JP2011103023A - Method and device for reading color arrangement code - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method for reading data when the arrangement of a constituent cell group of a 1D color bit code is partially disordered or even if it is difficult to provide a cut-out Qz around the constituent cell group. <P>SOLUTION: An image including the 1D color bit code 10 of a reading target is imaged to obtain a capture image 12. Next, a long visual field frame B is set in the capture image 12, and a narrow image in the visual field frame B is taken out. Next, the takeout narrow image is divided into a plurality of color areas by a value of each pixel in the narrow image, and color arrangement information described with an order from one end of the narrow image to the other end and a color of each color area is generated. Next, when the generated color arrangement information satisfies an establishment condition prescribing the establishment of the 1D color bit code, the color arrangement information is regarded as the arrangement of the colors of the 1D color bit code of the reading target, decoding is performed based on the color arrangement information, and data are outputted. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to an optical automatic recognition code. In particular, the present invention relates to a reading method in which the reading rate of an optical automatic recognition code called a 1D color bit code (Japanese Patent Application No. 2006-196705) already proposed by the present inventor is further improved. The present invention also relates to a method for optically reading and decoding an automatic recognition code.

  The present inventor has already filed a plurality of applications relating to optical automatic recognition codes that represent data by color arrangement (transition).

  The present inventor proposed a 1D color bit code in Japanese Patent Application No. 2006-196705. An area to which colors constituting the 1D color bit code are attached is called a “cell”. That is, a 1D color bit code is configured by arranging cells with predetermined colors. Each 1D color bit code itself (a geometrical figure formed by a plurality of cells) is called a “code symbol” or “color array”.

  When a color array is captured in order to read data, the color array itself and a portion that is not the color array appear in one image. In order to read the color array, it is necessary to cut out the color array portion from the image.

  However, depending on the situation, it may be difficult to cut out the color array to be read.

  Therefore, the present inventor has proposed a reading method using a quiet zone provided around the color array in the previous application (Japanese Patent Application No. 2007-130504). The quiet zone is an area of colors other than the colors constituting the color array. The quiet zone may be described as “Qz”. In this reading method, image data including a color array and a quiet zone is divided into a plurality of color regions based on the definition. Then, the quiet zone is excluded from the divided area group, and it is determined whether or not the remaining area group is a cell constituting a color array using a boundary condition, a cell number condition, a termination condition, and the like. According to this method, even if the captured image includes a portion that is not a color array, only the color array itself can be cut out.

Method for cutting out a color arrangement without providing a quiet zone In addition, in the previous application (Japanese Patent Application No. 2007-258430), the present inventor has a color arrangement that is linear and the surrounding quiet zone is small or exists. We proposed a method for extracting and decoding the presence of a color array in the case of not doing so. In this cutting-out method, a plurality of sheet-like objects to be color-marked on end faces are captured while being shifted little by little. In addition to the color arrangement, the captured image includes a side surface and a background of the sheet-like object to be printed, and one of the technical features is that the side surface and the background form a quiet zone. According to this method, it is possible to read the color arrangement by devising how to arrange and cut out the objects to be printed even in a situation where a wide Qz is difficult or does not exist like the end face of the sheet-like object.

Known Patent Document The following Patent Document 1 discloses a recognition assist device that monitors the presence of an obstacle from a traveling vehicle. In particular, this recognition assisting device sets a window having a narrower area than the image in the image acquired by the image sensor, and controls the position of the window so that the object to be monitored is located in the window. Is described.

  Patent Document 2 below discloses a method for detecting and positioning a target pattern by image recognition within a camera field of view where the target pattern and the non-target pattern are imaged. In particular, in Patent Document 2, a recognition window is set around the target pattern m based on the position of the non-target pattern n, and the target pattern m is recognized based on the centroid data and area data of the target pattern m. It is described that the position of the recognition window is corrected by determining whether it is within the range.

JP 2005-244399 A Japanese Patent Laid-Open No. 9-5022

  In the reading method described in Japanese Patent Application No. 2007-130504 filed earlier by the present inventor, the 1D color bit code needs to be surrounded by a color area (quiet zone) other than the color used by the 1D color bit code. was there. This is because it is easy to cut out the 1D color bit code. Further, in the reading method described in Japanese Patent Application No. 2007-130504, each color region constituting the 1D color bit code is cut out based on the “boundary condition” of the 1D color bit code. The “boundary condition” is, for example, a condition that “one cell contacts one or two other cells”.

  However, if the 1D color bit code is disturbed for some reason and only a part of the 1D color bit code does not meet the “boundary condition”, the entire 1D color bit code may not be read. This processing is preferable from the viewpoint of ensuring accuracy in reading the 1D color bit code.

  However, depending on the situation, the reading method described in Japanese Patent Application No. 2007-130504 may not be preferable.

  FIG. 1 shows a captured image 12 obtained by capturing the 1D color bit code 10. In this figure, eleven color regions are shown. Of these eleven color regions, three color regions 14b, 14d, and 14h are in contact with three or more other color regions. Therefore, since these three color regions do not meet the boundary condition described above, the conventional method often cannot read the 1D color bit code 10. As described above, since many 1D color bit codes cannot be read depending on the situation, the reading method described in the above Japanese Patent Application No. 2007-130504 is not convenient depending on the situation of use, and may not be preferable in reading operation. It was.

  In addition, when the 1D color bit code is marked in an extremely narrow place (for example, an extremely narrow place such as an end face of a plate-like body), it is not easy to provide Qz or cannot provide Qz. There is. As a result, it may be difficult to cut out the 1D color bit code. In addition, when various colors of the background are captured in the image capturing the end face, it may be difficult to determine the boundary of the 1D color bit code.

  The above-mentioned Japanese Patent Application No. 2007-258430 describes a method for reading a 1D color bit code marked on the end face of a plate-like body. However, this reading method shows that Qz exists even if it is very narrow. This is a presumed reading method. Specifically, for example, when a plurality of plate-like bodies marked with a 1D color bit code are arranged in close contact with each other, the side and background of the plate-like body are set as Qz on the captured image of the 1D color bit code. By handling, “there is Qz even if very narrow” is realized. Therefore, the reading method described in Japanese Patent Application No. 2007-258430 may not be able to handle the case where Qz does not exist.

Objects of the invention A first object of the present invention is to provide a method for reading a color array which is marked on a suitable straight section and forms a code. A second object is to provide a method of reading a color arrangement in which no quiet zone exists. A third object is to provide a method for reading a color arrangement even when a part of a cell group constituting the color arrangement is disturbed and marked.

  Moreover, it is providing the technique relevant to the said 1st-3rd objective.

(1) In order to solve the above-described problem, the present invention provides a method for reading a color array code that represents data by color transition, a capture step for capturing an image including the color array code, and a preset in the image A narrow image extracting step for extracting a narrow image inside the elongated field frame, a dividing step for dividing the narrow image into a plurality of color regions according to the value of each pixel in the narrow image, and for each color region, A color array information generating step for generating color array information describing the colors of the color area and the order from one end to the other end of the narrow image, and the color of the color area in the color array information, and Assuming that the order is the order and color of the cell group constituting the color arrangement code, the color arrangement code based on the color and order A decoding step of outputting the decoded data; a color disposition code reading method, which comprises a.

(2) Further, according to the present invention, in a method for reading a color arrangement code that represents data by color transition, a capture step for capturing an image including the color arrangement code, an elongated field frame preset in the image, and A narrow image extracting step for extracting an internal narrow image, a dividing step for dividing the narrow image into a plurality of color regions according to the value of each pixel in the narrow image, a color of the color region for each color region, and A color arrangement information generating step for generating color arrangement information describing the order from one end to the other end of the narrow image, and for cutting out indicating the start of the color arrangement code in the color arrangement information Color information indicating the color region to which the color of the redundant cell is attached, and the color of the redundant cell for extraction indicating the end of the color array code The color information representing the color area to which is attached, and the color and order of the color area in the color arrangement information between them are considered to be the order and color of the cell group constituting the color arrangement code, And a decoding step of decoding the color arrangement code based on the color arrangement information and outputting the data.

(3) Further, in the color array code reading method according to the above (1) or (2), the present invention relates to the field frame in which one or a plurality of the narrow image extracting steps are set in the image. A first visual field frame moving step for moving the position of the image in accordance with a user's instruction within the image, wherein the narrow image extracting step is configured to move the narrow image by the visual field frame moved in the visual field frame moving step. It is a color array code reading method characterized in that it is extracted.

(4) Further, in the color array code reading method according to the above (1) or (2), the narrow image extraction step may include the preset field frame for each captured image. A second visual field frame moving step for moving to a different position, and a narrow image inside the visual field frame set at the different position is constituted by colors constituting the color arrangement code, The color arrangement code reading method is characterized in that a narrow image inside the field frame is taken out.

(5) Further, in the color arrangement code reading method according to (4), the present invention provides the color arrangement code, wherein the field frame moves to a different position in a parallel direction for each image. Is the reading method.

(6) In the color arrangement code reading method according to (4), the field frame is moved to a different position in the direction of rotation about the center point of the field frame for each image. The color array code reading method is characterized in that:

(7) In the color arrangement code reading method according to (4), the field frame is rotated in a parallel direction for each image and with the center point of the field frame as an axis. The color arrangement code reading method is characterized in that the color arrangement code moves.

(8) In the color array code reading method according to (1) or (2), in the present invention, the narrow image extracting step includes a step of scanning the preset field frame on the image. 3 field frame moving step, and when the narrow image inside the field frame to be scanned in the field frame moving step is composed of colors constituting the color arrangement code, A color arrangement code reading method characterized in that a narrow image is taken out.

(9) In the color array code reading method according to (8), the field frame scans the image in parallel in the vertical direction or the horizontal direction on the image. How to read the code.

(10) Further, the present invention provides the color array code reading method according to (8) above,
The field frame is a method of reading a color arrangement code, wherein the field frame is scanned in a direction rotating around a center point of the field frame.

(11) In the color array code reading method according to (8), the field frame is rotated in a parallel direction for each image and with the center point of the field frame as an axis. The color array code reading method is characterized in that scanning is performed.

(12) In the color array code reading method according to any one of (1) to (11), the field frame may have a shape having a predetermined width and length, and may be in a longitudinal direction. One of the above is the one end, the other is the other end, and the length is longer than the width.

(13) In the color array code reading method according to (12), any one side of the field frame is a curve, and the field frame is arranged in a curved shape. A color array code reading method characterized by the above.

(14) In the color arrangement code reading method according to (12) or (13), the color arrangement information generation step may further include the narrow image of each color region in the narrow image. A length addition step for generating the color arrangement information described for each color area, and if the length is smaller than a predetermined reference value, the color area A data deletion step of deleting color information representing the color arrangement information from the color arrangement information, and a color arrangement information correction step of filling other color information adjacent before and after the deleted color information. It is a reading method.

(15) Further, in the color array code reading method according to any one of (1) to (14), the present invention is configured such that the constituent colors constituting the color array code are mixed in the color array information. When there is color information of the color area and the color of the area adjacent to the mixed color area is the component color, the mixed color is the color of any color area adjacent to the color area. A method for reading a color arrangement code, comprising: a color deeming step for deeming that

(16) Moreover, in the color array code reading method according to any one of (1) to (15), the present invention provides a method for extracting the narrow image and the color array information generating step. A uniformizing step for replacing the value of each pixel in the narrow image with an average value of a pixel row including the pixel and extending in the width direction of the narrow image, and the color array information generating step includes the uniform The color arrangement code reading method, wherein the color arrangement information is generated based on the narrow image after performing the conversion process.

