outer tangent of the guide cell, which is represented by the above equations (2) and (3) and which passes through the passing point Pxg and is inclined by the rotation angle φ _{0 with} respect to the Y axis of the two-dimensional code, An external tangent to the guide cell that passes through the point Pyg and is inclined by the rotation angle θ _{0 with} respect to the X axis of the two-dimensional code is generated.

As described above, in the guide cell external tangent detection step S24 (guide cell external tangent detection unit 24), the projection axis is set to the two-dimensional code rotation angle determination step S23.
(Two-dimensional code rotation angle determination unit 23) performs projection processing by inclining by the rotation angle obtained, and determines a point satisfying the condition as a passing point of the guide cell external tangent based on the projection intensity obtained as a result. ing. Therefore, even if each guide cell has a deformation such as bleeding or chipping or disappears, it is possible to obtain the external tangent of the guide cell independently. As a result, for example, even if the originally square two-dimensional code is deformed into a diamond shape due to the influence of printing, distortion of the imaging optical system, or the like, the external tangent of each guide cell can be accurately obtained.

By setting Rg% to, for example, 50% and setting the predetermined number Ng to about the width of a quiet zone (for example, if the quiet zone has a width of one cell, Ng = one cell), it is more accurate. Passing point Pxg, Py
g can be determined.

<Detection of Timing Cell External Tangential Line> Next, the timing cell external tangent detecting step S25 (timing cell external tangent detecting section 25) will be described. This step is a step of detecting an external tangent of the timing cell included in the two-dimensional code, and includes a plurality of sub-steps shown in the flowchart of FIG.

First, in step S121, the projection axis is set to 2
Dimension code rotation angle determination step S23 (two-dimensional code
The rotation angle with respect to the X axis determined by the rotation angle determination unit 23)
Degree θ ₀By performing the projection process by tilting only the projection axis
Obtain the above projected intensity. Next, in step S122,
Time based on the projection intensity obtained in step S121
The passing point of the outer tangent line of the cell is determined.

For example, in the case of the projection intensity shown in FIG. 15A, the projection intensity is examined for each pixel in order from the point 101 where the projection intensity is maximum in the direction opposite to the guide cell vertex 102, and the projection intensity is checked. Is less than a predetermined ratio Rt% of the maximum value (projection intensity of the point indicated by reference numeral 101) (104), and further, each projection intensity of a predetermined number (Nt) of cells in the same direction is a predetermined ratio of the maximum value. If the projection intensity of the point 104 that is lower than Rt% does not exceed the predetermined intensity Rt%, the point 104 where the projection intensity is lower than the predetermined ratio Rt% is determined as the passing point Pxt of the tangent to the timing cell. When there are a plurality of similar points, each point is set as a passing point Pxt (i) (i = 0, 1,...) Of a temporary tangent line outside the timing cell.

Similarly, the projection axis is set to a two-dimensional code rotation angle determination step S23 (two-dimensional code rotation angle determination unit 23).
The projection process is performed by inclining by the rotation angle φ _{0 with} respect to the Y axis determined in the above, thereby obtaining the passing point Ptg of the tangent to the timing cell that intersects the passing point Ptg. Also in this case, when there are a plurality of similar points, each point is set to a passing point Pyt (i) (i = 0, 1,
…). The passing point Pxt, thus obtained,
In step S123, an expression representing each tangent to the timing cell is obtained from Pyt.

Y = tan φ ₀ × {x−Pxt (i)} (4) x = tan θ ₀ × {y−Pyt (i)} (5) (In the above equation, Pxg is only the x component. Represents
Pxy represents only the y component. Also, i = 0, 1,...)

Next, in step S124, it passes through the passing point Pxt (i) expressed by the above equations (4) and (5) and is parallel to one of the external tangents of the guide cell, that is, the inclination is 2
The tangent line of the temporary timing cell having the rotation angle φ ₀ with respect to the Y axis of the two-dimensional code passes through the passing point Pyt (i) and is parallel to the other tangent line of the other guide cell. An external tangent to the temporary timing cell at the angle θ ₀ is generated. At this time, when there are a plurality of passing points of the temporary timing cell outer tangent, a plurality of timing cell outer tangents are generated. For example, as shown in FIG.
The outer tangents 132 and 133 of the temporary timing cell parallel to one guide cell outer tangent 131 and the outer tangent 13 of the temporary timing cell parallel to the other guide cell outer tangent 134.
5,136 is generated.

Next, at step S125, step S125
From the temporary timing cell outer tangents generated at 124, two normal timing cell outer tangents that intersect with the respective guide cell outer tangents are selected. At this time,
The first condition that the length of each side of a rectangle formed by two guide cell external tangents and two intersecting temporary timing cell tangents is closest to the design value of the two-dimensional code, The second that the timing cell is touching
The external tangent to the timing cell that satisfies the above condition is selected as a normal external tangent to the timing cell. In other words, a combination that satisfies the above two conditions is adopted as a combination of normal timing cell external tangents from an arbitrary combination of two temporary timing cell external tangents that intersect each other.

The design value of the two-dimensional code includes information such as the number of pixels on the image, the number of guide cells of the two-dimensional code, the type of the two-dimensional code, the cell area and the color of the cell.
The length of one side of the two-dimensional code can be obtained, for example, by multiplying the number of guide cells of the two-dimensional code by the diameter of the cell.

As described above, the timing cell external tangent detection step S25 (timing cell external tangent detection unit 2)
In 5), the projection axis is subjected to projection processing based on the rotation angle obtained in the two-dimensional code rotation angle determination step S23 (two-dimensional code rotation angle determination unit 23), and a condition is determined based on the projection intensity obtained as a result. Are defined as passing points of the tangent to the outside of the timing cell. Therefore, even if each of the timing cells has a deformation such as bleeding or chipping or disappears, the external tangents of the timing cells can be obtained independently. As a result, for example, even if the originally square two-dimensional code is deformed into a diamond shape due to the influence of printing or distortion of the imaging optical system, the external tangent of each timing cell can be accurately obtained.

When a plurality of passing points of the timing cell external tangent are determined, the timing cell external tangent most suitable to the above two conditions is selected from the combination of the temporary timing cell external tangents that intersect each other. By employing this, this is set as a normal timing cell external tangent. Therefore, even if the timing cell is missing or noise is mixed in the timing cell row, the external tangent of the timing cell can be accurately detected. For example, when the vertical aspect ratio of the two-dimensional code is different, and when any of the above two conditions is satisfied, the temporary timing cell external tangent may be determined as a normal timing cell external tangent.

