JP2008282339A - Code detection program, code detection method, and image forming apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a code detection program, a code detection method, and an image forming apparatus, capable of detecting simply and surely a one-dimensional code and a two-dimensional code from a printed matter, even when scanning diagonally the printed matter mixed with image information such as a photograph and text information. <P>SOLUTION: A reference line specifying part extracts an edge by executing one level of binomial wavelet conversion to a scan image, selects the edge with a prescribed value or more of high-frequency intensity, out of the extracted edges, Hough-transforms the selected edge to extract a straight line, and specifies a reference line, out of the extracted straight line, a candidate area specifying part specifies an area with the prescribed number or more of specified-directional straight lines with respect to the reference line, as a candidate area, and a code detecting part detects a code 31 by carrying out pattern-matching to the candidate area, or detects the code 31 by carrying out pattern-matching to the reference line. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a program and method for detecting a code such as a one-dimensional barcode or a two-dimensional code, and an image forming apparatus having a function for detecting the code.

  Many printed materials such as magazines and free paper in recent years are printed with a two-dimensional code such as QR code (registered trademark) in which URL (Uniform Resource Locator) information of stores is recorded. By reading the two-dimensional code using the photographing function and accessing the URL recorded in the two-dimensional code using the network connection function, it is possible to receive a service such as obtaining a discount ticket. However, for example, some mobile phones with simplified functions for easy operation and mobile phones for business use do not have a shooting function, and users using such mobile phones cannot use the above services. I can't receive it.

  On the other hand, since image forming apparatuses having a function of connecting to a communication network such as the Internet are widespread and installed in various places, a user who uses a mobile phone without a photographing function can enter a convenience store or the like. By scanning a free paper or the like with the installed image forming apparatus and extracting a two-dimensional code, the above service can be received. In this case, in order to reliably extract the two-dimensional code, it is necessary to perform appropriate image processing in the image forming apparatus. As an example of such image processing, for example, Patent Document 1 below extracts an image edge. In addition, a method is disclosed in which candidate lines are determined by searching for two straight lines that intersect with the Hough transform and least square approximation.

Japanese Patent Application Laid-Open No. 07-220081

  Generally, as a method of extracting a QR code from a scanned image, a method of searching for “cutout symbols” provided at corners of the QR code by pattern matching is used. However, the area where the QR code is printed is not constant. For this reason, pattern matching has to be performed on all areas of the scanned image, and there is a problem that it takes time for image processing.

  In addition, when the printed material is placed on the scanner and scanned, the printed material may be set obliquely, and the QR code is not necessarily parallel or perpendicular to the angle of view of the scanned image. In order to extract it reliably, pattern matching has to be performed at every angle, and there is a problem that it takes time for image processing.

  For such a problem, it is conceivable to use the image processing method disclosed in Patent Document 1, but image information such as photographs is often printed on printed matter such as magazines and free paper. Since there are a large number of edges, a huge amount of straight lines are extracted by the Hough transform. As a result, there is a problem that analysis cannot be performed and image processing takes a long time.

  The present invention has been made in view of the above problems, and the main object of the present invention is to provide a one-dimensional barcode from a printed material even when the printed material including image information such as a photograph and text information is scanned obliquely. Another object of the present invention is to provide a code detection program, a code detection method, and an image forming apparatus capable of easily and reliably detecting a two-dimensional code.

  In order to achieve the above object, the code detection program of the present invention extracts edges from a scan image obtained by scanning a printed material on which a computer is printed with at least a one-dimensional barcode or two-dimensional code. A straight line is extracted from the reference line, a reference line specifying unit that specifies the longest straight line among the extracted straight lines as a reference line, pattern matching is performed on the scanned image in a specific direction with respect to the reference line, and the one-dimensional It functions as a code detection unit that detects a barcode or the two-dimensional code.

  Further, the code detection program of the present invention extracts edges from a scanned image obtained by scanning a printed material on which at least a one-dimensional barcode or two-dimensional code is printed, and extracts a straight line from the extracted edges. A reference line specifying unit for specifying the longest straight line as a reference line among the extracted straight lines; an area including a predetermined number or more of straight lines in a specific direction with respect to the reference line among the divided areas of the scan image It functions as a candidate area specifying unit that specifies as a candidate area, and a code detection unit that performs pattern matching on the candidate area of the scanned image and detects the one-dimensional barcode or the two-dimensional code.