(17) Further, the present invention provides a method for reading a color array code representing data by color transition, a capture step for capturing an image including the color array code, and the image based on a value of each pixel in the image A step of dividing the image into color regions, a first start cell deeming step in which a color region including a pixel at a predetermined position in the image is regarded as a start cell of the color arrangement code, and the start cell A second cell deeming step that regards a color region of a predetermined color as a second cell of the color arrangement code among other color regions adjacent to, a centroid point of the (k + 1) th cell, and the first A straight line passing through each centroid point of each of the other color regions adjacent to the (k + 1) cell is drawn for each centroid point, and the (k) +1) The straight line having the smallest angle with respect to the straight line passing through the centroid point of the cell and the centroid point of the kth cell is selected, and the (k + 1) th centroid point among the two centroid points through which the selected straight line passes. The cell selection step of regarding a color region to which a barycentric point other than the barycentric point of a cell belongs as the (k + 2) th cell of the color arrangement code, and the cell group from the start cell to the nth cell For each color area, a color array information generating step for generating color array information describing the color of each color area and the order from the start cell to the nth cell, and in the color array information Assuming that the color and order of the color area are the colors and order of the cells constituting the color array code, the color array code is decoded based on the color and order and the data is output. And code step,
A method of reading a color arrangement code. (K is a natural number of 1 or more, and n is a natural number greater than k.)
(18) Further, according to the present invention, in a method for reading a color array code representing data by color transition, a capture step for capturing an image including the color array code, and the image based on a value of each pixel in the image A division step for dividing the color region into color regions, and selecting a color region of a predetermined color from a group of color regions in contact with a predetermined image edge in the image, and selecting the selected color region in the color array A second start cell deeming step which is regarded as a start cell of the code, and a color region of a predetermined color among the other color regions adjacent to the start cell is the second cell of the color arrangement code A second cell considering step, a centroid point of the (k + 1) th cell, and centroid points of other color regions adjacent to the (k + 1) th cell; A straight line with the smallest angle formed with respect to the straight line passing through the centroid point of the (k + 1) -th cell and the centroid point of the k-th cell is selected from the centroid points. A cell selection step in which a color region to which a centroid other than the centroid of the (k + 1) th cell among the two centroids through which the selected straight line passes is regarded as the (k + 2) th cell of the color arrangement code. And for each of the color areas regarded as a cell group from the start cell to the nth cell, the color of each color area and the order from the start cell to the nth cell, The color arrangement information generation step for generating the described color arrangement information, and the color and order of the color areas in the color arrangement information are regarded as the color and order of the cell group constituting the color arrangement code, and the color as well as A color disposition code reading method, which comprises a, a decoding step of outputting the decoded the data the color sequence code based on the turn. (K is a natural number of 1 or more, and n is a natural number greater than k.)
(19) Further, in the color array code reading method according to the above (17), the present invention provides a plurality of the predetermined positions in the first start cell considering step,
In the first start cell deeming step, when a color at any one of the plurality of predetermined positions is a predetermined color, the color area is a start cell of the color arrangement code. This is a color array code reading method characterized in that

(20) Further, in the color array code reading method according to the above (18), the present invention provides an article to which the predetermined color of the second start cell deeming step is attached with the color array code. Or a background color obtained by coloring the front surface of the ground color.

(21) Further, according to the present invention, in a color arrangement code reading device that represents data by color transition, a capture unit that captures an image including the color arrangement code, and an elongated field frame preset in the image A narrow image extracting means for extracting an internal narrow image, a dividing means for dividing the narrow image into a plurality of color regions according to the value of each pixel in the narrow image, and for each color region, the color of the color region Color arrangement information generating means for generating color arrangement information describing the order from one end to the other end of the narrow image, and the color and order of the color areas in the color arrangement information are the color arrangement It is assumed that it is the color and order of the cells constituting the code, and the color array code is decoded based on the color and order and the data is output. And over de unit, a reader of color disposition code, which comprises a.

(22) Further, according to the present invention, in a color arrangement code reading device that represents data by color transition, capture means for capturing an image including the color arrangement code;
A narrow image extracting means for extracting a narrow image inside an elongated field frame set in advance in the image; a dividing means for dividing the narrow image into a plurality of color regions according to values of each pixel in the narrow image; For each color area, color arrangement information generating means for generating color arrangement information describing the order from one end of the narrow image to the other end and the color of the color area, and the color arrangement information Among them, the color information indicating the color area to which the color of the redundant cell for extraction indicating the start of the color arrangement code is attached, and the color area to which the color of the redundant cell for extraction indicating the end of the color arrangement code is attached. The color information indicating the color arrangement and the color area in the color arrangement information between them are considered to be the color and the order of the cell group constituting the color arrangement code, and the color And decoding means for decoding the color sequence encoding outputs the data based on the column information, a reading device color sequence code, which comprises a.

(23) Further, the present invention provides the color arrangement code reader according to (21) or (22) above, wherein the position of the field frame set in the image is set in the image. The color arrangement code reading device includes: a field frame moving unit that moves in accordance with a user instruction, wherein the narrow image extracting unit extracts the narrow image by the moved field frame.

  According to the reading method of the present invention, even if a part of the color arrangement does not match the boundary condition, a portion that matches the boundary condition can be cut out and decoded using the field frame.

  Also, according to the reading method of the present invention, even if a part of the color arrangement does not match the boundary condition, the tracking is performed using the linearity of the color arrangement, so that it can be decoded.

It is a principle diagram of a 1D color bit code reading method. It is a figure which shows 1D color bit code reading apparatus. It is a figure explaining the processing method of noise. It is a figure explaining the processing method of the area | region where a color is mixed. It is a figure explaining the averaging process of a color. It is a figure explaining the short side direction color equalization process. It is a figure explaining the marking method of the cell in case white is made into one of the constituent colors of 1D color bit code. It is a figure explaining the redundant cell for cut-out. It is a figure explaining the reading method using the redundant cell for cutting. It is a figure which shows the variation of the number, arrangement | positioning, and shape of a visual field frame. It is a figure explaining the method of tracking the structure cell of 1D color bit code. It is an enlarged view explaining the method of tracking the structure cell of 1D color bit code. It is a figure explaining the method of tracking 1D color bit code containing an endpoint cell. It is a figure explaining the search method of 1D color bit code. It is a figure explaining the reading method of a serial 1D color bit code. It is a figure explaining the reading method of a serial 1D color bit code. It is a figure which shows the arrangement | sequence of a color and the decoding result. It is a figure explaining the effect of the reading method which concerns on this Embodiment.

(1) Features of the reading method according to this embodiment
Feature 1
The 1D color bit code is cut out from an image including the 1D color bit code by the method described in Japanese Patent Application No. 2007-130504 filed earlier by the present inventor. One of the characteristic matters in the present embodiment is that a part of the image of the 1D color bit code is taken out using the field frame before performing this clipping.

  If the 1D color bit code satisfies predetermined boundary conditions and fulfillment conditions, the decoding means can read the 1D color bit code. Here, the boundary condition is, for example, a condition that “each cell is in contact with one or two other cells”. Further, the establishment conditions are conditions such as the number of R, the number of G, the number of B, the color at the end of the start, the color at the end of the end, and the error correction code, as in the conventional case.

  However, if only a part of the cells of the 1D color bit code are in contact with, for example, three cells, the cells do not satisfy the boundary condition, and the entire 1D color bit code may not be read.

  On the other hand, if the reading range is narrowed and a portion that does not satisfy the boundary condition is excluded from the reading target, the 1D color bit code may be read. When only a part of the 1D color bit code does not satisfy the boundary condition, it is often preferable to read the remaining part except for the part.

  Therefore, in the present embodiment, an application running in the decoding means performs a process of masking on the image data of the 1D color bit code by software, and the image data is not masked with the masked portion. Divide into parts. In the present embodiment, a portion that is not masked is particularly referred to as a “narrow field of view”. One of the features in the present embodiment is that only the image with the narrow field of view is read and decoded.

Feature 2
In the present embodiment, the position of the narrow visual field is presented to the user by the decoding means displaying it on the image with a visual field frame. The shape of the field frame needs to be a shape that excludes a portion that does not satisfy the boundary condition of the 1D color bit code.

  For example, when reading a 1D color bit code marked on the end face of a plate-like body, it is preferable to use a field frame in which the short side is set narrower than the width of the 1D color bit code. This is to prevent a portion that does not satisfy the boundary condition of the 1D color bit code from entering the field frame. In this case, the shape of the field frame is a narrow linear shape. The narrow linear field frame corresponds to a preferable example of the shape of the “field frame” described in the claims of the present application.

  In this way, if only the portion located inside the narrow linear field frame is read out of the linear 1D color bit code, the portion that prevents the seriality of the 1D color bit code may be excluded from the reading target. It is highly possible. As a result, the 1D color bit code can be satisfactorily read without being affected by the portion that obstructs the seriality of the 1D color bit code, and the probability can be improved.

Feature 3
Another feature of the present embodiment is that the cut-out operation and the decoding operation of the 1D color bit code are performed by paying attention only to the color change in the longitudinal direction (linear direction) of the field frame.

  In the present embodiment, “cutout” refers to detecting and extracting a cell group constituting a 1D color bit code from a color arrangement (color area group) in a linear field frame. In the present embodiment, “decoding” refers to recognizing the arrangement order of the colors assigned to the cell group constituting the extracted 1D color bit code, and outputting data from the recognized arrangement order. .

(2) Conventional Reading Method In the conventional reading method, a 1D color bit code is read by tracking a cell that satisfies the boundary condition “each cell touches one or two other cells” in a captured image. It was. However, if a small part of the 1D color bit code does not satisfy the boundary condition due to dirt, printing mistakes, noise, or the like, the entire 1D color bit code may not be read.

  FIG. 1 shows a principle diagram of a 1D color bit code reading method according to the present embodiment. In this figure, 11 color regions 14 groups are shown. Of these 11 color area 14 groups, 9 color areas (color areas 14a, 14b, 14c, 14d, 14f, 14g, 14h, 14i, 14j) are arranged in series from left to right in FIG. It is arranged. These nine color regions 14a, 14b, 14c, 14d, 14f, 14g, 14h, 14i, and 14j are assigned R, G, B, R, B, R, G, R, and B, respectively. These nine color area 14 groups are constituent cells of the 1D color bit code 10, and a change in color between the cells represents data.

  The remaining two color regions 14e and 14k are {G}, {color (Q)} that is not any of R, G, and B, respectively. These two color areas 14 e and 14 k are not constituent cells of the 1D color bit code 10. The surface of the object to be marked with the 1D color bit code 10 is reflected around the 11 color regions 14 group. The portion in which the surface of the object is shown plays a role of distinguishing the 1D color bit code 10 from the surrounding area (not shown in FIG. 1) that is not the 1D color bit code.

  By the way, among the ten color regions to which the constituent colors of the 1D color bit code 10 are attached, the color regions 14b, 14d, 14f, 14g, 14h, and 14i are in contact with three or more other color regions. If some cells in the constituent cell group of the 1D color bit code 10 are in contact with three or more other cells, two or more directions in which the cells are connected exist, and the seriality of the cells is prevented.

  In the conventional 1D color bit code segmentation algorithm, the boundary condition “each cell touches one or two other cells” needs to be satisfied, but these color regions 14b, 14d, 14f, 14g, 14h, Since 14i contacts three or more other cells, this boundary condition is not satisfied. For this reason, the 1D color bit code 10 is not a target for extraction by the conventional reading method using the extraction algorithm.

  In many cases, it is not appropriate that the entire 1D color bit code cannot be read because a small part of the 1D color bit code does not satisfy the boundary condition due to dirt, printing mistakes, noise, or the like. Even if a very small part of the 1D color bit code is contaminated, it is very convenient if the 1D color bit code can be read by removing the dirt from the reading target.

  In this embodiment, the reading method is highly likely to be able to read the 1D color bit code even if only a part of the 1D color bit code is contaminated.

(3) Principle of Reading Method According to this Embodiment (3-1) Reading Device The reading device in this embodiment will be described with reference to FIG. FIG. 2 shows a decoding means 16 for decoding the 1D color bit code. A capture means 18 is connected to the decoding means 16. The capture means 18 is fixed on a tripod in FIG. 2, and the group of objects to be stamped 20 is successively moved in front of the capture means 18. The capture unit 18 captures each 1D color bit code 22 marked on the group of objects to be marked 20 and transmits the image data to the decoding unit 16. The capture means 18 does not need to be fixed. For example, it is preferable that the user captures each 1D color bit code 22 while holding the capture means 18.

(3-2) Reading Method The decoding unit 16 activates a predetermined application for decoding the 1D color bit code 22. The image data 24 received from the capture means 18 is shown on the screen of the decoding means 16. The decoding means 16 performs a process of masking on the image data 24 by software, and divides the image data 24 into a masked part and an unmasked part. In the present embodiment, a portion that is not masked is called a narrow field of view. The decoding means 16 displays the position of the narrow visual field in the visual field frame 26 on the screen data according to the user's instruction. In the present embodiment, the image inside the field frame is referred to as a narrow image.

  FIG. 1 described above shows a captured image 12 obtained by capturing a 1D color bit code 10 in which cells are arranged in series. First, as an example of the simplest arrangement, a method of reading the serial 1D color bit code 10 with the field frame B will be described.

  When the decoding unit extracts a part of the captured image 12 in the field frame B, the narrow images arranged in the order of “color regions 14a → 14b → 14c → 14d → 14f → 14g → 14h → 14i → 14j” from left to right. Get. Next, the decoding means divides the narrow image into a plurality of color regions based on the value of each pixel in the narrow image. From the left end to the right end of the narrow image, {R color region}, {G color region}, {B color region}, {R color region}, {B color region}, {R color region} It is recognized that {color region}, {G color region}, {R color region}, and {B color region} exist in this order.