By setting Rt% to, for example, 30% and setting the predetermined number Nt to be about the width of a quiet zone (for example, if the quiet zone has a width of one cell, Nt = one cell). Passing point Pxg, Py
g can be determined.

<Detection of Timing Cell> Next, a timing cell detection step S26 (timing cell detection section 2)
6) will be described. This step is a step of detecting a timing cell tangent to the external tangent of the timing cell detected by the external cell tangent detecting step S25 (timing cell external tangent detecting unit 25), and includes a plurality of sub-cells shown in the flowchart of FIG. It consists of steps.

First, in step S141, FIG.
As shown in (a), a timing cell external tangent 151 detected in the timing cell external tangent detection step S25 (timing cell external tangent detection unit 25) and about one cell from the timing cell external tangent 151 to the guide cell vertex side. The straight line 152 shifted in parallel by one minute and the other external tangent line 1 to the timing cell
The image surrounded by 53 and the guide cell outer tangent 154 facing the timing cell outer tangent 153 is projected on the timing cell outer tangent 151 side. This step S1
The result of the projection at 41 is shown in FIG. As shown in the figure, the portion where the timing cell exists has a high projection intensity.

Next, in step S142, step S
The arithmetic mean of the maximum value and the minimum value of the projection intensity obtained in 141 is calculated as a threshold for binarization. Next, in step S143, the projection intensity is binarized using the arithmetic average value calculated in step S142, and a temporary timing cell is obtained. The temporary timing cell obtained at this time is shown in FIG. In the figure, the center of the section where the logical state is ON (or) is set to the position P of the temporary timing cell.
a (i), and the width of the section is the width W of the temporary timing cell.
Let a (i). (However, i = 0, 1, ...)

Next, in step S144, the position Pb of the timing cell obtained from the design value of the two-dimensional code
(J) and the width Wb (j) of the timing cell are obtained. (Where j = 0, 1,...) For example, the distance between the temporary tangent line 153 of the timing cell and the tangent line 154 of the guide cell in FIG. The position Pb (j) of the timing cell is calculated by dividing equally by the number of timing cells included in one side, and the design value of the two-dimensional code indicates the distance between the external tangent 153 of the timing cell and the external tangent 154 of the guide cell. The width Wb (j) of the timing cell is obtained by equally dividing the number of guide cells included in one side of the two-dimensional code. This step S144
The position Pb (j) and width W of the timing cell obtained in
FIG. 20D shows b (j). Hereinafter, the position Pb (j)
The timing cell having the width Wb (j) is referred to as a calculated timing cell.

Next, in step S145, step S
The presence / absence of a provisional timing cell that corresponds one-to-one with the calculated timing cell calculated in 144 is determined, and the provisional timing cell that corresponds one-to-one with the calculated timing cell is adopted as a regular timing cell. In this embodiment,
The condition that the position difference between the position Pa (i) of the temporary timing cell and the position Pb (j) of the calculated timing cell is smaller than the allowable error δP, and the width Wa of the temporary timing cell
A temporary timing cell that satisfies the condition that the width difference between (i) and the width Wb (j) of the calculated timing cell is smaller than the allowable error δW is defined as a normal timing cell. In addition,
When the temporary timing cell satisfies one of the conditions, it may be used as a normal timing cell.

The above condition can be expressed by the following equation. The allowable errors δP and δW are, for example, 0.5 cell width. | Pb (j) −Pa (i) | <δP (6) | Wb (j) −Wa (i) | <δW (7) (where i, j = 0, 1,...)

Next, in step S146, step S146
The number of regular timing cells obtained at 145 is counted. The number of the timing cells obtained in step S146 is the largest when the timing cell external tangent is correctly obtained in the timing cell external tangent detection step S25 (timing cell external tangent detection unit 25) (FIG. 20). However, when the timing cell outer tangent is not correctly obtained as shown in FIG. 21, the number of timing cells is smaller than when the timing cell is correctly obtained. If the number of timing cells counted in this step S146 is smaller than a predetermined value, it is determined as "error",
Stop processing.

As described above, in the timing cell detecting step S26 (timing cell detecting unit 26), the timing cell detecting step S25 (timing cell external tangent detecting unit 25) obtains the timing cell external tangent detected based on the timing cell external tangent detected in the timing cell external tangent detecting step S25. The normal timing cell is detected by comparing the obtained provisional timing cell with a calculated timing cell obtained based on the design value of the two-dimensional code. Therefore, even if noise is mixed in the timing cell row or the timing cell is deformed such as bleeding or chipping, the timing cell can be accurately obtained.

More specifically, the mixing of noise is canceled by binarizing the projection intensity in step S142 by the average effect of the projection processing in which the logical state is not turned on if the noise is small, and the timing cell Deformation such as bleeding or chipping does not affect the detection of normal timing cells if the bleeding or chipping is a little, because the temporary timing cell and the calculated timing cell are compared in step S145.

<Completion of Erasure Timing Cell> Next, the erasure timing cell complementation step S27 (erasure timing cell complementer 27) will be described. This step is a step for complementing the lost timing cell of the two-dimensional code, and includes a plurality of sub-steps shown in the flowchart of FIG.

First, in step S161, for example, FIG.
As shown in FIG. 3, when there is no normal timing cell corresponding to the calculated timing cell calculated in step S144, the position Pb (k) of the calculated timing cell corresponding to the non-existent normal timing cell is replaced by the lost timing cell. As a temporary position of Next, step S1
In step 62, the positions Pa (k ') and Pa (k ") of regular timing cells which are in the same position as the position Pb (k) of the calculated timing cell and which are close to the non-existent timing cell are obtained. , Pa (k ′), Pa of the normal timing cell obtained in step S162
An inner dividing point Pa (k) or outer dividing point Pa (k) of (k ″) is obtained, and a timing cell is generated at that position.

When calculating the position Pa (k) of the timing cell which has disappeared due to the internal division, it is possible to obtain Pa (k) = {Pa (k ′) + Pa (k ″)} / 2 (8) When obtaining the position Pa (k) of the timing cell which has been lost due to the external part, it can be obtained by the following expression: Pa (k) = 2 × Pa (k ′) + Pa (k ″) (9) However, Pa (k ') <P
a (k) <Pa (k ″), and k = 0, 1,..., N−
1 (where N is the number of timing cells in the design)
k ′, k ″ = 0, 1,...