  In the present invention, the specific direction may be a direction substantially parallel and / or substantially perpendicular to the reference line.

  In the present invention, the reference line specifying unit performs one-level binomial wavelet transform on the scanned image to extract edges, and selects edges having a high frequency intensity equal to or higher than a predetermined value from the extracted edges. The selected edge is subjected to Hough transform to extract a straight line, and the straight line having the largest number of Hough transform votes is specified as the reference line from the extracted straight lines.

  The present invention also provides a code detection method for detecting a one-dimensional barcode or two-dimensional code printed on a printed material, the step of scanning the printed material to obtain a scanned image, and a straight line from the extracted edge. Extracting, identifying the longest straight line among the extracted straight lines as a reference line, identifying the longest straight line among the converted straight lines as a reference line, and a specific direction with respect to the reference line And performing pattern matching on the scanned image and extracting the one-dimensional barcode or the two-dimensional code.

  The present invention also provides a code detection method for detecting a one-dimensional barcode or two-dimensional code printed on a printed material, the step of scanning the printed material to obtain a scanned image, and an edge from the scanned image. Extracting, a step of extracting a straight line from the extracted edges, a step of identifying the longest straight line among the extracted straight lines as a reference line, a step of dividing the scan image into a plurality of regions, Specifying a region that includes a predetermined number or more of straight lines in a specific direction with respect to the reference line, as a candidate region, pattern matching is performed on the candidate region of the scanned image, and the one-dimensional bar Extracting at least a code or the two-dimensional code.

  In the present invention, in the step of extracting the edge, a one-level binomial wavelet transform is performed on the scanned image to extract an edge, and in the step of extracting the straight line, a high frequency intensity is predetermined from the extracted edge. In the step of selecting edges that are equal to or greater than the value, performing a Hough transform on the selected edges to extract a straight line, and identifying the reference line, the straight line having the largest number of votes for the Hough transform is selected from the extracted straight lines. It can be set as the structure specified as.

  The present invention is also an image forming apparatus including at least a scanner unit, and extracts edges from a scanned image obtained by scanning a printed material on which at least a one-dimensional barcode or two-dimensional code is printed. A straight line is extracted from the reference line, a reference line specifying unit that specifies the longest straight line among the extracted straight lines as a reference line, pattern matching is performed on the scanned image in a specific direction with respect to the reference line, and the 1 And a code detector for detecting the two-dimensional code.

  The present invention is also an image forming apparatus including at least a scanner unit, and extracts edges from a scanned image obtained by scanning a printed material on which at least a one-dimensional barcode or two-dimensional code is printed. A straight line is extracted from the reference line, a reference line specifying unit that specifies the longest straight line among the extracted straight lines as a reference line, and a straight line in a specific direction with respect to the reference line is predetermined among the divided areas of the scan image A candidate region specifying unit that specifies a region included in a number as a candidate region, a code detection unit that performs pattern matching on the candidate region of the scanned image, and detects the one-dimensional barcode or the two-dimensional code; Is provided.

  As described above, in the present invention, pattern matching is performed in a specific direction with respect to a reference line, or pattern matching is performed on a candidate region including a predetermined number or more of straight lines with a specific direction with respect to the reference line. Therefore, the one-dimensional barcode and the two-dimensional code can be easily and reliably detected from the printed material.

  According to the code detection program, the code detection method, and the image forming apparatus of the present invention, a one-dimensional barcode is printed from a printed material even when the printed material including image information such as a photograph and text information is scanned obliquely. And a two-dimensional code can be detected easily and reliably.

  The reason is that a scan image obtained by scanning a printed matter on which a code such as a one-dimensional barcode or a two-dimensional code is scanned performs binomial wavelet transform to extract an edge, and the high frequency intensity is extracted from the extracted edge. Edges that exceed a predetermined value are selected, a Hough transform is performed on the selected edges, a straight line is extracted, a reference line is specified from the extracted straight lines, and a specific direction (for example, a reference) is specified with respect to the reference line. Because the code is detected by performing pattern matching on the scanned image in parallel and / or at right angles to the line, the processing can be made easier compared to the conventional method in which pattern matching is performed in all directions. is there.

  Also, the binomial wavelet transform is performed on the scanned image to extract the edges, the edges having a high frequency intensity of a predetermined value or more are selected from the extracted edges, and the Hough transform is performed on the selected edges to extract a straight line. , Specifying a reference line from the extracted straight lines, specifying a region including a predetermined number or more of straight lines in a specific direction (for example, parallel and / or perpendicular to the reference line) as a candidate region, This is because the code is detected by performing pattern matching on the candidate area of the scan image, and therefore, the processing can be facilitated as compared with the conventional method in which pattern matching is performed on all areas.