  The decoding means determines whether or not the boundary condition “each cell is in contact with one or two other cells” is satisfied for the group of color regions 14 in the narrow image. In FIG. 1, since the short side of the field frame B is set to be shorter than the length (in the vertical direction in the figure) of the color region 14 group, the color region 14 group in the narrow image in the field frame B is a boundary condition. Meet.

  Next, the decoding means assumes that the narrow image is the 1D color bit code 10, and generates color information for each color region. The color information is information in which a number (ID) assigned to the color area, information representing the color, the number of the color area adjacent in the forward direction, and the number of the color area adjacent in the backward direction are described. It is. Since nine color regions exist in the narrow image taken out in the field frame B, nine pieces of color information are generated. When the boundary condition is not satisfied, the decoding unit does not create color information, and the decoding unit performs the same process on the next image data received from the capture unit.

  In FIG. 1, the start of the 1D color bit code is the leftmost color area 14a, and the end is the rightmost color area 14j. Therefore, in FIG. 1, the color area adjacent to the left of the color area is the color area adjacent in the forward direction, and the color area adjacent to the right is the color area adjacent in the backward direction. Therefore, the decoding means, based on the narrow image taken out in the field frame B, “1, R, (none), 2”, “2, G, 1, 3”, “3, B, 2, 4”, “4, R, 3, 5”, “5, B, 4, 6”, “6, R, 5, 7”, “7, G, 6, 8”, “8, R, 7, 9”, Color information “9, B, 8, (none)” is generated. These nine pieces of color information constitute one piece of color arrangement information.

  Thereafter, the decoding means regards the generated color arrangement information as representing the color arrangement of the 1D color bit code 10, decodes the 1D color bit code 10 based on the color arrangement information, and outputs data.

  According to the reading method of the present embodiment, even if a part of the 1D color bit code is contaminated and a part does not satisfy the boundary condition of the 1D color bit code, the decoding means determines the remaining boundary condition. Since only the satisfying part is read, the possibility of reading using the conventional decoding algorithm is dramatically increased.

  In the present embodiment, the decoding means corresponds to a preferable example of the narrow image extracting means, the dividing means, and the color arrangement information generating means described in the claims.

(3-3) Field frame
As described above the length of the short side of the field frame, the reading method of this embodiment, as one of features to read only narrow image. Therefore, the length of the short side of the field frame is preferably a length that allows only a portion satisfying the boundary condition to enter the narrow field of view. In general, the shorter the length of the field frame in the short side direction (width direction), the less likely the portion of the field frame that does not satisfy the boundary condition of the 1D color bit code is to enter, and the decoding means uses 1D color. The probability that the bit code can be read increases. Therefore, in the present embodiment, the length of the short side of the field frame is preset to about 1 pixel or several pixels which is the shortest length on the capture screen.

  The decoding means determines whether or not the narrow image inside the field frame satisfies the boundary condition, and presents the determination result to the user with an indicator. When the narrow image satisfies the boundary condition, the decoding means performs decoding based on the narrow image.

  When the narrow image does not satisfy the boundary condition, the user performs various operations so that the narrow image satisfies the boundary condition while viewing the presented determination result, the narrow image, and the captured image. Can do.

  When the short side of the field frame is set to a small number of pixels such as one pixel or several pixels, the S / N ratio of a narrow image is lowered. As a result, it is expected that the decoding means is strongly influenced by noise or the like existing inside the field frame during decoding and cannot output data accurately.

  Therefore, the decoding means performs appropriate processing on the noise in the narrow image using a median filter or the like according to the user's instruction, thereby preventing the influence of the noise from reaching the output data. As a result, even if the short side of the field frame is set to be as short as a few pixels, the possibility of correct decoding increases.

The length of the long side of the field frame The length of the long side of the field frame is selected by the user from two types set in advance in the decoding means.

  One type is a narrow field of view in which the length of the long side of the field frame is longer than the length of the 1D color bit code. The other type is a field frame in which the length of the long side of the field frame is shorter than the length of the 1D color bit code.

  FIG. 1 shows the former type of field frame A and the latter type of field frame B. The field frame A is an example of a sufficiently long one for the 1D color bit code 10. Both ends in the longitudinal direction of the field frame A are in the ground region. On the other hand, the field frame B is an example which is shorter than the 1D color bit code 10 and whose both ends in the longitudinal direction are within the end point cell region. The color region 14a is sufficiently longer than the region where the color region 14a and the field frame B overlap. The color region 14j is sufficiently longer than the region where the color region 14j at the other end point overlaps the field frame B. That is, in other words, the field frame A is an example in the case where the surface of the object to be marked (Qz) is present. The field frame B is an example when there is no surface (Qz) of the object to be marked.

  First, in principle, the field frame B is selected. This is because the color area groups in the narrow image cut out by the field frame B are all color areas of the constituent colors of the 1D color bit code. However, if the 1D color bit code is longer than the field frame B and the color area constituting the 1D color bit code cannot be contained inside the field frame B, the field frame A may be selected. The user can select either the field frame A or B as appropriate in consideration of the ease of “cutting out” or “decoding” of the 1D color bit code.

  When a part of the captured image 12 is extracted from the field frame B, narrow images arranged in the order of “color regions 14a → 14b → 14c → 14d → 14f → 14g → 14h → 14i → 14j” are obtained from the left to the right. This narrow image does not include a portion that disturbs the seriality of the cell group.

  Further, when a part of the captured image 12 is taken out from the field frame A, from the left to the right, “the surface of the object to be printed → the color region 14a → 14b → 14c → 14d → 14f → 14g → 14h → 14i → 14j → marked” Narrow images arranged in the order of “object surface” are obtained. This narrow image also does not include a portion that disturbs the seriality of the cell group.

  In principle, the length of the long side of the field frame is set to B in advance before the decoding means performs the reading operation of the 1D color bit code. You may make it choose the type of a visual field frame suitably on a screen.

  As described above, according to the reading method according to the present embodiment, the portion that disturbs the seriality of the cell group is excluded from the reading target, and only the remaining portion is read. The 1D color bit code can be read without receiving.

Number of Field Frames FIGS. 1 and 2 described above show an example in which one narrow image is cut out using one field frame. In the present embodiment, the position of the field frame needs to match the position of the 1D color bit code. However, the possibility that the 1D color bit code in the captured image coincides with the position of only one field frame set on the screen is not necessarily high.

  Therefore, in the present embodiment, the number of field frames is increased, for example, five field frames are set as shown in FIG. 10 (2), and five narrow images are cut out from one captured image. Is also preferable. If one 1D color bit code is detected simultaneously with a plurality of field frames having different positions, a plurality of different narrow images are cut out for one 1D color bit code. If there are a plurality of cut out narrow images, the number of candidates to be read increases. As long as any one of a plurality of narrow images satisfies the boundary condition of the 1D color bit code, the decoding unit can decode the 1D color bit code. To increase.

In the case where the capture screen itself is a field frame In the above-described example, the method has been described in which the decoding unit masks the image data and extracts a part of the image data with the field frame. On the other hand, instead of extracting a part of the image data, a reading device having a screen with a narrow field of view may be used. This also makes it possible to cut out the 1D color bit code without being affected by the portion that hinders the seriality of the cells.

(4) Noise Processing Method As described above, the reading method according to the present embodiment is characterized in that a part of a captured image is extracted using a field frame and decoding is performed based on the extracted narrow image. One. Since the narrow image is image data of an image smaller than the captured image, the amount of information obtained from the image data is small. Therefore, since the S / N ratio of the narrow image is smaller than that of the captured image, the reading method in the present embodiment may be susceptible to noise.

  Therefore, a method for processing noise mixed in a narrow image will be described.

(4-1) Processing Method Utilizing Length of Color Area FIG. 3 shows a narrow image 28 obtained by cutting out a part of a captured image of a 1D color bit code with a narrow field of view. As shown in FIG. 3, the narrow image 28 includes an R color area, a G color area, a Q color area, a B color area, an R color area, a B color area, a Q color area, R Eleven color regions, namely, the color region G, the color region G, the color region Q, and the color region G exist in this order.

  Based on the color areas from the left end to the right end of the narrow image 28, the decoding means “1, R, (none), 2”, “2, G, 1, 3”, “3, Q, 2, 4 ”,“ 4, B, 3, 5 ”,“ 5, R, 4, 6 ”,“ 6, B, 5, 7 ”,“ 7, Q, 6, 8 ”,“ 8, R, 7, Color information of “9”, “9, G, 8, 10”, “10, Q, 9, 11”, “11, G, 10, (none)” is generated. These 11 pieces of color information constitute one piece of color arrangement information.

  In FIG. 3, the decoding means further measures the length in the longitudinal direction of the narrow image in each color region from the left end to the right end of the image 28, and adds the length to the color information. This is because the color information generated from the noise is later removed using the length information. The color information after adding the length describes the number (ID) assigned to the color area, the color information, the number of the color area adjacent in the forward direction, the number of the color area adjacent in the backward direction, and the length. Information.

  The lengths of the color regions from the left end to the right end of the image 28 are 10, 7, 4, 10, 1, 5, 3, 4, 8, 2, 5 respectively. Therefore, each color information after adding each length is “1, R, (none), 2, 10”, “2, G, 1, 3, 7”, “3, Q, 2, 4, 4 ”,“ 4, B, 3, 5, 10 ”,“ 5, R, 4, 6, 1 ”,“ 6, B, 5, 7, 5 ”,“ 7, Q, 6, 8, 3 ”,“ 8, R, 7, 9, 4 ”,“ 9, G, 8, 10, 8 ”,“ 10, Q, 9, 11, 2, ”,“ 11, G, 10, (none), 5 ".

The noise type narrow image 28 includes two types of noise.

  The first type of noise is a color region to which colors other than the constituent colors of the 1D color bit code are added and whose length is shorter than the reference. In the narrow image 28, the maximum value 10 of the length of the color area is used as a reference value. A color region whose length is equal to or greater than a reference value is a region sandwiched between two 1D color bit codes.

  The second type of noise is a color region to which the colors of the constituent colors of the 1D color bit code are added and the length is very short. In the narrow image 28, a color region having a length of 1 is considered to be very short.

  The original color arrangement of the 1D color bit code in FIG. 3 is “R → G → B → R → G”, but as a result of mixing the two types of noise described above, a color different from the original color arrangement is obtained. Color arrangement information representing the arrangement of For this reason, even if the decoding means tries to decode the color arrangement information, it may not be decoded correctly.

  Therefore, in the present embodiment, the above two types of noise are removed by the following two steps.

(First Step) First, the decoding means performs processing of ignoring the color information representing the first type of noise (deleting the color information from the color arrangement information) and filling the color information before and after that. In the present embodiment, there are three first type noises (“3, Q, 2, 4, 4”, “7, Q, 6, 8, 3”, “10, Q, 9, 11, 2 "), the decoding means deletes these three pieces of color information from the color arrangement information.

  Next, the decoding means obtains the “number of color areas adjacent in the backward direction” in the color information adjacent in the forward direction of each deleted color information and the “number of color areas adjacent in the backward direction” in the color information adjacent in the backward direction of the deleted color information. The color area number (ID) "is corrected. For example, the color information adjacent to the front direction of the color information “3, Q, 2, 4, 4” (before correction) is “2, G, 1, 3, 7”. To 1, 4, 7 ".

  In addition, the decoding means displays the “color area number adjacent in the forward direction” in the color information adjacent in the backward direction of the deleted color information, and the “color area in the color information adjacent in the forward direction of the deleted color information. To "number (ID)". For example, the color information (before correction) adjacent to the backward direction of the color information “3, Q, 2, 4, 4” is “4, B, 3, 5, 11”. To 2, 5, 11 ".

  In the present embodiment, “1, R, (none), 2, 10”, “2, G, 1, 4, 7”, “4, B, 2, 5, 11 ”,“ 5, R, 4, 6, 1 ”,“ 6, B, 5, 8, 5 ”,“ 8, R, 6, 9, 4 ”,“ 9, G, The processing results of “8, 11, 8”, “11, G, 9, (none), 5” are obtained.

  This is the processing content of the first step.

(Second Step) Next, the decoding means determines that a color area having a length of 1 or less among the processing results of the first step is the second type noise and determines that it is the second type noise. Ignore the color information corresponding to the determined color area (delete the color information from the color array information). And the process which narrows between the color information adjacent to the front direction of the color information to ignore and the color information adjacent to the back direction is performed. In the present embodiment, since “5, R, 4, 6, 1” is the color information of the second type of noise, it is deleted. Also, the decoding means ignores “the number of the color area adjacent in the backward direction” in the color information adjacent in the forward direction of the color information to be ignored. "Number (ID)" is corrected, and "number of color area adjacent in the forward direction" in the color information adjacent in the backward direction of the color information to be ignored is changed to "number in the color information adjacent in the forward direction of the color information to be ignored" The color area number (ID) "is corrected.