As described above, the erasure timing cell complementation step S27 (the erasure timing cell complementer 27)
In, when there is no regular timing cell corresponding to the calculated timing cell, the timing cell is generated at the position where the lost timing cell was. Therefore, even if the timing cells disappear, the positions of all the timing cells can be accurately obtained.

<Setting of Data Cell Area> Next, a data cell area setting step S28 (data cell area setting section 2)
8) will be described. This step is a step of setting a data cell area used for identifying the logical state of the data cell of the two-dimensional code, and includes a plurality of sub-steps shown in the flowchart of FIG.

First, in step S171, as shown in FIG. 25, a straight line 1 passing through the center of each timing cell in one timing cell row and parallel to the other timing cell row
81, or a straight line 182 that passes through the midpoint of each timing cell in one of the timing cell rows, and passes through the center of each timing cell in the other timing cell row, and Parallel straight line 1
An intersection 185 with the straight line 184 that passes through the center point of each timing cell of the other timing cell column 83 or 83 and is parallel to one timing cell column is obtained and set as the center point of each data cell region.

Next, in step S172, the distance D1 between one external tangent to the timing cell and the external tangent to the guide cell facing it, and the distance D2 between the external tangent to the other timing cell and the external tangent to the guide cell facing it. Are obtained by dividing the distances D1 and D2 by the number of guide cells included in one side of the two-dimensional code indicated by the design value of the two-dimensional code.
2 is defined as the length of one side, and an area around the intersection 185 is generated as a data cell area 186. Although FIG. 25 specifically shows only the data cell area 186 centered on the intersection 185, a plurality of data cell areas similar to the data cell area 186 are arranged in a matrix.

As described above, in the data cell area setting step S28 (data cell area setting section 28), a data cell area centered on a point where a straight line passing through the center or the midpoint of each timing cell intersects is generated. Therefore, an area for determining the logic state of the data cell can be set.

<Identification of Data Cell Logic State> Finally, the data cell logic state identification step S29 (data cell logic state identification unit 29) will be described. The steps are:
This is a step of identifying the logical state of the data cell in the two-dimensional code as either ON or OFF, and includes a plurality of sub-steps shown in the flowchart of FIG.

First, in step S191, a label is generated by performing labeling processing in an area surrounded by two guide cell outer tangents and timing cell outer tangent. Next, in step S192, out of all the labels obtained in step S191, the area of one label is R1% or more and R2% or less (for example, 20% or more and 200% or less) of the cell area indicated by the design value of the two-dimensional code. ) Is added, and the added value is divided by the number of added labels to calculate an added average area S ₀ . Next, in step S193, of all labels generated in step S191, the area is ± R3% relative summing the average area S ₀ (e.g., ± 30%) to extract the label of the extracted label independent Considered to represent one data cell.

In step S192, when individual data cells are obtained with values close to the design value of the two-dimensional code as shown in FIG. 27A, when adding the label area, the cell area is close to the cell area. It is desirable to add only labels having an area.

Also, in step S192, the addition average area S
By calculating ₀ , it is possible to determine the data cell according to the brightness of the read two-dimensional code. That is, actually, as shown in FIG. 27, not only the case where one label area is equal to the cell area indicated by the design value (a) but also the case where one label area is equal to the cell area indicated by the design value due to insufficient exposure or the like. In some cases, the entire label area is smaller than the area (b), and in another case, the entire label area is larger than the cell area indicated by the design value due to excessive exposure or the like (c).

However, in the case shown in FIGS. 27B and 27C, there is a possibility that the data cell cannot be accurately determined from the label generated in step S191. Therefore, by extracting the label using the averaging area S ₀ calculated in step S192 rather than using the one cell area indicated by the design value in step S193, one data cell can be extracted regardless of the exposure state. Can be determined.

Next, in step S194, as the first data cell logical state determination, the label extracted in step S193 and the data cell area generated in data cell area setting step S28 (data cell area setting unit 28) are determined. , The logic state of the data cell area is determined. For example, when a label exists in one data cell area, the logic state of this data cell area is determined to be ON. However, as shown in FIG. 28A, when the label 201 extends over a plurality of data cell areas, the logic state of the data cell area 202 where the label 201 representing the data cell overlaps the widest is set to ON. .

In this step S194, since the threshold value is not used for determining the logical state of the data cell area, as shown in FIG. 28B, even if the area of one label is small due to insufficient exposure, etc. The logic state of a data cell region having a large occupied area is determined to be on. Therefore, even if the label is small and there is no label at the center of the data cell area, the logical state can be accurately obtained.

Next, in step S195, as the second data cell logic state determination, the center of the data cell area 203 (for example, the data cell width) is determined for each data cell area whose logic state is determined to be off in step S194. 2
0%) When the image of 204 is ON, that is, in the state where a label is present, and when the pixels in the ON state are equal to or more than Rth% (for example, 30%) of all the pixels in the same data cell area 203, the same data The logic state of the cell region 203 is set to ON. In this step S195, as shown in FIG. 28C, even if the area of one label is large due to overexposure or the like, the logical state of the data cell area is not turned on unless the center of the data cell area is turned on. Is not determined to be ON, the logical state of the data cell area can be accurately obtained even if adjacent data cells protrude due to bleeding or misalignment of the data cells.

As described above, data cell logic state identification step S29 (data cell logic state identification unit 29)
Then, the label is extracted using the averaging area S ₀ according to the size of the label, and the extracted label is regarded as a data cell. When the label singly extends over a plurality of data cell areas, the label is most widely used. The logic state of the overlapping data cell region is set to ON, and a plurality of labels having a large area
When a part of one data cell region is turned on, the logic state of the data cell region is turned on unless the center of the data cell region is turned on even if the area occupied by the label is equal to or larger than the threshold value. Does not judge the state.

Therefore, the logic state of the data cell area can be accurately obtained even if there is no data cell at the center of the data cell area due to lack, reduction, displacement, or the like of the data cell. Further, even if adjacent data cells overhang due to bleeding, enlargement, displacement, or the like of the data cells, the logical state of the data cell area can be accurately obtained.

Note that the first data cell logical state determination in step S194 determines the distance from the center point of the data cell region to the label, that is, the distance from the center point to the pixel in the ON state, regardless of the occupied area. May be calculated, and the logical state of the data label area in which the shortest distance is obtained may be determined to be on.