  As shown in the prior art, image forming apparatuses are used for reading codes such as one-dimensional barcodes and two-dimensional codes printed on printed materials such as magazines and free paper. In addition, various information such as image information and text information is printed, and the printed material may be set obliquely on the scanner base, so the image forming apparatus scans the printed material. There is a problem in that it is necessary to perform pattern matching on the entire region from all directions on the image, and image processing takes time.

  On the other hand, printed matter such as magazines and free papers are usually printed with long straight lines such as border lines and dividing lines indicating boundaries, taking into account design and ease of viewing. Are printed parallel and / or perpendicular to the straight line. Accordingly, if appropriate image processing can be performed on the scan data to detect a frame line, a partition line, or the like, a code can be easily detected based on the line.

  Therefore, in the present embodiment, using binomial wavelet transform and Hough transform, which will be described later, an edge is extracted by performing binomial wavelet transform on the scanned image, and an edge whose high-frequency intensity is a predetermined value or more from the extracted edge The selected edge is subjected to Hough transform to extract a straight line, a reference line is specified from the extracted straight line, and pattern matching is performed on the scanned image in a specific direction with respect to the reference line. Or code detection, or by performing binomial wavelet transform on the scanned image and extracting edges, selecting edges with a high-frequency intensity greater than or equal to a predetermined value from the extracted edges, and huffing the selected edges Perform a conversion to extract a straight line, identify a reference line from the extracted straight lines, and select an area that contains a predetermined number or more of straight lines with a specific orientation relative to the reference line Specified as auxiliary region, employs a method of detecting the code performs pattern matching against a candidate region of the scanned image, to be able to detect easily and reliably code from the printed material. Hereinafter, the binomial wavelet transform and the Hough transform used for specifying the reference line will be outlined.

[Binary wavelet]
The wavelet transform is a wavelet function (formula (1)) that oscillates in a finite range as shown in FIG. 8, and the wavelet transform coefficients <f, ψa, b> for the input signal f (x) are expressed by formula (2). This is a transformation that decomposes the input signal into the sum of wavelet functions as shown in Equation (3).

  In the above formula, a represents the scale of the wavelet function, and b represents the position of the wavelet function. As illustrated in FIG. 8, the larger the value of the scale a, the smaller the frequency of the wavelet function ψa, b (x), and the position where the wavelet function ψa, b (x) oscillates moves according to the value of the position b. To do. Therefore, equation (3) means that the input signal f (x) is decomposed into the sum of wavelet functions ψa, b (x) having various scales and positions.

  There are many known wavelet functions that enable such transformations, but orthogonal wavelets and biorthogonal wavelets are widely used in the image processing field. . The wavelet function of the orthogonal wavelet / biorthogonal wavelet is defined as shown in Equation (4).

  Comparing equation (4) and equation (1), in the orthogonal wavelet and biorthogonal wavelet, the value of the scale a is discretely defined as 2 to the power of i, and the minimum moving unit of the position b is discretely 2i. You can see that it is defined. This value of i is called a level.

  On the other hand, a wavelet function of a binary wavelet (Dyadic Wavelet) used in the present invention is defined as shown in Expression (5) (for example, JP-A-9-212623).

  Here, the wavelet function of the orthogonal wavelet / biorthogonal wavelet is discretely defined with 2i as the minimum moving unit of the position at the level i as described above, whereas the binomial wavelet has the position regardless of the level i. The minimum movement unit is constant. Due to this difference, the binomial wavelet transform has the following characteristics.

Feature 1: The signal amount of each of the high frequency band component W _i and the low frequency band component S _i generated by the one-level binomial wavelet transform shown in Equation (6) is the same as the signal S _i-1 before the conversion. It is.

Feature 2: The relationship of Expression (7) is established between the scaling function φ _{i, j} (x) and the wavelet function ψ _{i, j} (x).

Therefore, the high frequency band component W _i generated by the binomial wavelet transform represents the first derivative (gradient) of the low frequency band component S _i .