  As a result, according to the processing of the second step, from the processing result of the first step, “1, R, (none), 2, 10”, “2, G, 1, 4, 7”, “4, B, 2, 6, 11 "," 6, B, 4, 8, 5 "," 8, R, 6, 9, 4 "," 9, G, 8, 11, 8 "," 11, G, 9, (None) 5 ”is obtained.

  This is the processing content of the second step.

  Thereafter, the decoding means determines whether or not the processing result of the second step satisfies the conditions for establishing the 1D color bit code. If the processing result of the second step satisfies the establishment condition, the decoding means determines that the processing result of the second step is correct, performs decoding, and outputs data. If the processing result does not satisfy the establishment condition, the decoding unit determines that the decoded 1D color bit code is another 1D color bit code that is not a reading target, and captures the 1D color bit code again. Correct and repeat decoding again.

(4-2) Method for processing a portion where two color regions overlap In general color printing, it is widely performed to slightly overlap color boundary portions. This is to prevent a gap from appearing at the color boundary even if the plate is shifted slightly. When two colors are printed with a combination of a dark color and a light color, the overlapped portion of the two colors often becomes a part of the dark color area, so that it is not particularly necessary to recognize the mixture of colors.

  On the other hand, when two colors are printed with a combination of light colors (colors with low luminance), a portion where the colors are superimposed may become another color region. For example, when two colors are printed with a combination of light yellow and light blue, a portion where the colors are superimposed becomes a green color region. As a result, instead of the original color arrangement “(light yellow area) − (light blue area)”, a different color arrangement “(light yellow area) − (green area) − (light blue area)” Region) ”, it is not appropriate.

  A schematic diagram of the boundary portion of the color region is shown in FIG. In FIG. 4A, the color region 30A (the portion indicated by the right oblique line) is the A color cell of the 1D color bit code constituent color, and the color region 30B (the portion indicated by the left oblique line) is the 1D color bit code constituent color. B cells. If a part of each color area 30A, 30B overlaps, two colors of A color and B color may be mixed in a certain area (area 32 in the figure).

  When the color of the area 32 is recognized by a normal method in terms of software, the color of the area 32 becomes another color C color different from the A color and the B color (when the A color and the B color are mixed, it becomes a C color). ). When the C color is not a constituent color of the 1D color bit code, the area 32 has been treated as noise in the conventional reading method. If the C color is regarded as one of the cells of the 1D color bit code, decoding is performed based on a color change different from the original color change, which may cause inconvenience.

  However, when it is known in advance that the C color is a color obtained by mixing the A and B colors that are cell constituent colors, it is not always appropriate to recognize the region 32 as noise. In particular, when the area of the C color region is large, it is not appropriate to determine that the C color region is a region sandwiched between two 1D color bit codes. It is appropriate to treat the C color area as either the A color area or the B color area.

  Therefore, in the present embodiment, the region 32 that is the C color is regarded as a part of either the color region 30A or the color region 30B through the steps described below.

  When the “color area number” assigned to the color area 30A shown in the lower diagram of FIG. 4 (1) is 3, the color information group “3, A, 2, 4”, “4, C, 3, 5” and “5, B, 4, 6” are generated.

  The decoding means corrects this color information group to “recognize the C color area as a part of the A color area” or “C color area as the B color based on the user's instruction. Is considered to be part of the color area ".

  Specifically, when the former process is performed, first, the decoding unit searches for color information whose color information is C and color information whose color information is A. In the present embodiment, the color information “3, A, 2, 4” and “4, C, 3, 5” are applicable.

  Next, the decoding means regards these two pieces of color information as originally constituting one A color region, generates color information “3, A, 2, 4”, and generates two pieces of color information “ 3, A, 2, 4 ”,“ 4, C, 3, 5 ”. In addition, correction is performed by appropriately shifting the numbers in the other color information.

  As a result, the color information group “3, A, 2, 4”, “4, C, 3, 5”, “5, B, 4, 6” has the color information group “3, A, 2, 4”. , “4, B, 3, 5”. By performing decoding based on the color information group, it is realized that “the region 32 that is the C color is regarded as a part of the color region 30A”.

  Note that the color arrangement information is not corrected as described above, but before the decoding unit divides the captured image into a plurality of color areas, “replace C color pixels with A color pixels” or “C It is also a preferred embodiment to perform image processing “replace color pixels with B color pixels”.

  The processing result 34a in FIG. 4 (2) considers the region 32 to be a part of the color region 30A by replacing the C color pixel in the region 32 with the A color pixel in image processing. This is an example. The processing result 34b is an example of a case where the area 32 is regarded as a part of the color area 30B by replacing the C color pixel in the area 32 with the B color pixel in image processing. is there. Based on either of these two processing results 34a and 34b, the decoding means generates the same color information group “3, A, 2, 4”, “4, B, 3, 5” as described above. Generate.

  According to the method described above, since the color information whose color information is C does not enter between the color information whose color information is A and the color information whose color information is B, there is no problem. The possibility of correctly recognizing the color bit code is increased.

(4-3) Color averaging process When capturing a 1D color bit code marked on an object, the background appears in the captured image or the color of the 1D color bit code changes slightly. There is. As a result, if no image processing is performed on the captured image, the decoding unit may read a color arrangement different from the original color arrangement of the 1D color bit code from the captured image, and correct data may not be obtained. There is.

  Therefore, in the present embodiment, in order to correct the colors in the captured image, a color averaging process is performed on the captured image. Specifically, the color averaging process includes, for example, a median filter, luminance averaging, and shading correction.

  FIG. 5 shows a captured image 38 obtained by capturing the end surface of the curved plate-like body 36. A 1D color bit code 40 is marked on the end face of the plate-like body 36. Since the 1D color bit code 40 is marked on the end face of the plate-like body 36 without a gap, the end face of the plate-like body 36 itself is not shown in the captured image 38. The periphery of the 1D color bit code 40 is a background.

  A two-dot chain line in the captured image 38 is a part of the field frame 42. In the drawing, a portion sandwiched between two two-dot chain lines is the inside of the field frame 42. The shaded portion in the figure shows the background or noise inside the visual field frame 42 and is not a color region constituting the 1D color bit code.

  The decoding unit extracts a part of the captured image 38 with the field frame 42. Next, the decoding means uses, for example, a median filter to replace the shaded portion in the field frame 42 with the constituent color of the 1D color bit code. As a result, the inside of the visual field frame 42 is a color region of the constituent colors of the 1D color bit code.

  Thereafter, the decoding unit divides the narrow image after the replacement into a plurality of color regions, executes the reading method described in (2) above, and acquires data from the 1D color bit code 40.

  According to the present embodiment, even if noise is mixed to some extent in the captured image, the decoding result is not affected. Therefore, even if the user performs a somewhat rough capture operation, a very beneficial effect of correctly decoding the 1D color bit code can be obtained.

Comparison of 1D Color Bit Code and One-Dimensional Barcode One of the features of the 1D color bit code reading method in this embodiment is that a part of an image is cut out by a linear field frame. On the other hand, this method of reading a code symbol with a linear field of view (frame) has already been used for reading a black and white barcode (also called a one-dimensional barcode).

  Therefore, in terms of using a linear field of view, the conventional one-dimensional barcode reading method and the 1D color bit code reading method of the present embodiment are partially common in that a field frame is used. There may be a matter.

  However, since the data representation is different between the 1D color bit code and the one-dimensional barcode, there is a great difference in whether color averaging processing (for example, pixel replacement) can be performed for noise color processing.

  If the 1D color bit code 40 in FIG. 5 is a one-dimensional barcode, inconvenience may occur. One-dimensional barcode data is represented by the thickness of a striped line. The width of each line (black and white (light / dark) symbol) is important. This is because when there is an ambiguity in the boundary portion between the black and white (bright and dark) symbols, or when the width of the black and white (bright and dark) symbols changes, decoding may not be performed correctly.

  On the other hand, since the 1D color bit code returns a code (data) by recognizing a color change in the longitudinal direction of the visual field image, the length of each cell of the 1D color bit code is not related to the data to be returned. Regardless of the length of each cell, it is possible to perform color averaging to remove noise. Therefore, the 1D color bit code can reduce the influence of the noise color and output data correctly even if noise is mixed in a narrow image to some extent.

(4-4) Color Uniform Processing in Short Side Direction The above-described color averaging processing is performed without distinguishing between constituent colors of 1D color bit code and colors that are not constituent colors of 1D color bit code. Process with. If it is known in advance that there is a color that is not a constituent color of the 1D color bit code, it is more preferable to perform a color averaging process after removing the color from the captured image.

  Therefore, in the present embodiment, as a noise countermeasure, there is provided a step of performing short side direction color equalization processing for uniformizing pixels in the short side direction of a narrow image. This short side direction color equalization process is performed before the color averaging process.

  FIG. 6A shows a captured image 44 obtained by capturing a 1D color bit code. This 1D color bit code is marked on the end face of the curved plate-like body 46 without a gap. Further, the position of the field frame 48 is shown in the captured image 44. Inside the field frame 48, a 1D color bit code and a background are shown. Note that, outside the field frame 48, only the outline of the plate-like body 46 is shown by a broken line.

  In many cases, the cell constituent colors of the 1D color bit code are, for example, red, green, and blue. This is because it also matches general image signal components and is easy to handle in terms of image processing. In many cases, the background portion has a color with low saturation and lightness (for example, gray) due to interference with room illumination light and reflection on various objects. This background portion is represented as Qz.

  Note that “Qz” in this section (4-4) has a different meaning from “Qz” in (3) above. “Qz” in (3) is a term indicating the “quiat zone” described in the background art. As described in the background art, this quiet zone plays a role of partitioning a 1D color bit code area and a surrounding area that is not a 1D color bit code. On the other hand, “Qz” in this section is a term used simply to indicate an excluded area. In short, “Qz” in this section refers to so-called “noise”.

  The decoding means first cuts out a part of the captured image 44 by the field frame 48.

  Next, the decoding unit reads the value of each pixel of the captured image 44. As a result of reading the value of each pixel, when the ratio of the number of pixels corresponding to Qz to the total number of pixels is higher than a predetermined reference, the decoding means It is determined that there is a high possibility that Qz exists in Then, the Qz pixels are subjected to color averaging processing for each pixel group that can represent one thin line in a direction parallel to the short side direction of the narrow image or a row of pixels arranged in the short side direction of the narrow image. A color averaging process (for example, a median filter) is performed only on a pixel group that is a constituent color of the 1D color bit code. Note that it is also preferable to perform the above processing for each of a plurality of columns of pixel groups arranged in the short side direction of the narrow image or a pixel group that can express a slightly thicker straight line in a direction parallel to the short side direction of the narrow image. In the present embodiment, this processing step is referred to as short side direction color equalization processing.

  The decoding means performs this short side direction color uniformity processing for all pixels from one end of the narrow image to the other end. This is shown in FIG. 6 (2).

  Next, the decoding means divides the narrow image into a plurality of color regions based on the value of each pixel after the short side direction color equalization processing. FIG. 6 (3) shows the narrowed image 50 after segmentation. The narrow image 50 has no Qz color region.

  As a result, the decoding unit can perform decoding based on an image in which the Qz color region does not exist or very little, and thus decoding processing with higher reading accuracy is possible.

  In addition, in FIG. 6A, when it is known in advance that the color of the boundary portion of the cell group constituting the 1D color bit code is a mixed color of colors (A color and B color) of adjacent color regions. As described in (4-2) above, it is preferable to perform a uniformization process peculiar to the 1D color bit code that recognizes a pixel group of Qz by a combination of values of “A color or B color”. If the A color and the B color are simply averaged (for example, the RGB values of each pixel are averaged, for example), the transition of the color may change, and the consistency with the process for detecting the color change may be deteriorated. It is.

  If it is known in advance that the color of Qz is a color in which the constituent colors of the 1D color bit code are mixed, processing with a method suitable for the method will make it possible to achieve even more accurate decoding characteristics. It becomes easy to realize.

Application Example If the short side direction color equalization process is used, a 1D color bit code can be read even if noise and background are mixed in a narrow image.