As described above, in each step (part),
Perform desired processing without being affected by luminance unevenness and noise of multi-valued images, deformation of two-dimensional code, deformation such as bleeding or chipping of cells, loss of cells, displacement of cells, etc. Thereby, the two-dimensional code in the multi-valued image can be accurately read.

[0189]

As described above, according to the multi-value image binarizing apparatus, the multi-value image binarization method and the multi-value image binarization program of the present invention, the multi-value image dividing means (multi-value image Dividing step), a multi-valued image including a symbol represented by a combination of a plurality of dots is divided into a plurality of small areas,
In the threshold value calculating means (threshold value calculating step),
Among the small regions, the binarized small region that includes the binarized small region to be binarized and has a larger area than the binarized small region is calculated based on the luminance value of the threshold calculation region. And in the small area binarizing means (small area binarizing step), the binarized small area is binarized by the threshold value calculated by the threshold value calculating means. When a symbol exists in the binarized small area, a quiet zone is generated outside the symbol.

Further, the multi-valued image including the two-dimensional code is divided into a plurality of small areas, and the threshold value calculating means (threshold value calculating step) selects the two-dimensional image to be binarized among the small areas. The threshold value of the binarized small area is calculated from the luminance value of the threshold value calculation area that includes the binarized small area and is larger than the binarized small area, In the region binarizing step), the binarized small region is binarized by a threshold value, and if a two-dimensional code exists in the binarized small region, a quiet outside of the two-dimensional code is performed. Creating a zone.

As described above, the threshold value is calculated based on the luminance value of the threshold value calculation region having a larger area than the binarized small region, and the threshold value is calculated using the threshold value. By binarizing, a symbol or 2 appears in the binarized small area.
When a dimensional code is present, a quiet zone is generated outside the symbol or the two-dimensional code, so that the outline of the symbol or the two-dimensional code is extremely clear. Therefore, even if the multi-valued image contains noise,
2 such as detection of position, inclination, shape, etc. of dimensional code
There is an effect that image processing after the binarization can be performed accurately. Further, since a threshold value is calculated for each of the divided small areas and the threshold value is used to binarize the binarized small area, it is necessary to accurately binarize even if the multi-valued image has uneven brightness. It has the effect of being able to

As a result, desired effects can be obtained without being affected by luminance unevenness or noise of a multi-valued image, deformation of a two-dimensional code, deformation such as bleeding or chipping of a cell, loss of a cell, or displacement of a cell. By performing the processing, the two-dimensional code in the multi-valued image can be accurately read.

According to the two-dimensional code reading device, the two-dimensional code reading method and the two-dimensional code reading program of the present invention, the binarization / quiet zone generation means (binarization / quiet zone generation step) The multi-level image is binarized, a quiet zone is generated around the two-dimensional code included in the multi-level image, and the timing cell detecting means (timing cell detecting step) detects the two-dimensional area surrounded by the quiet zone. A timing cell in the code is detected, and a data cell area setting means (data cell area setting step) determines an intersection of a straight line passing through the center or midpoint of each timing cell detected by the timing cell detection unit with another straight line. The predetermined area at the center is used to identify the logical state of each data cell in the two-dimensional code. A label obtained by labeling the two-dimensional code area in the data cell logic state identification means (data cell logic state identification step), and the data set in the data cell area setting step. The two-dimensional code is read from the multi-valued image including the two-dimensional code by identifying the logical state of each data cell area based on the overlap with the cell area.

The binarization / quiet zone generation means (2
In the binarization / quiet zone generation step), a quiet zone is generated around the two-dimensional code.
The contour of the dimensional code becomes very clear, and even if noise is included in the multi-valued image, it is not so affected. Also,
In the data cell area setting means (data cell area setting step), a predetermined area centered on the intersection of a straight line passing through the center or midpoint of the timing cell detected by the timing cell detecting means (timing cell detecting step) and another straight line Is set as a data cell area, and in the data cell logic state identification means (data cell logic state identification step), each data cell area is determined based on the overlap between the label obtained by the labeling process and the set data cell area. , The two-dimensional code can be accurately read from the multi-valued image including the two-dimensional code.

In the two-dimensional code approximate existence area determining means (two-dimensional code approximate existence area determination step),
In the binarized image obtained by the binarization / quiet zone generating means, a quadrangular region including the entire region surrounded by the quiet zone is determined as a two-dimensional code general existence region,
In the two-dimensional code rotation angle determining means (two-dimensional code rotation angle determining step), the rotation angle of the two-dimensional code in the two-dimensional code general existence area is determined for each of two substantially orthogonal sides of the two-dimensional code general existence area. In the guide cell outer tangent detecting means (guide cell outer tangent detecting step), the rotation angle obtained by the two-dimensional code rotation angle determining means and the guide cell outer tangent tangent to the guide cells arranged in a line in the two-dimensional code are determined. A guide cell outer tangent is detected based on the passing point, and a timing cell outer tangent detecting means (timing cell outer tangent detecting step) detects the rotation angle obtained by the two-dimensional code rotation angle determining means;
A timing cell outer tangent is detected based on a passing point of the timing cell outer tangent circumscribing the timing cells arranged in a line in the dimension code, and the guide cell outer tangent detected by the guide cell outer tangent detecting means and the timing cell outer tangent are detected. The area surrounded by the timing cell external tangent detected by the line detecting means is defined as a two-dimensional code.

The two-dimensional code approximate existence area determining means (two-dimensional code approximate existence area determination step) extracts a part of the binarized image as a two-dimensional code approximate existence area, and reads the two-dimensional code other than the area. Therefore, it is not affected by noise mixed in an area other than the area. The guide cell outer tangent detecting means (guide cell outer tangent detecting step) detects the guide cell outer tangent based on the rotation angle and the passing point obtained by the two-dimensional code rotation angle determining means (two-dimensional code rotation angle determining step). And timing cell external tangent detection means (timing cell external tangent detection step)
Since the tangent line outside the timing cell is detected based on the rotation angle and the passing point obtained by the two-dimensional code rotation angle determining means (two-dimensional code rotation angle determination step), the position, shape and area of the two-dimensional code can be accurately determined. Can be specified.