  Therefore, on the premise that noise and a background portion enter the inside of the field frame, the decoding unit can give the field frame a certain width (longer short side). When a 1D color bit code having a slight curve is read, some constituent cells of the 1D color bit code may not be able to fit inside the field frame. In the present embodiment, since the decoding means reads only the inside of the field frame, the constituent cells that cannot fully enter the field frame are not decoded. If the arrangement of the 1D color bit code and the shape of the field frame are slightly different, the inability to read the 1D color bit code may be less convenient for the user.

  Therefore, if the decoding means adjusts the length of the short side of the field frame to be longer in accordance with the user's instruction and the noise and background in the narrow image are removed, the decoding means is a 1D color bit code having a slight curve. Can also be read. As a result, the decoding means can correctly read the data based on the image including the entire bent 1D color bit code, so that it is possible to obtain a very beneficial effect that the convenience is greatly improved by the user.

Order of performing short side direction color uniformization processing and color quantization processing In the 1D color bit code reading method described in Japanese Patent Application No. 2007-130504 filed earlier by the present inventor, Perform color quantization. This color quantization refers to a process of dividing a captured image into a plurality of color regions.

  On the other hand, in the reading method according to the present embodiment, a pattern in which short side direction color equalization processing is performed after color quantization (that is, color quantization is performed first), and after short side direction color equalization processing, It is conceivable that the pattern is converted (that is, color quantization is performed later), but the latter pattern is typical for the reason described below.

  As shown in FIG. 6A, a captured image may include a 1D color bit code and a background. A 1D color bit code is often composed of cells of a color with high saturation and lightness such as red, green, and blue. On the other hand, the color of the background often becomes a color with low brightness and saturation (for example, gray). When comparing the saturation and brightness of these colors, it is considered that the color of the cell forming the 1D color bit code is higher than the background color in both saturation and brightness.

  Since there is a big difference between the pixel value of the background portion and the pixel value of the constituent cell portion of the 1D color bit code, the color quantization is performed after removing Qz by the short side direction color equalization processing first. Is advantageous for the purpose of highlighting the color cells forming the 1D color bit code. That is, in the present embodiment, it is preferable to perform processing in the order of “short side direction color uniformity processing → quantization”.

  Further, since it may be difficult to completely remove the noise in the short side direction only by the uniformization process in the short side direction of the narrow image, it is more preferable to perform the noise reduction process in the longitudinal direction.

  Therefore, a typical processing order in the short side direction color uniformization processing is “capture” → “short side direction uniformization processing” → longitudinal direction “noise reduction processing” → quantization.

(5) 1D color bit code marking method and improvement example of reading method (5-1) Constituent colors of 1D color bit code In a conventional 1D color bit code reading method not using a field frame, for example, “each cell Was tracking a cell that satisfies the boundary condition of “is touching one or two other cells”. A cell next to the cell is searched, then another cell adjacent to the cell is searched for, a cell adjacent to the cell is further searched, and the cell is traced one after another. If the series of cells branches into two and the number of adjacent cells is three or more, the boundary condition is not satisfied, so that the 1D color bit code cannot be cut out.

  On the other hand, in the present embodiment, the decoding means detects and decodes the 1D color bit code based only on the color change in the longitudinal direction (linear direction) of the narrow image. The decoding means regards the color in the short side direction (width direction) of the narrow image as uniform by averaging (for example, pixel replacement) and ignores the color change in the short side direction (width direction) of the narrow image. ing. In the reading method according to the present embodiment, it is assumed that a cell in the longitudinal direction of a narrow image is an adjacent cell.

  Therefore, since it is only necessary to capture a one-dimensional color change in the longitudinal direction (straight direction) of the narrow image, it is not always necessary to provide Qz in the short side direction (width direction) of the narrow image.

  On the other hand, in the longitudinal direction of the narrow image, it is necessary to provide Qz representing the front end and the end of the 1D color bit code. This is because an image other than the 1D color bit code may appear in the longitudinal direction of the narrow image, and it is necessary to clarify the boundary between the 1D color bit code and the image other than the 1D color bit code.

  Of course, when it is defined that “the longitudinal direction of the narrow image is filled with the image of the 1D color bit code”, it is not necessary to provide Qz in the vicinity of the long side of the narrow image. Since the length of the field frame in the longitudinal direction is finite, Qz is not necessary if all the 1D color bit code cell groups fit within the field frame without excess or deficiency.

  Here, “over” of “over / under” indicates a case where the long side of the field frame is longer than the 1D color bit code. When the long side of the field frame is longer than the 1D color bit code, it is necessary to provide Qz in order to clarify the boundary between the 1D color bit code and other images. On the other hand, “insufficient” refers to a case where the short side of the field frame is too short and only a part of the 1D color bit code can enter the field frame. In this case, the 1D color bit code outside the field frame is not read. Therefore, it is necessary to provide Qz representing the start of the 1D color bit code and Qz representing the end, and use them as marks to place the entire 1D color bit code inside the field frame.

  When Qz is unnecessary, the color used for Qz may be used to represent data. In general, white and black are often reserved as the color of Qz, but in addition to white constituting the color of the 1D color bit code, the 1D color bit code is composed of three colors {red, green, white}. You may do it.

(5-2) Marking method when white is one of the constituent colors of the 1D color bit code
Example 1
In a real scene, marking of a 1D color bit code using three colors of A, B, and white is generally performed by first applying white over a wide area, and then A cell and B color on it. This is done by applying a cell.

  FIG. 7A is a schematic view of the surface of the object 52 to be marked. It is shown that a white background 54 is formed over a wide range of the surface of the object 52 by applying W ink, which is a constituent color of the 1D color bit code, as a background color. Here, the wide range means wider than the field frame. The hatched portion in FIG. 7 (1) indicates the background color.

  FIG. 7B is a schematic view of the surface of the object 52 after forming the constituent cell group 56 (A color cell and B color cell). The A-color ink or the B-color ink is overcoated on the white base 54 to constitute a group of constituent cells 56. The A-color cells and B-color cells constituting the group of constituent cells 56 are formed with a space therebetween. The white base 54 is exposed as it is between the A-color cell and the B-color cell, and this exposed portion becomes a white cell constituting the 1D color bit code.

  FIG. 7 (3) shows a captured image 58 of a 1D color bit code. The field frame 60 is displayed in advance on the capture screen of the decoding unit, and the decoding unit cuts out a narrow image inside the field frame 60 using the field frame 60.

  As shown in FIG. 7 (3), there are only A-color cells, B-color cells, and white portions in the field frame 60. As described above, if the color used for Qz is used as one of the constituent colors of the 1D color bit code, the situation that “the 1D color bit code can be accommodated in the field frame without excess or deficiency” can be easily realized. Can do.

  FIG. 7 (4) shows a narrowed image 62 that has been cut out. The decoding means performs color averaging processing or short side direction color uniformity processing on the narrow image 62, and divides the narrow image 62 into a plurality of color regions. When it is determined that the narrow image 62 satisfies the conditions for establishing the 1D color bit code, decoding is performed by the reading method described in (2) above.

Example 2
In order to further improve the accuracy of reading, it is also preferable that white cells are marked again on white cells other than both ends of the 1D color bit code. Even when the white background is soiled when marking the A-color cell and the B-color cell, an advantageous effect is obtained in that the white cell is marked from above the stain so that it can be read correctly.

  In addition, when the constituent colors of the 1D color bit code are three types of {red}, {green}, and {color that is neither red nor green}, white cells other than both ends of the 1D color bit code are used. The same effect can be obtained by marking the yellow color.

Effect of using background color as constituent color of 1D color bit code As described above, in the present embodiment, white is applied to the surface of the object to be printed as the base color, and the portion where the base is exposed is configured of the 1D color bit code. It is also preferable to use it as a cell. If the background color is used as one of the 1D color bit codes in this way, the following three effects can be obtained.

  The first effect is that an achromatic color (white, black, gray, etc.) or a bright color (white, bright yellow, etc.) can be used as a constituent color of the 1D color bit code. is there.

  As the color of Qz, conventionally, achromatic and bright colors are often used. The achromatic and light colors are common as background colors and can be captured stably, so the achromatic and light colors are used for Qz in order to maintain decoding accuracy and quality. Because was the most preferred.

  In addition to the Qz color, it is necessary to select a color that is not an achromatic color system or a light color system (for example, R (red), G (green), and B (blue)). A color that is not an achromatic color system or a light color system may be more difficult to separate colors than white. For this reason, if R, G, and B are cell constituent colors, decoding may not be performed correctly.

  On the other hand, in the present embodiment, since it is not necessary to provide Qz, it is possible to use the achromatic color and light color that have been assigned to Qz as the constituent colors of the 1D color bit code. Since a stable color can be used as a constituent color of the 1D color bit code as a capture target, the reading accuracy is improved as compared with the conventional case.

  The second effect is that it is easy to distinguish colors. As the constituent colors of the 1D color bit code, three colors of R (red), G (green), and B (blue) are often employed. This is because these three colors are the three primary colors, and color recognition is easy. However, these three types of hues of R, G, and B may be misaligned between human vision and the sensitivity measured by the camera. In particular, it is relatively difficult to distinguish {G} and {B} using a camera. It is possible that the order of arrangement of the G cell and the B cell is erroneously recognized. On the other hand, the distinction between {R} and {G}, {R} and {B}, and {R, G, B} and {W} are often relatively easy. .

  Therefore, if W is used as a background color, a 1D color bit code can be configured by a combination of {R, G, B} and {W}. Since the colors constituting the 1D color bit code can be more easily classified using a camera, a beneficial effect that the reading characteristics are remarkably improved can be obtained.

  The third effect is that the number of colors required for attaching the 1D color bit is reduced, and the selection of colors becomes very easy. In a conventional 1D color bit code in which three colors are cell constituent colors, a total of four types of colors, that is, three types of colors constituting the 1D color bit code and one color representing Qz are required. On the other hand, in the present embodiment, it is not necessary to provide a color for expressing Qz. As a result, there was obtained a beneficial effect that the types of colors necessary for attaching the 1D color bit code could be reduced from four to three. In addition, there is an effect that the marking of the 1D color bit code is further simplified.

(5-3) Redundant cell for cutout In FIG. 7 (3), when the length of the long side of the field frame 60 is longer than the white background portion (white background 54), the non-background portion (white background) The portion where 54 is not marked also enters the field frame 60 and is captured. The color of the non-background portion is indefinite, and the non-background portion itself may have some color pattern (pattern). Therefore, in this case, it is preferable to exclude the non-background portion from the decoding target. Here, a method of excluding the non-background portion from the decoding target (that is, decoding only the 1D color bit code) using the cut-off redundant cells 68a and 68b will be described.

  FIG. 8 shows a plate-like body 64. A white background 66 is marked on a part of the end face of the plate-like body 64, and cells of colors other than white are marked on the white background 66 among the cell constituent colors of the 1D color bit code. Thus, as shown in the figure, the cutout redundant cell group 68a, the 1D color bit code 70, and the cutout redundant cell group 68b are marked in this order from the left end to the right end.

  The cutout redundant cell groups 68a and 68b are each composed of five cells (including white cells) and represent the same color arrangement. These cut-out redundant cell groups 68a and 68b indicate that a 1D color bit code 70 exists between them.

FIG. 9 shows a captured image 72 obtained by capturing the 1D color bit code 70 from above the end face of the plate-like body 64. In FIG. 9, the length of the long side of the field frame 74 is longer than that of the white base 66. For this reason, the non-background portion (indefinite color) enters the field frame and is captured. Even in such a case, the 1D color bit code can be easily cut out by indicating the existence of the 1D color bit code by the redundant cell groups for cutting 68a and 68b. When the decoding means determines that the image inside the preset field frame 74 satisfies the conditions for establishing the 1D color bit code, the cell group between the clipping redundant cells 68a and 69b is the 1D color bit code. Decoding is performed assuming that 70 colors are represented.
According to the present embodiment, the 1D color bit code can be read regardless of the color pattern (pattern) of the marking object itself marking the 1D color bit code. .

(6) Improved example of operability of decoding means (6-1) Variations in number, arrangement, and shape of field frames In this embodiment, it is easy to display a captured image on the screen of the decoding means. It is also easy to display an image inside the field frame. When reading the 1D color bit code with the field frame, it is preferable that the user can simultaneously recognize the positions and ranges of both the captured image and the field frame on the capture screen of the decoding unit. If both the captured image and the image inside the field frame are displayed on one screen at the same time, it is convenient for the user to easily operate and instruct. Further, if the field frame is displayed in the captured image, it is easy for the user to aim and further facilitate the operation.

  FIG. 10 shows an example of variations in the number, arrangement, and shape of the field frames.

  FIG. 10A shows an example in which one linear field frame 78 is set on one capture screen 76. The field frame in the present embodiment is not necessarily linear. Further, the number of field frames is not limited to one. If it is more convenient in terms of reading operation to have a plurality of field frames on one capture screen, a plurality of field frames can be set on the capture screen. When there are a plurality of field frames on one capture screen, there can be many variations for improving operability such as arranging the field frames in parallel or by changing the angle.

  The capture screen 80 in FIG. 10B is an example in which five linear field frames are provided in parallel. To read the 1D color bit code, it is necessary to match the image of the cell group constituting the 1D color bit code with the field frame. If there are multiple field frames, one of the field frames matches the 1D color bit code. Just do it. As a result, the 1D color bit code can be easily captured, so that the user can easily operate.

  The capture screen 82 shown in FIG. 10 (3) is an example in which nine linear field frames are provided with a set of three, each having a slightly different angle. As described above, in order to read the 1D color bit code, it is necessary to match the position of the field frame with the 1D color bit code, but it is also necessary to match the angle. However, it may not be easy to match the angle of the 1D color bit code with the field frame. Therefore, if a plurality of field frames are provided with the angles slightly shifted as in the capture screen 82, the angle between the field frame and the 1D color bit code can be easily adjusted, and the operation is easier for the user. become.

  The capture screen 84 in FIG. 10 (4) is an example in which one field frame of a “he” shape is arranged. In addition, the “variant shape” described in FIG. 10 means that it is not linear. When the 1D color bit code is marked in a “h” character shape, it is easier to read by using a “h” character field frame accordingly.

  The capture screen 86 in FIG. 10 (5) is an example in which one “U” -shaped field frame is arranged. Similarly to the capture screen 84, when the 1D color bit code is arranged in a “U” shape, it is easy to cut out if the shape of the field frame is set to the “U” shape in accordance with the 1D color bit code. It becomes easier to read.

  In the present embodiment, it is important that the narrow image taken out by the field frame has the directivity of the short side direction and the long side direction. This is because the decoding is performed based on the color change only in the longitudinal direction of the narrow image. Even if one of the edges of a narrow image is curved or the width in the short side direction changes (even if it is not constant), the two directions of the narrow image (longitudinal direction and short side direction) are clear. If so, reading is possible.

  The field frame does not need to be fixed within the field of view of the capture screen, and it is conceivable that the decoding means moves or rotates the field frame so as to scan up and down in the capture screen. In the present embodiment, the decoding means corresponds to a preferred example of the field frame moving means described in the claims. Moreover, there may be a plurality of field frames on one capture screen. It is also preferable to change the width and length of the field frame in accordance with the arrangement of the 1D color bit code. If the user can change the shape of the field frame according to the width and length of the 1D color bit code, it is assumed that there are many cases where it is convenient.

  The variation of the number and shape of the field frames can be set by software after capturing an image. It is possible for the user to freely select a variation of the field frame or to change it arbitrarily.

(6-2) A method of associating the position of the 1D color bit code with the position of the field frame When there are a plurality of tags marked with one 1D color bit code in the captured image, the decoding means If the data represented by the 1D color bit code and the position information are associated and presented on the captured image, convenience for the user is improved. In the reading method using the field frame, it is assumed that many of the read 1D color bit codes are generally arranged in parallel. This is because the capture camera is often moved up, down, left and right for reading. If it is possible to recognize the position of a plurality of 1D color bit codes that have been read and the position of the field frame when these 1D color bit codes are read in association with each other, it is even more preferable because the positional relationship between the tags can also be recognized.

  As described in (6-1) above, the user can freely set the position of the field frame in the captured image. In addition, it is conceivable to set the position of the field frame by various methods as described below.

  a. While capturing the 1D color bit code as a moving image, the field frame is automatically moved (scanned) within the captured video.

  Specifically, on each continuously captured image, a field frame is set by moving the position little by little for each image, and the position of the field frame and the color arrangement code matches. If decoding processing is performed on such a captured image, this processing can be performed. More specifically, out of the field frames set in each image, for those in which the color region in the field frame is configured by the colors constituting the color arrangement code, a narrow image in the field frame is extracted, It is the object of decoding. In this case, the method of moving the field frame in each captured image may be moved in parallel or rotated. The number of field frames may be singular or plural.

  b. The field frame is automatically scanned on the captured still image. The number of field frames may be singular or plural.

  c. The field frame is fixed at a predetermined position in the captured image, and the user moves (scans) the entire moving image capture field by moving the camera body. The number of field frames may be singular or plural.

  With the above method, even if there are a plurality of tags, any tag can be read.

  However, if the field frame position is set by the methods a to c and the tag is captured, it may be difficult to perform the following three processes in the subsequent processes.

1. Recognition of tag positions
Since both the position of the field frame and the tag move on the captured image, it is difficult to recognize the positional relationship between the two. In addition, when there are a plurality of tags, it is difficult to recognize the positional relationship between the tags.

2. Same tag distinction
Since the position of the field frame moves, it is difficult to determine that one tag has been recognized a plurality of times.

3. Processing of tags that failed to read It is desirable to retry reading a tag that has failed to be read. However, since the position of the tag that has failed to be read is not tracked, it is difficult to retry reading again once the tag whose field frame has not been read is left.

  In order to perform these three processes, it is preferable to use the following two methods together.

Method 1 (tag position recognition)
Method 1 is a method of recognizing the tag position using the position information of the field frame.

  The decoding means stores the position information of the visual field frame (decoded visual field) when the 1D color bit code can be decoded linked to (attached to) the captured image. The position information is coordinate information of each pixel existing inside the field frame. Each time the decoding means obtains a decoding result from the 1D color bit code, the decoding result and the position information of the field frame when decoding is possible are linked (pasted) to the captured image and stored. When decoding can be performed a plurality of times, a plurality of pieces of position information are linked to the captured image.

  As a result, all the decoded 1D color bit code positions are stored linked to the captured image, so that the decoding means reads all the positional information linked to the captured image and displays the position on the capture screen. If a frame or the like indicating is displayed, it is easy for the user to recognize all of the read tag positions.

Method 2 (Distinction of the same tag)
Method 2 is a method in which when the decoding unit determines that the same tag is based on the position information of the field frame stored by executing method 1, the number is counted as one. There are two ways of judging whether or not they are the same tag.

  The first way of thinking is based on the criterion “more than a factor of the width of the field frame”. The position information of the field frame newly obtained by the decoding means and the position information of the field frame already obtained by the decoding means are compared. If they are separated by “more than a factor of two”, it is determined that the tags appearing in the field frames are not the same. If it is within the standard, the decoding means determines that the same tag has been read, and discards the newly obtained position information and decoding result.

  The second concept is based on the “predetermined distance”. Similar to the first idea, if there is a separation equal to or greater than the reference value between the new field frame and the previous field frame, it is determined that the tags are not the same. If it is within the standard, the decoding means determines that the same tag has been read, and discards the newly obtained position information and decoding result.

  The user can appropriately adopt either or both of the concepts according to the length of the field frame or the density of tags to be read. As a result, since the decoding means can present the position of the read tag to the user, the operability for the user is improved. In addition, since the user can know which tag is the tag that has not been read by knowing the position of the read tag, the operability for the user is greatly improved.

(7) Tracking of 1D color bit code In the most basic reading method of 1D color bit code, the color area of the constituent color of the 1D color bit code in the captured image is tracked based on only the color, and 1D color bit code is recorded. Cut out.

  However, in this tracking method based only on color, the direction in which the 1D color bit code is tracked is not restricted, so that it may be difficult to read depending on the arrangement of the 1D color bit code.

  Therefore, the present inventor has developed two improved examples in which the tracking direction of the 1D color bit code is regulated.

  As described above, the improvement example 1 has a narrow field frame (the field frame may be arranged in a curved shape like the above-described “U” -shaped field frame or “He” -shaped field frame). Is included). As described above, in this reading method, the tracking direction is regulated by assigning a field frame to the captured image of the 1D color bit code.

  The improvement example 2 is a reading method using the linearity of the 1D color bit code. The arrangement of the 1D color bit code is not related to data, but generally the 1D color bit code is often marked in a straight line. Therefore, when a 1D color bit code marked in a straight line is read, tracking can be performed even more easily by using the linearity of the 1D color bit code. Hereinafter, two specific examples of the improved example 2 will be described.

Tracking example 1
FIG. 11 shows a captured image 88 including a 1D color bit code. The 16 color regions 90a, 90b, 90c, 90d, 90e,..., 90k, 90k + 1, 90k + 2,..., 90n-1, 90n in the figure constitute a 1D color bit code. The color area 90q is a color area having the same color as the constituent color of the 1D color bit code, but does not constitute the 1D color bit code. In addition, the periphery of the color region 90 group is a color region of a color that is not a constituent color of the 1D color bit code.

  In the present embodiment, it is known that (A) the color area 90a is the first cell as an initial condition. Since the color area 90a includes “a pixel at a predetermined position”, it is known from the beginning that the color area 90a is the first cell of the 1D color bit code. (B) Each cell constituting the 1D color bit code is marked based on a boundary condition that “each cell is in contact with one or two other cells”. (C) Marking is performed based on the rule that “the cell group constituting the 1D color bit code does not bend sharply”. The decoding unit tracks the 1D color bit code by sequentially tracking the color range (color region) having the highest linearity using these three rules.

  The decoding means first selects a color area of the constituent color of the 1D color bit code from other color areas adjacent to the color area 90a, and regards the color area as the second cell of the 1D color bit code. In the present embodiment, the color region 90b is the second cell.

  Similarly, the decoding means sequentially selects the color areas of the constituent colors of the 1D color bit code, and sets the color areas 90c, 90d, 90e,..., 90k, 90k + 1 to the third cell, the fourth cell, and the fifth cell, respectively. Cell,..., Kth cell, k + 1th cell.

  Next, the decoding means tracks the color area which is the k + 2 cell. As shown in FIG. 11, the color region 90k + 1 cell is in contact with the color region 90k + 2, the color region 90q, and other surrounding color regions. The color region 90k + 2 and the color region 90q are both color regions of the constituent colors of the 1D color bit code, and there are two color region candidates that are the k + 2 cell. I can't track it. Therefore, the decoding means tracks the 1D color bit code using the above-mentioned definition (C).

  FIG. 12 shows an enlarged view of the inside of the frame 92 in FIG. The decoding means first calculates the positions of the centroid points 94k + 1, 94k, 94k + 2, and 94q in the color regions 90k + 1, 90k, 90k + 2, and 90q.

  Next, the decoding means draws straight lines L (k), L (k + 2), and L (q) passing through the barycentric point 94k + 1 and the barycentric points 94k, 94k + 2, and 94q for each barycentric point.

  Next, the decoding means calculates an angle α (k + 2) formed by the straight line L (k) and the straight line L (k + 2) and an angle α (q) formed by the straight line L (k) and the straight line L (q). The straight line having the highest linearity with respect to the straight line L (k) is selected by comparison. In this embodiment, since the angle α (k + 2) is smaller, the decoding means determines that the straight line L (k + 2) has the highest linearity and selects it.

  Next, the decoding means determines that, out of the two centroid points on the straight line L (k + 2), the color region 90k + 2 to which the centroid point 94k + 2 that is not the centroid point 94k + 1 belongs is the k + 2 cell of the 1D color bit code. .

  The decoding means further tracks the color area up to the color area 90n. As a result, according to the present embodiment, it is possible to more easily track and decode the color area group constituting the 1D color bit code.

Tracking example 2
FIG. 13 shows a plate-like body 96. On the end surface of the plate-like body 96, an end point cell 98a, a cutout redundant cell group 98b, and a data cell group 98c are marked from left to right in the drawing. These three cell groups constitute a 1D color bit code 98. One of the features of the present embodiment is that the end point cell 98a is marked in a color range (region) that is longer in the longitudinal direction of the end face than the other cells and much larger than the other cells. The end point cell is sometimes called a start cell.

  FIG. 14 shows a captured image 100 in which the 1D color bit code 98 is captured by the decoding unit. In the drawing, a color region group shown around the 1D color bit code 98 indicates a background pattern.

  In the present embodiment, a 1D color bit code is normally captured and read by a field of view without using a field frame. The initial conditions are the following two.

  The first condition is that the end point cell represents a predetermined color and exists in contact with a predetermined image end in the captured image. The predetermined image edge means an edge designated by the user among the edges of the captured image. It may be a part of the edge. In the present embodiment, the end point cell is conditioned to be in contact with the left edge of the captured image 100.

  The second condition is a condition in which redundant cell groups are continuously cut out from the end point cells. The cut-out redundant cell group is composed of four cells and represents an array of colors defined in advance.

  The user captures the 1D color bit code 98 by moving the decoding means so that the end point cell 98 a covers the left edge of the captured image 100. As described above, since the color region of the end point cell 98a is very large, it is easy to move and capture the color region of the end point cell 98a so as to cover the left end of the capture scene 100.