Further, when a part of the timing cell in the two-dimensional code is lost, a position where the timing cell is lost is calculated in a lost timing cell complementing step (erasing timing cell complementing means). The position is complemented by a timing cell. Therefore, even if the timing cells disappear, the positions of all the timing cells can be accurately obtained.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram showing a data matrix code.

FIG. 2 is a flowchart illustrating a flow of processing in the multi-value image binarization device.

FIG. 3 is an explanatory diagram showing processing in a multi-value image binarization device.

FIG. 4 is an explanatory diagram illustrating a binary image in which a quiet zone is generated.

FIG. 5 is an explanatory diagram of binarization using a threshold value obtained by a conventional threshold value calculation method.

FIG. 6 is an explanatory diagram of binarization based on a threshold value obtained by the threshold value calculation method of the present invention.

FIG. 7 is an explanatory diagram showing a data matrix code.

FIG. 8 is an explanatory diagram illustrating a two-dimensional code reader according to the present invention.

FIG. 9 is a flowchart showing a main routine of the two-dimensional code reading method according to the present invention.

FIG. 10 is a flowchart showing a multilevel image binarization / quiet zone generation step S21 which is one of subroutines of a two-dimensional code reading method.

FIG. 11 is a flowchart showing a two-dimensional code approximate existence area determining step S22 which is one of subroutines of a two-dimensional code reading method.

FIG. 12 is a schematic two-dimensional code existence area determining step S2.
FIG. 4 is an explanatory diagram showing an L-shaped mask used in No. 2;

FIG. 13 is a diagram showing a two-dimensional code approximate existence area determining step S2.
FIG. 4 is an explanatory diagram showing a schematic existence area of a two-dimensional code determined by 2;

FIG. 14 is a flowchart showing a two-dimensional code rotation angle determining step S23 which is one of subroutines of a two-dimensional code reading method.

15A is an explanatory diagram showing a state of projection processing when the rotation angle of the projection axis is set to θ _0, and FIG. 15B is a diagram illustrating a state on the projection axis when the rotation angle of the projection axis is not θ _0. FIG. 9 is an explanatory diagram showing the projection intensity of the image.

FIG. 16 is a flowchart showing a guide cell external tangent detection step S24 which is one of subroutines of a two-dimensional code reading method.

FIG. 17 is a flowchart showing a timing cell external tangent detection step S25, which is one of subroutines of a two-dimensional code reading method.

FIG. 18 is an explanatory diagram showing a temporary timing cell outer tangent including a guide cell outer tangent and a normal timing cell outer tangent of a two-dimensional code.

FIG. 19 is a flowchart showing a timing cell detection step S26 which is one of subroutines of a two-dimensional code reading method.

FIG. 20 is an explanatory view showing position detection of a timing cell based on a timing cell outer tangent obtained correctly.

FIG. 21 is an explanatory diagram showing timing cell position detection based on a timing cell external tangent that is not correctly obtained.

FIG. 22 is a flowchart showing a lost timing cell complementing step S27, which is one of subroutines of a two-dimensional code reading method.

FIG. 23 is an explanatory diagram showing generation of a timing cell when the timing cell has disappeared.

FIG. 24 is a flowchart showing a data cell area setting step S28 which is one of subroutines of a two-dimensional code reading method.

FIG. 25 is an explanatory diagram showing a data cell area generated in a two-dimensional code.

FIG. 26 is a flowchart showing a data cell logical state identification step S29 which is one of subroutines of a two-dimensional code reading method.

FIG. 27A shows a two-dimensional code when the label area is substantially equal to the cell area indicated by the design value of the two-dimensional code,
(B) shows a two-dimensional code when the label area is smaller than the cell area indicated by the design value as a whole,
(C) is an explanatory view showing a two-dimensional code when the label area is larger than the cell area indicated by the design value as a whole.

FIG. 28A shows an example of a relationship between an example of a label and a data cell area, FIG. 28B shows an example of a two-dimensional code when the area of one label is small, and FIG. FIG. 4 is an explanatory diagram showing an example of a two-dimensional code when the area of the is large.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Data matrix code 2 Data cell 3 Timing cell 4 Guide cell 21 Multi-valued image division means 22 Threshold value calculation means 23 Small area binarization means 31 Multi-valued image 32 Small area 33 Binarized small area 34 Threshold Calculation area 35 Character 36 Quiet zone 121 Multi-valued image binarization / quiet zone generator 122 Two-dimensional code approximate existence area determiner 123 Two-dimensional code rotation angle determiner 124 Guide cell external tangent detector 125 Timing cell external tangent detector 126 Timing cell detection unit 127 Erasure timing cell complement unit 128 Data cell area setting unit 129 Data cell logical state identification unit

Claims (38)

[Claims] 1. A multi-level image binarizing apparatus for binarizing a multi-level image including a symbol represented by a combination of a plurality of dots, wherein the multi-level image is divided into a plurality of small areas. Value image dividing means, and a luminance value of a threshold value calculation region including a binarized small region to be binarized among the small regions and having an area larger than the binarized small region A threshold value calculating means for calculating a threshold value of the binarized small area from a threshold value calculated by the threshold value calculating means; A multi-valued image binarizing apparatus, comprising: a small area binarizing means for generating a quiet zone outside a symbol when the symbol exists in the binarized small area. 2. A multi-valued image binarization device for binarizing a multi-valued image including a two-dimensional code, the multi-valued image dividing means for dividing the multi-valued image into a plurality of small areas, The threshold of the binarized small area is determined based on the luminance value of the threshold value calculation area which is larger than the binarized small area and which includes the binarized small area to be binarized. Threshold value calculating means for calculating a value; and, when the two-dimensional code is present in the binarized small area, the two-dimensional code is And a small area binarizing unit for generating a quiet zone outside the multi-valued image. 3. The threshold value calculation unit according to claim 1, wherein the threshold value is an arithmetic average of a maximum luminance value and a minimum luminance value in the threshold value calculation area. The multivalued image binarization device according to the above. 4. A multi-valued image binarization method for binarizing a multi-valued image including a symbol represented by a combination of a plurality of dots, wherein the multi-valued image is divided into a plurality of small areas. Value image dividing step; and a luminance value of a threshold value calculation region including a binarized small region to be binarized among the small regions and having an area larger than the binarized small region. A threshold value calculating step of calculating a threshold value of the binarized small region from the following; and, at the same time, binarizing the binarized small region with the threshold value calculated by the threshold value calculating means, A binarizing step of generating a quiet zone outside the symbol when the symbol exists in the binarized small region. 5. A multi-valued image binarization method for binarizing a multi-valued image including a two-dimensional code, the method comprising: dividing a multi-valued image into a plurality of small regions; The threshold of the binarized small area is determined based on the luminance value of the threshold value calculation area which is larger than the binarized small area and which includes the binarized small area to be binarized. A threshold value calculating step of calculating a value; and, when the two-dimensional code is present in the binarized small area, the two-dimensional code is A small area binarizing step of generating a quiet zone outside of the multi-valued image. 6. The threshold value calculating step, wherein the threshold value is an arithmetic average of a maximum luminance value and a minimum luminance value in the threshold value calculation region.
Or the multilevel image binarization method according to 5. 7. A multi-level image binarization program for causing a computer to execute the multi-level image binarization method according to claim 4. 8. A multi-valued image including a two-dimensional code,
A two-dimensional code reader for reading a two-dimensional code, wherein the two-dimensional image is binarized and a quiet zone is generated around a two-dimensional code included in the multi-value image. Means, a timing cell detecting means for detecting a timing cell in the two-dimensional code surrounded by the quiet zone, a straight line passing through a center or a midpoint of each timing cell detected by the timing cell detecting means, and the other Data cell area setting means for setting a predetermined area centered on an intersection with a straight line as a data cell area used for identifying a logical state of each data cell in the two-dimensional code; Of the label obtained by performing the labeling process on the data cell area set by the data cell area setting means. 2. A two-dimensional code reader, comprising: data cell logical state identification means for identifying a logical state of each data cell area based on a preamble. 9. A two-dimensional area which defines a quadrangular area including the entire area surrounded by the quiet zone in the binary image obtained by the binary / quiet zone generating means as a two-dimensional code roughly existing area. Code approximate existence area determining means; two-dimensional code rotation angle determining means for determining a rotation angle of a two-dimensional code in the two-dimensional code approximate existence area with respect to each of two sides of the two-dimensional code approximate existence area that intersect each other; A rotation angle obtained by the two-dimensional code rotation angle determining means;
And a guide cell outer tangent detecting means for detecting the guide cell outer tangent based on a passing point of a guide cell outer tangent circumscribing the guide cells arranged in a line in the two-dimensional code; and determining the two-dimensional code rotation angle. Rotation angle obtained by means,
And a timing cell external tangent detecting means for detecting the timing cell external tangent based on a passing point of a timing cell external tangent circumscribing the timing cells arranged in a line in the two-dimensional code. The area surrounded by the guide cell outer tangent detected by the line detector and the timing cell outer tangent detected by the timing cell outer tangent detector is defined as
The two-dimensional code reader according to claim 8, wherein the two-dimensional code reader is a two-dimensional code. 10. When a part of a timing cell in a two-dimensional code is lost, a position where the timing cell is lost is calculated, and a lost timing cell complementer for complementing the timing cell with the calculated position is provided. The two-dimensional code reader according to claim 9, further comprising: 11. The binarizing / quiet zone generating means divides a multi-valued image into a plurality of small areas, and converts the multi-valued image into a plurality of small areas, and the binarized object to be binarized. A threshold calculation region including a small region and having a larger area than the binarized small region is set, and the threshold value calculation region is set based on a luminance value in the threshold calculation region.
Calculating a threshold value for the binarized small area, binarizing the binarized small area by comparing the threshold value with the luminance value of each pixel in the binarized small area. The two-dimensional code reader according to claim 8, 9 or 10, wherein: 12. When a guide cell is arranged at an inner edge of the two-dimensional code, the two-dimensional code general area determining means determines a binarized image obtained by the binarization / quiet zone generation means. By performing a pattern matching process using a mask, a pattern matching value of each coordinate of the binarized image is obtained, a point having the maximum pattern matching value is detected as a guide cell vertex, and the guide cell vertex is included. 12. The two-dimensional code reading device according to claim 9, wherein a quadrangular area including the entire area surrounded by the quiet zone is determined as the two-dimensional code general existence area. 13. The two-dimensional code reader according to claim 12, wherein when the guide cells are arranged in a substantially L shape, a substantially L-shaped mask is used as the mask. 14. The two-dimensional code rotation angle determining means rotates a projection axis by a predetermined angle at a predetermined angle with respect to each of two mutually intersecting sides of the two-dimensional code roughly existing area. By performing projection processing on the dimensional code approximate existence area, a projection intensity on the projection axis is obtained for each rotation angle, a maximum value of the projection intensity is obtained for each rotation angle of the projection axis, and A rotation angle at which a maximum value having the highest value is obtained, and the rotation angle is a rotation angle of the two-dimensional code with respect to each side of the two-dimensional code approximate existence area. 14. The two-dimensional code reader according to 11, 12, or 13. 15. The guide cell outer tangent detection means, for each of two sides of the two-dimensional code roughly existing area, which intersects each other, sets the projection axis to the two-dimensional code corresponding to the target side. By performing projection processing by inclining by the rotation angle of, the projection intensity on the projection axis is obtained, and the passing point of the guide cell external tangent is obtained based on the projection intensity, and the rotation angle of the passing point and the two-dimensional code is obtained. 10. A guide cell external tangent is generated from the following.