  The decoding unit divides the captured image 100 into a plurality of color regions based on the values of the respective pixels, and searches for the end point cell 98a based on the values of the pixels in the color region group in contact with the left edge of the captured image 86. When a color area having the same color as the color assigned to the end point cell 98a is found, the color area is regarded as the end point cell 98a.

  Next, the decoding means finds the first cell of the cut redundant cell group 98b based on the value of each pixel from the four color regions adjacent to the end point cell 98a. Further, the decoding means sequentially tracks the second cell, the third cell, and the fourth cell of the cutout redundant cell group 98b based on the pixel value.

  Next, the decoding means recognizes the positional relationship between the end point cell 98a and the cell group constituting the cutout redundant cell group 98b on the captured image 100, and based on the positional relationship, the decoding direction ( Angle).

  Next, in the same manner as in Tracking Example 1 described above, the decoding means successively selects the color areas existing at the position having the highest linearity with respect to the arrangement direction of the cutout redundant cell group 98b, thereby configuring the data cell group 98c. The color area group to be tracked is tracked. Since the cell group column of the 1D color bit code 98 has a relatively small curvature and is arranged in a straight line, the 1D color bit code 98 can be cut out by the method described in the tracking example 1.

  In the present embodiment, the end point cell 98a has an initial condition that it touches the field edge (image edge) of the captured image, but is not at the image edge, for example, a pixel that exists at a determined position within the field of the captured image. Even if the color area including is defined as an end point cell, the 1D color bit code can be cut out without any problem. In this case, if the “predetermined position” is displayed on the screen by, for example, an “x” mark, the user can easily adjust the end point cell to the “predetermined position”, so that the operability is greatly improved. To do. The operability can also be improved by moving the “predetermined position” so as to automatically scan within the capture screen. Specifically, a plurality of “predetermined positions” are defined in the field of view of the captured image, and if any of the “predetermined positions” is a color representing an end point cell, it is regarded as an end point cell. . As a result, even if the end point of the color arrangement code is not located at one “predetermined position”, if the end point is located at any other “predetermined position”, the end point is “automatically” from within the image. Cells can be extracted and color array codes can be cut out. Note that this processing is performed by selecting “a color area of a predetermined color from a plurality of predetermined positions in the claims”, and the selected color area is a start cell of the color arrangement code. It is a preferable example of “considering”.

(8) Improvement example of 1D color bit code (8-1) Write-once type 1D color bit code For example, in a manufacturing process in which each process is complicated, a code may be added to a product for each process unit. Often desired. This is for easily and accurately managing what process the manufactured product has taken.

  Therefore, the present inventor first developed a write-once type 1D color bit code configured by consecutively marking a plurality of code symbols (1D color bit code) on the same color sequence.

Specifically, for example,
(A) A write-once type 1D color bit code comprising four types of cells, ie, a cell of Qz color and three types of cells not of Qz color, has been developed. This write-once type 1D color bit code is composed of a plurality of 1D color bit codes, and each 1D color bit code is composed of three types of cells that are not Qz color. A Qz cell is inserted between each 1D color bit code. At the time of reading, the decoding means divides one recordable 1D color bit code into a plurality of 1D color bit codes with Qz as a mark. Thus, the decoding means can simultaneously read a plurality of 1D color bit codes.

  However, in this embodiment, as described above with reference to FIG. 7 and the like, the color of Qz may be the constituent color of the cell of the 1D color bit code. In this case, it may not be possible to know whether the cell of Qz is a cell representing data or a cell representing a break in the 1D color bit code. Therefore, the 1D color bit code of (A) may not be suitable for this embodiment.

  (B) In addition, the present inventor disclosed a write-once type code in Japanese Patent Application No. 2008-200624 filed earlier. This write-once code is composed of a plurality of 1D color bit codes. Each 1D color bit code has a predetermined number of constituent cells and start and end colors. Then, a plurality of 1D color bit codes are continuously marked to constitute one write-once code. At the time of reading, the decoding means can read the write-once code by dividing it into individual 1D color bit codes based on the cell color and the number of cells. According to this write-once code, it is useful to add another 1D color bit code in series.

  Therefore, in this embodiment, a reading method in which the technical matter of the (1D) 1D color bit code described in Japanese Patent Application No. 2008-200264 is applied to the 1D color bit code in this embodiment will be described.

Example 1
FIG. 15 shows an original 1D color bit code 102 and a postscript 1D color bit code 104.

  In the present embodiment, it is defined in advance by the user that “the number of cells constituting the original 1D color bit code is ten”. Further, it is defined in advance that “one white cell is additionally marked at each end of the original 1D color bit code”. These added white cells are end point cells representing the start or end of the original 1D color bit code.

  Further, in the present embodiment, it is previously defined by the user that “the number of cells constituting the appendable 1D color bit code is five”. Further, it is defined in advance that “one white cell is marked adjacent to both ends of the additionally-recorded 1D color bit code”. These additional white cells are end point cells representing the start or end of the additional 1D color bit code.

  Further, in the present embodiment, “a serial 1D color bit code is configured by adding one or more additional 1D color bit codes to one original 1D color bit code” is defined in advance by the user. . In the present embodiment, it is defined that only one end point cell is provided between each 1D color bit code. An end point cell sandwiched between two 1D color bit codes has two roles: an end point cell that represents the end of one adjacent 1D color bit code, and an end point cell that represents the start of the other 1D color bit code. Bear.

  Now, as shown in FIG. 15 (3), a serial 1D color bit code 106 having 24 cells is marked on the surface of the object. The decoding means captures an image including the serial 1D color bit code 106 and generates color arrangement information by the reading method described in (3) or (4) above. In the present embodiment, since the serial 1D color bit code 106 includes 24 cells, the generated color arrangement information includes 24 color information. Note that the color information is information including the number assigned to the color area, the color information, the number of the previous color area, and the number of the subsequent color area, as described above.

  The decoding unit calculates the number of additional write 1D color bit codes included in the serial 1D color bit code 106 based on the number of color information.

  The serial 1D color bit code 106 always includes one original 1D color bit code. Accordingly, of the 24 pieces of color information, 10 pieces of color information are color information obtained from the original 1D color bit code. The other two pieces of color information are color information of end point cells of the original 1D color bit code.

  The postscript 1D color bit code is composed of five cells, and one terminal cell is provided at the end. Therefore, the decoding means calculates that the number of additional write 1D color bit codes included in the serial 1D color bit code 106 is (24-10-2) / (5 + 1) = 2.

Example 2
If the arrangement order of the original 1D color bit code and the additional 1D color bit code is predetermined by the user, the decoding means also determines the delimiter position of each 1D color bit code in the serial 1D color bit code. Can do.

  For example, the serial 1D color bit code 106 in FIG. 15 (3) is defined as “the left end of the serial 1D color bit code is the original 1D color bit, and additional write 1D color bit codes are successively added to the right”. In the case of marking based on the color information, the decoding means outputs the color information from the second to the eleventh among the color arrangement information generated from the serial 1D color bit code 106, and the color information of the original 1D color bit code The 13th to 17th color information is the color information of the additional 1D color bit code, and the 19th to 23rd color information is the color information of the other additional 1D color bit code. It can be determined.

Example 3
When the arrangement order of the original 1D color bit code and the additional 1D color bit code is not determined, the type of each 1D color bit code is distinguished by attaching a specific color to the cells at both ends of each 1D color bit code. can do.

  FIG. 16 shows an example 112 group of the original 1D color bit code 108, the additional write 1D color bit code 110, and the serial 1D color bit code.

  The number of cells of the original 1D color bit code and the number of cells of the additional 1D color bit code are defined in advance by the user in the same manner as in Example 1 described above.

  Characteristic matters in this example 3 are that the cells at both ends of the original 1D color bit code are defined as “A”, and the cells at both ends of the additional write 1D color bit code are defined as “B”. It is. By using these rules, the decoding unit can determine the arrangement order of the original 1D color bit code and the additional 1D color bit code in the serial 1D color bit code.

  FIG. 16 shows an example 112 group of serial 1D color bit codes. First, a description will be given based on the serial 1D color bit code 112a.

  The decoding means captures an image including the serial 1D color bit code 112a, and generates color arrangement information by the reading method described in (3) or (4) above. Since the serial 1D color bit code 112a includes 30 cells, the generated color arrangement information includes 30 pieces of color information.

  Of the 30 pieces of color information, the information on the first to second colors is “W” and “A”. “A” indicates that the original 1D color bit code starts. Since the original 1D color bit code is composed of 10 cells, the decoding means determines that the second to eleventh color information represents the color information of the original 1D color bit code. Since the 11th to 12th color information is “A” and “W”, the 11th color information may be positively determined as the color information of the original 1D color bit code.

  Next, the twelfth to thirteenth color information is “W” and “B”. “B” indicates a postscript 1D color bit code. Since the additional recording 1D color bit code is composed of five cells, the decoding means determines that the 13th to 17th color information represents the color information of the additional recording 1D color bit code. Since the 17th to 18th color information is “B” and “W”, it may be confirmed that the 17th color information is the color information of the additional 1D color bit code.

  Thereafter, the decoding means determines in the same manner until reaching the color information at the end of the color arrangement information. As a result, in the decoding means, the serial 1D color bit code 112a arranges "original 1D color bit code", "additional 1D color bit code", "additional 1D color bit code", and "additional 1D color bit code" in this order. It can be determined that

  Similarly, for the serial 1D color bit codes 112b, 112c, and 112d, the decoding unit can determine the arrangement order of the original 1D color bit code and the additional 1D color bit code.

  As described above, according to the present embodiment, a plurality of 1D color bit codes are arranged in series in a straight line without providing a Qz color region for dividing the 1D color bit code between the 1D color bit codes. Is possible. As a result, even if a plurality of 1D color bit codes are marked on an article, they can be distinguished and recognized, so that, for example, management of operations including a plurality of processes can be easily performed.

(8-2) Data density improvement In the above (5-2), the Qz color (Qz color) is added to the color component colors (A color and B color) of the 1D color bit code, and a total of three colors are used. An example showing data is shown.

  As the number of constituent colors of the 1D color bit code increases, the amount of data that can be expressed increases and the data density can be increased.

  As an example in which the types of colors used for the marking of the 1D color bit code are more than the above three types, Japanese Patent Application No. 2007-130504 filed by the present inventor includes A color, B color, C color, and Qz. A 1D color bit code marked using four colors of color is shown. In this 1D color bit code, data is represented by three types of colors, A color, B color, and C color. The Qz color does not represent data.

  In this Japanese Patent Application No. 2007-130504, when the A color, B color, and C color are, for example, R (red), G (green), and B (blue), respectively, they also coincide with general image signal components. It is easy to handle in terms of image processing. Also, if there are three types of colors to be used, a wide color range per color can be obtained, which is practically convenient. If the number of types of colors is further increased, the color range per color is narrowed. Therefore, depending on the illumination light, camera adjustment, etc., contamination of each color is likely to occur, and the practicality may be lost.

  In the reading method in the present embodiment, as described in (5-1) above, since the portion to be read by the field frame is limited, there is a case where Qz may not be provided. In the case where it is not necessary to provide Qz, in addition to the colors constituting the data, the color reserved for Qz, the data can be represented by four types of colors of A color, B color, C color, and Qz color. . If four types of colors are used, the amount of data per cell can be further increased than when three types of colors are used.

  The upper row of the table 114 in FIG. 17 shows the color arrangement of the 1D color bit code. The color arrangement of this 1D color bit code is “W-A-B-W-A-W-A-C-B-B-C-C-B-C-B-C-B-C-W”. . In the table 114, this color arrangement is divided into a start, each color of the data portion, and an end. The start is “W-A-b”, the data part is “W-A-W-A-C-B-W-C-B-C-C-W”, and the end is “b-C-W”. . In the lower row of Table 114, the relationship between the color arrangement of the 1D color bit code and the data is shown.

  “B” in the start “W-A-b” is set so that the color of “A” immediately before and the color of the cell “W” behind “W-A-b” do not become the same color. This is a cell to be inserted. That is, the color of “b” changes depending on the colors before and after. In this embodiment, since “b” is sandwiched between “A” and “W”, it is a color that is neither A nor W (ie, B or C). Similarly, “b” in the end “b-C-W” is also inserted so that the color of the cell “C” in front of “b-C-W” does not become the same color. It is. In the present embodiment, since “b” is sandwiched between C, “b” is a color other than “C” (ie, W or A or B).

  In the present embodiment, the three colors A, B, and C are defined to represent numerical values “0”, “1”, and “2”, respectively. Qz (W) is defined to represent a numerical value represented by the color immediately after that.

  Based on these definitions, by decoding the color array “W-A-W-A-C-B-W-C-C-B-C-C”, the ternary data “000012122222” is obtained. can get. This is shown in the lower row of table 114 in FIG.

  Thus, according to the present embodiment, it is possible to obtain a beneficial effect that the data density can be further increased as compared with the conventional 1D color bit code.

(9) Effect FIG. 18A shows a 1D color bit code 116 marked on the surface of the object to be marked. The periphery of the 1D color bit code 116 (the portion indicated by the oblique lines in the figure) is Qz. If an unnecessary color area is added to a cell of a 1D color bit code, or if the arrangement of some constituent cells is disturbed, the conventional reading method may not be able to read accurately.

  In particular, as shown in FIG. 18 (2), when a 1D color bit code marked on the end face of the plate-like body 118 is captured, as shown in FIG. 18 (3), 1D color is added to the captured image. The background color pattern may be shown along with the bit code image. Since the background color pattern may interfere with reading of the 1D color bit code, it may be difficult to read with a conventional normal 1D color bit code reading means that does not use a field frame.

  However, according to the present embodiment, it is possible to exclude a portion in which the arrangement of cells is disturbed from the reading target. Therefore, even if the background color pattern is reflected in the captured image, the 1D color bit code can be easily read. It can be performed.

DESCRIPTION OF SYMBOLS 10 1D color bit code 12 Captured image 14a-14k Cell 16 Decoding means 18 Capture means 20 Captured object 22 1D color bit code 24 Image data 26 Field frame 28 Narrow image 30A-30C Color area 32 area 34a, 34b Processing result 36 Plate Shape 38 Captured image 40 1D color bit code 42 Field frame 44 Capture image 46 Plate-like body 48 Field frame 50 Narrow image 52 Object 54 White base 56 Component cell group 58 Capture image 60 Field frame 62 Narrow image 64 Plate body 66 White background 68a, 68b Redundant cell for cutting 70 1D color bit code 72 Capture image 74 Field frame 76 Capture screen 78 Field frame 80, 82, 84, 86 Capture screen 88 Capture image 90a-90e, 0k, 90k + 1, 90k + 2, 90n-1, 90n, 90q Color area 92 Frame 94k, 94k + 1, 94k + 2, 94q Center of gravity 96 Plate body 98 1D color bit code 98a End point cell 98b Cutout redundant cell group 98c Data cell group 100 Captured image 102 Original 1D color bit code 104 Additional 1D color bit code 106 Serial 1D color bit code 108 Original 1D color bit code 110 Additional 1D color bit code 112a to 112d Serial 1D color bit code 114 Table 116 1D color bit code 118 Plate A , B Field frame L (k), L (k + 2), L (q) Straight line α (k + 2), α (q) Angle

Claims (23)

In a method of reading a color array code representing data by color transition,
A capture step for capturing an image including the color arrangement code;
A narrow image extracting step of extracting a narrow image inside a long and narrow field frame preset in the image;
A division step of dividing the narrow image into a plurality of color regions according to the value of each pixel in the narrow image;
A color arrangement information generating step for generating color arrangement information describing the color of the color area and the order from one end to the other end of the narrow image for each color area;
Assuming that the color and order of the color area in the color array information are the order and color of the cell group constituting the color array code, the color array code is decoded based on the color and order, and the data is A decoding step to output;
A method for reading a color arrangement code, comprising: In a method of reading a color array code representing data by color transition,
A capture step for capturing an image including the color arrangement code;
A narrow image extracting step of extracting a narrow image inside a long and narrow field frame preset in the image;
A division step of dividing the narrow image into a plurality of color regions according to the value of each pixel in the narrow image;
A color arrangement information generating step for generating color arrangement information describing the color of the color area and the order from one end to the other end of the narrow image for each color area;
Among the color arrangement information, color information indicating the color area to which the color of the redundant cell for extraction indicating the start of the color arrangement code is attached, and the color of the redundant cell for extraction indicating the end of the color arrangement code are attached. The color arrangement representing the color area, and the color arrangement and the order of the color areas in the color arrangement information between them are considered to be the order and color of the cell group constituting the color arrangement code. A decoding step of decoding the color arrangement code based on the information and outputting the data;
A method for reading a color arrangement code, comprising: In the reading method of the color arrangement | sequence code of Claim 1 or 2,
The narrow image extracting step includes a first visual field frame moving step of moving the position of the visual field frame set in the image according to a user instruction within the image,
Further including
The method of reading a color arrangement code, wherein the narrow image extracting step extracts the narrow image by the field frame moved in the field frame moving step. In the reading method of the color arrangement | sequence code of Claim 1 or 2,
The narrow image extracting step includes a second field frame moving step of moving the preset field frame to a different position for each captured image,
Further including
When the narrow image inside the field frame set at the different position is composed of colors constituting the color array code, the color array code is extracted from the narrow image inside the field frame How to read. The method for reading a color arrangement code according to claim 4,
The method for reading a color arrangement code, wherein the field frame moves to a different position in a parallel direction for each image. The method for reading a color arrangement code according to claim 4,
The method of reading a color arrangement code, wherein the field frame moves to a different position in the direction of rotation about the center point of the field frame for each image. The method for reading a color arrangement code according to claim 4,
The method of reading a color arrangement code, wherein the field frame moves in a direction parallel to each image and rotates around a center point of the field frame. In the reading method of the color arrangement | sequence code of Claim 1 or 2,
The narrow image extracting step includes a third field frame moving step of scanning the image on the preset field frame,
Further including
A color array, wherein a narrow image inside the field frame is taken out when a narrow image inside the field frame scanned in the field frame moving step is composed of colors constituting the color array code. How to read the code. The method for reading a color arrangement code according to claim 8,
The method of reading a color arrangement code, wherein the field frame scans the image in parallel in a vertical direction or a horizontal direction. The method for reading a color arrangement code according to claim 8,
The method of reading a color arrangement code, wherein the field frame is scanned in a direction rotating around a center point of the field frame. The method for reading a color arrangement code according to claim 8,
The method of reading a color arrangement code, wherein the field frame scans in a direction parallel to each image and rotates about the center point of the field frame as an axis. In the reading method of the color arrangement | sequence code of any one of Claims 1-11,
The field frame has a shape having a predetermined width and length, wherein one of the longitudinal directions is the one end, the other is the other end, and the length is longer than the width. How to read color array code. The method for reading a color arrangement code according to claim 12,
One of the sides of the field frame is a curved line, and the field frame is arranged in a curved line. The method for reading a color arrangement code according to claim 12 or 13,
The color arrangement information generation step includes
Further, a length addition step for measuring the length of each narrow color image in the narrow image in the longitudinal direction of the narrow image, and generating the color arrangement information describing the measured length for each color region, and
When the length is smaller than a predetermined reference value, a data deletion step of deleting color information representing the color area from the color arrangement information;
A color arrangement information correction step for packing other adjacent color information before and after the deleted color information;
A method for reading a color arrangement code, comprising: In the reading method of the color arrangement | sequence code of any one of Claims 1-14,
In the color arrangement information, when there is color information of a color area in which the constituent colors constituting the color arrangement code are mixed, and the color of the area adjacent to the mixed color area is the constituent color, A color judging step that considers the mixed color to be a color of any color area adjacent to the color area;
A method for reading a color arrangement code, comprising: In the reading method of the color arrangement | sequence code of any one of Claims 1-15,
Between the narrow image extraction step and the color arrangement information generation step,
A uniformizing process step of replacing the value of each pixel in the narrow image with an average value of a pixel row including the pixel and extending in the width direction of the narrow image;
Including
The color arrangement information reading step is characterized in that the color arrangement information is generated based on the narrow image after the uniformization process is performed. In a method of reading a color array code representing data by color transition,
A capture step for capturing an image including the color arrangement code;
A division step of dividing the image into color regions based on the value of each pixel in the image;
A first start cell deeming step in which a color region including a pixel at a predetermined position in the image is regarded as a start cell of the color arrangement code;
A second cell deeming step of regarding a color area of a predetermined color among other color areas adjacent to the start cell as a second cell of the color arrangement code;
A straight line passing through the centroid point of the (k + 1) th cell and each centroid point of each of the other color regions adjacent to the (k + 1) th cell is drawn for each centroid point. The straight line having the smallest angle formed with respect to the straight line passing through the centroid point of the (k + 1) cell and the centroid point of the k-th cell is selected, and the (k + 1) -th centroid point among the two centroid points through which the selected straight line passes. ) A cell selection step that regards a color region to which a barycentric point other than the barycentric point of the cell belongs as the (k + 2) th cell of the color arrangement code;
For each color area considered to be a cell group from the start cell to the nth cell, the color of each color area and the order from the start cell to the nth cell are described. A color array information generation step for generating color array information;
Assuming that the color and order of the color region in the color array information is the color and order of the cell group constituting the color array code, the color array code is decoded based on the color and order, and the data is A decoding step to output;
A method for reading a color arrangement code, comprising: (K is a natural number of 1 or more, and n is a natural number greater than k.) In a method of reading a color array code representing data by color transition,
A capture step for capturing an image including the color arrangement code;
A division step of dividing the image into color regions based on the value of each pixel in the image;
A color area of a predetermined color is selected from a color area group in contact with a predetermined image edge in the image, and the selected color area is regarded as a start cell of the color arrangement code. No start cell deeming step,
A second cell deeming step of regarding a color area of a predetermined color among other color areas adjacent to the start cell as a second cell of the color arrangement code;
A straight line passing through the centroid point of the (k + 1) th cell and each centroid point of each of the other color regions adjacent to the (k + 1) th cell is drawn for each centroid point. The straight line having the smallest angle formed with respect to the straight line passing through the centroid point of the (k + 1) cell and the centroid point of the k-th cell is selected, and the (k + 1) -th centroid point among the two centroid points through which the selected straight line passes. ) A cell selection step that regards a color region to which a barycentric point other than the barycentric point of the cell belongs as the (k + 2) th cell of the color arrangement code;
For each color area considered to be a cell group from the start cell to the nth cell, the color of each color area and the order from the start cell to the nth cell are described. A color array information generation step for generating color array information;
Assuming that the color and order of the color region in the color array information is the color and order of the cell group constituting the color array code, the color array code is decoded based on the color and order, and the data is A decoding step to output;
A method for reading a color arrangement code, comprising: (K is a natural number of 1 or more, and n is a natural number greater than k.) The color array code reading method according to claim 17,
There are a plurality of the predetermined positions in the first start cell deemed step,
In the first start cell deeming step, when a color at any one of the plurality of predetermined positions is a predetermined color, the color area is a start cell of the color arrangement code. A method of reading a color arrangement code, characterized by The color array code reading method according to claim 18,
The predetermined color of the second start cell deeming step is a ground color of an article to which the color arrangement code is attached or a ground color in which the front surface is colored with respect to the ground color. How to read color array code. In a color arrangement code reader that represents data by color transition,
Capture means for capturing an image including the color arrangement code;
A narrow image extracting means for extracting a narrow image inside a long and narrow field frame preset in the image;
A dividing unit that divides the narrow image into a plurality of color regions according to a value of each pixel in the narrow image;
Color arrangement information generating means for generating color arrangement information describing the color of the color area and the order from one end to the other end of the narrow image for each color area;
Assuming that the color and order of the color region in the color array information is the color and order of the cell group constituting the color array code, the color array code is decoded based on the color and order, and the data is Decoding means for outputting;
A device for reading a color arrangement code, comprising: In a color arrangement code reader that represents data by color transition,
Capture means for capturing an image including the color arrangement code;
A narrow image extracting means for extracting a narrow image inside a long and narrow field frame preset in the image;
A dividing unit that divides the narrow image into a plurality of color regions according to a value of each pixel in the narrow image;
Color arrangement information generating means for generating color arrangement information describing the order from one end of the narrow image to the other end and the color of the color area for each color area;
Among the color arrangement information, color information indicating the color area to which the color of the redundant cell for extraction indicating the start of the color arrangement code is attached, and the color of the redundant cell for extraction indicating the end of the color arrangement code are attached. The color arrangement representing the color area, and the color arrangement and the order of the color areas in the color arrangement information between them are considered to be the color and order of the cell group constituting the color arrangement code, and the color arrangement Decoding means for decoding the color arrangement code based on the information and outputting the data;
A device for reading a color arrangement code, comprising: In the reading device of the color arrangement code according to claim 21 or 22,
Visual field frame moving means for moving the position of the visual field frame set in the image in accordance with a user's instruction within the image;
And the narrow image take-out means takes out the narrow image by the moved field frame.

Images