The two-dimensional code reader according to 0, 11, 12, 13 or 14. 16. The timing cell outer tangent detection means, wherein the projection axis is determined by the two-dimensional code rotation angle determination means for each of two mutually intersecting sides of the two-dimensional code existence area. By performing projection processing by inclining by the rotation angle of the two-dimensional code corresponding to the side being determined, a projection intensity on the projection axis is determined, and a passing point of a timing cell external tangent is determined based on the projection intensity, 16. The method according to claim 9, wherein a tangent line external to the timing cell is generated from the passing point and the rotation angle of the two-dimensional code.
Dimension code reader. 17. When a plurality of timing cell external tangents are generated by the timing cell external tangent detecting means, the timing cell external tangent detecting means replaces each of the generated timing cell external tangents with a temporary timing cell external tangent. The length of each side of the quadrangle formed by the intersecting temporary cell tangents intersecting each other and the intersecting guide cell external tangents detected by the guide cell external tangent detecting means is two-dimensional as a design value. Intersecting each other which satisfies one or both of the first condition that is closest to the length of one side of the code and the second condition that the number of timing cells existing inside the tentative timing cell outer tangent is the largest. 2. The method according to claim 1, wherein the tentative timing cell external tangent is selected as a normal timing cell external tangent.
6. The two-dimensional code reader according to 6. 18. The timing cell detecting means includes: a timing cell external tangent detected by the timing cell external tangent detecting means; and a guide cell facing the one timing cell external tangent from the one timing cell external tangent. The binarized image surrounded by the straight line translated by one cell toward the external tangent side, the other external tangent to the timing cell, and the external tangent to the guide cell opposite to the external tangent to the other timing cell, Projecting to the outer tangent side,
Calculating the projection intensity on the outer tangent to the one timing cell; calculating a threshold based on the projection intensity; binarizing the projection intensity using the threshold to obtain a temporary timing cell; Timing calculated from the distance between the external tangent of the other timing cell and the external tangent of the guide cell opposed to the external tangent of the other timing cell, and the number of guide cells and the number of timing cells included in one side of the two-dimensional code as a design value 18. A cell is obtained, and the temporary timing cell and the calculated timing cell are compared, and a temporary timing cell corresponding to the calculated timing cell on a one-to-one basis is set as a normal timing cell. 2. The two-dimensional code reader according to claim 1. 19. The lost timing cell complementing means sets a position of the calculated timing cell corresponding to the lost timing cell as a temporary position of the lost timing cell, and sets the calculated timing cell to a position close to the same position as the temporary position of the lost timing cell.
19. The two-dimensional code reading apparatus according to claim 18, wherein two regular timing cells are obtained, and the timing cell is complemented at a position indicated by an internal dividing point or an external dividing point of the two adjacent regular timing cells. 20. The data cell area setting means obtains an intersection between a straight line passing through the center or midpoint of each timing cell detected by the timing cell detecting means and another straight line, and determines a predetermined point around each intersection. 9. An area having an area is generated as the data cell area.
9, 10, 11, 12, 13, 14, 15, 16, 1
20. The two-dimensional code reader according to 7, 18 or 19. 21. The data cell logical state identification means generates a label by performing labeling processing on an area surrounded by the guide cell external tangent and the timing cell external tangent, and generates a label for all the generated labels. The area of one label calculates the average area of the labels within a predetermined range, and among the generated labels, extracts the label of the predetermined range whose area is centered on the average area. The logic state of each data cell area is determined based on the overlap between the label and the data cell area set by the data cell area setting means. 13, 14, 15, 1
The two-dimensional code reader according to 6, 17, 18, 19 or 20. 22. The data cell logical state identifying means, when the generated label alone extends over a plurality of data cell areas, turns on the logical state of the data cell area where the label overlaps most widely. In the determination, the ratio of the area occupied by the label other than the single label to the data cell region in which the logical state is determined to be OFF is equal to or greater than a predetermined value, and is equal to or greater than a predetermined value in the center of the data cell region. 22. The two-dimensional code reader according to claim 21, wherein when the labels overlap with each other at a ratio of the area, the logic state of the data cell area is determined to be on. 23. A two-dimensional code reading method for reading a two-dimensional code from a multi-valued image including a two-dimensional code, wherein the multi-valued image is binarized and the two-dimensional code included in the multi-valued image is included. A binarizing / quiet zone generating step of generating a quiet zone around the code; a timing cell detecting step of detecting a timing cell in the two-dimensional code surrounded by the quiet zone; and a timing cell detecting step. A predetermined area centered on the intersection of a straight line passing through the center or midpoint of each timing cell and the other straight line is defined as a data cell area used to identify the logical state of each data cell in the two-dimensional code. By setting a data cell area setting step as: and performing a labeling process on the area of the two-dimensional code. And a data cell logical state identifying step of identifying a logical state of each data cell area based on an overlap between the label and the data cell area set in the data cell area setting step. Dimension code reading method. 24. A square area including the entire area surrounded by the quiet zone in the binary image obtained in the binary / quiet zone generating step is defined as a two-dimensional code general existence area. Determining a rotation angle of a two-dimensional code in the two-dimensional code approximate existence area with respect to each of two crossing sides of the two-dimensional code approximate existence area; Based on the rotation angle obtained in the two-dimensional code rotation angle determination step and the passing point of the guide cell external tangent circumscribing the guide cells arranged in a line in the two-dimensional code, A guide cell outer tangent detection step, a rotation angle obtained in the two-dimensional code rotation angle determination step, and A timing cell external tangent detecting step of detecting the timing cell external tangent based on a passing point of a timing cell external tangent circumscribing the timing cells arranged in a line in the two-dimensional code; An area surrounded by the guide cell outer tangent detected in the line detecting step and the timing cell outer tangent detected in the timing cell outer tangent detecting step is a two-dimensional code in the timing cell detecting step. 24. The two-dimensional code reading method according to claim 23. 25. When a part of a timing cell in a two-dimensional code is lost, a position where the timing cell is lost is calculated, and a lost timing cell complementing step of complementing the timing cell with the calculated position is performed. 25. The two-dimensional code reading method according to claim 24, further comprising: 26. The binarizing / quiet zone generating step includes: a multi-valued image dividing step of dividing a multi-valued image into a plurality of small regions; and a small region obtained in the multi-valued image dividing step.
A threshold value calculation region setting step of setting a threshold value calculation region including a binarized small region targeted for binarization and having a larger area than the binarized small region; Based on the luminance value in the value calculation area,
A threshold value calculating step for calculating a threshold value for the binarized small area; and comparing the threshold value with a luminance value of each pixel in the binarized small area to determine the binarized small area. Small area 2 for binarizing the area
3. A quantification step.
The method for reading a two-dimensional code according to 3, 24 or 25. 27. When a guide cell is arranged at an inner edge of the two-dimensional code, the step of determining a general area of the two-dimensional code includes the step of determining a binarized image obtained in the binarization / quiet zone generation step. Performing a pattern matching process using a mask to obtain a pattern matching value for each coordinate of the binarized image, wherein the pattern matching value obtained in the pattern matching process step is a maximum. And a guide cell vertex detecting step of detecting a point as a guide cell vertex. A quadrangular area including the guide cell vertex and including the entire area surrounded by the quiet zone is defined as the two-dimensional code general existence area. The two-dimensional code reading method according to claim 24, 25 or 26, wherein: 28. The two-dimensional code reading method according to claim 27, wherein when the guide cells are arranged in a substantially L-shape, a substantially L-shaped mask is used as the mask. 29. The two-dimensional code rotation angle determining step comprises: rotating a projection axis by a predetermined angle at a predetermined range angle with respect to each of two mutually intersecting sides of the two-dimensional code roughly existing area; A projection processing step of obtaining a projection intensity on the projection axis for each rotation angle by performing a projection process on the dimensional code approximate existence area; and a maximum projection for obtaining a maximum value of the projection intensity for each rotation angle of the projection axis. An intensity detection step, and a rotation angle detection step of obtaining a rotation angle at which the highest maximum value is obtained among the maximum values obtained at the maximum projection intensity detection step. 25. The rotation angle of the two-dimensional code with respect to each side of the two-dimensional code approximate existence area, wherein the obtained rotation angle is set as the rotation angle of the two-dimensional code.
The two-dimensional code reading method described in 5, 26, 27 or 28. 30. The guide cell outer tangent detection step includes: setting the projection axis for each of two mutually intersecting sides of the two-dimensional code approximate existence region in an object obtained in the two-dimensional code rotation angle determination step. By performing projection processing by inclining by the rotation angle of the two-dimensional code corresponding to the side that has been set, a projection processing step of obtaining a projection intensity on the projection axis, and a projection intensity obtained in the projection processing step A guide cell outer tangent passing point detecting step for obtaining a passing point of the guide cell outer tangent based on the basis of: a guide cell outer tangent generating step of generating a guide cell outer tangent from the passing point and the rotation angle of the two-dimensional code. 24. The method according to claim 24, 25, 26, or 2,
The two-dimensional code reading method according to 7, 28 or 29. 31. The timing cell outer tangent detection step includes: determining a projection axis for each of two mutually intersecting sides of the two-dimensional code approximate existence area in the two-dimensional code rotation angle determining step; By performing projection processing by inclining by the rotation angle of the two-dimensional code corresponding to the side considered, a projection processing step of obtaining a projection intensity on the projection axis, and a projection intensity obtained in the projection processing step A timing cell outer tangent passing point detecting step of obtaining a timing cell outer tangent passing point based on the timing cell outer tangent passing point detecting step; 24. The method according to claim 24, wherein
The method for reading a two-dimensional code according to 5, 26, 27, 28, 29 or 30. 32. When a plurality of timing cell external tangents are generated in the timing cell external tangent generation step, each of the generated plurality of timing cell external tangents is used as a temporary timing cell external tangent, and The length of each side of the quadrangle formed by the outer tangent line and the guide cell outer tangent line that intersects each other detected in the guide cell outer tangent detection step is the length of one side of the two-dimensional code as the design value. The temporary timing cell external tangents that intersect each other satisfying one or both of the first condition of closeness and the second condition that the number of timing cells existing inside the temporary timing cell external tangent is the largest. Claims: A method comprising the step of selecting a normal timing cell external tangent to select as a normal timing cell external tangent. 31
The two-dimensional code reading method described in the above. 33. The timing cell detecting step includes: a timing cell outer tangent detected in the timing cell outer tangent detecting step; and a guide facing the one timing cell outer tangent from the one timing cell outer tangent. The binarized image surrounded by a straight line translated by one cell toward the cell outer tangent, the other timing cell outer tangent, and the guide cell outer tangent facing the other timing cell outer tangent is the one timing. A projection processing step of projecting onto the cell tangent side to obtain the projection intensity on the one timing cell tangent, and a threshold value calculation calculating a threshold value based on the projection intensity obtained in the projection processing step A temporary timing cell detecting step of binarizing the projection intensity using the threshold value and obtaining a temporary timing cell. A distance between the other timing cell outer tangent and the guide cell outer tangent facing the other timing cell outer tangent;
A calculation timing cell deriving step of obtaining a calculation timing cell from the number of guide cells and the number of timing cells included in one side of the two-dimensional code as a design value; and comparing the provisional timing cell and the calculation timing cell to calculate a calculation timing. 33. The two-dimensional code reading method according to claim 31, further comprising a normal timing cell setting step of setting a temporary timing cell corresponding to the cell one-to-one as a normal timing cell. 34. The erasure timing cell complementing step, wherein a erasure timing cell temporary position determination step of setting a position of a calculated timing cell corresponding to the lost timing cell as a tentative position of the lost timing cell; 2 near the same position as the temporary position
A timing cell determining step of determining two regular timing cells; and a timing cell complementing step of complementing a timing cell at a position indicated by an internal dividing point or an external dividing point of the two adjacent normal timing cells. The two-dimensional code reading method according to claim 33, characterized in that: 35. The data cell area setting step, wherein: an intersection setting step of obtaining an intersection of a straight line passing through a center or a midpoint of each timing cell detected by the timing cell detecting step and another straight line; 27. A data cell region generating step of generating, as the data cell region, each region having a predetermined area centered on each intersection obtained by the setting step. 2
The two-dimensional code reading method according to 7, 28, 29, 30, 31, 32, 33, or 34. 36. The data cell logical state identification step includes: a labeling processing step of generating a label by performing a labeling processing on an area surrounded by the guide cell external tangent and the timing cell external tangent; and the labeling processing. Of all the labels generated in the step, the area of one label is an averaged area calculation step of calculating the averaged area of the labels within a predetermined range, of the labels generated in the labeling processing step,
A label extraction step of extracting a label in a predetermined range centered on the averaging area, and a label extracted by the label extraction step;
25. A data cell logical state determining step of determining a logical state of each data cell area based on an overlap with the data cell area set in the data cell area setting step. , 25, 2
6, 27, 28, 29, 30, 31, 32, 33, 34
35. The two-dimensional code reading method according to 35. 37. The data cell logic state determination step, wherein when the label generated in the labeling processing step extends over a plurality of data cell areas by itself, the logic of the data cell area where the label overlaps the most widely. A first data cell logical state determining step of determining the state to be on, and a label other than the single label for a data cell region whose logical state has been determined to be off in the first data cell logical state determining step. When the ratio of the area occupied by is greater than or equal to a predetermined value, and the label overlaps the center of the data cell region by the ratio of the predetermined area or more, the second data for determining that the logic state of the data cell region is ON 4. A cell logic state determining step.
6. The two-dimensional code reading method according to 6. 38. The method according to claim 23,24,25,26,2.
7, 28, 29, 30, 31, 32, 33, 34, 3
A program for causing a computer to execute the two-dimensional code reading method described in 5, 36, or 37.