Feature 3: With respect to W _i · γ _i (hereinafter referred to as a corrected high frequency band component) obtained by multiplying a high frequency band component by a coefficient γ _i determined according to the level i of the wavelet transform, the input signal Depending on the singularity of the signal change, the relation between the signal intensity levels of the corrected high frequency band components W _i · γ _i after the conversion follows a certain law. That is, the corrected high frequency band component W _i · γ _i corresponding to a gentle (differentiable) signal change corresponds to a step-like signal change while the signal intensity increases as the level number i increases. The corrected high frequency band component W _i · γ _i has a constant signal intensity regardless of the number of levels i, and the corrected high frequency band component W _i · γ _i corresponding to the δ function-like signal change has the level number i. The signal strength decreases as it increases.

Features 4: 1 level dyadic wavelet transform method in the two-dimensional signal such as an image signal is different from the method of orthogonal wavelet biorthogonal wavelets above by one level wavelet transform, the low frequency band component S _{n- 1} two high frequency band components _Wx n, is decomposed into Wy _n and one low frequency band component _{S n.} Two high-frequency band component corresponds to the x and y components of the change vector V _n in the two-dimensional low frequency band component S _n. Deflection angle _{A n} the size _{M n} of change vector _{V n} is given by equation (8), equation (9).

From the above characteristics, since the magnitude of the signal intensity of the corrected high frequency band component W _i · γ _i differs depending on the type of information to be processed, the scanner image is configured with a gentle curve. Image information can be excluded.

[Hough transform]
The Hough transform is a process of determining a straight line that passes through the most feature points of an image from straight lines that pass through a specific point. In the standard Hough transform, as shown in Expression (10), one straight line is obtained. Is represented by two parameters (the length and angle of the normal drawn on the straight line, r and θ).

  Here, if θ∈ [0, π] and r∈R, or θ∈ [0, 2π] and r ≧ 0, a set of (r, θ) for a certain straight line is uniquely determined. In other words, a straight line is expressed as being perpendicular to θ and at a distance r from the origin.

  Therefore, Hough transform is performed on the edges extracted by binomial wavelet transform, and data with a small number of votes is cut on the Hough transformed data, and then a specific edge is extracted as a straight line by performing inverse transform. By considering the data with the largest number of votes as the longest straight line, it is possible to identify frame lines, partition lines, etc. as reference lines from the edges extracted by the binomial wavelet transform. Thus, pattern matching can be performed quickly and easily by setting the direction of pattern matching or specifying candidate areas.

  In order to describe the above-described embodiment of the present invention in more detail, a code detection program, a code detection method, and an image forming apparatus according to an embodiment of the present invention will be described with reference to FIGS. FIG. 1 is a diagram schematically illustrating a configuration of an information providing system according to the present exemplary embodiment, FIG. 2 is a block diagram illustrating a configuration of an image forming apparatus according to the present exemplary embodiment, and FIG. 3 is a control unit of the image forming apparatus. It is a block diagram which shows the structure of these. FIG. 4 is a diagram illustrating a configuration example of a printed matter according to the present embodiment, and FIG. 5 is a flowchart illustrating an information providing method using the information providing system according to the present embodiment. FIG. 6 is a flowchart showing a procedure for specifying candidate areas in the code detection method of this embodiment, and FIG. 7 is a diagram showing a specific example thereof.

  As shown in FIG. 1A, the information providing system according to the present embodiment scans a printed material 30 on which various information is printed on a printing medium such as paper or plastic, such as a magazine or free paper, and prints 30. A copier or printer having a scanner function for detecting a one-dimensional barcode, two-dimensional code, color two-dimensional code, etc. (hereinafter collectively referred to as a code) printed on Collectively referred to as an image forming apparatus 10). Further, when a URL is recorded in the code and information is acquired by accessing the URL, as shown in FIG. 1B, the image forming apparatus 10 and a server 20 storing information to be provided to the user, Are connected by a communication network such as a WAN (Wide Area Network) or a LAN (Local Area Network).

  The one-dimensional bar code includes JAN, CODE39, ITF, NW-7, CODE128, CODE93, etc. The two-dimensional code is classified into a stack type two-dimensional code and a matrix type two-dimensional code. Two-dimensional codes include PDF417, CODE49, CODE16K, and matrix type two-dimensional codes include QR code, DATA MATRIX, VERI CODE, MAXI CODE, CP CODE, CODE1, box graphic code, ULTRA CODE, AZTECH CODE, etc. However, the code in this embodiment is not particularly limited as long as information is encoded and recorded. Further, the code of this embodiment includes a hologram for optically recording information.

  As shown in FIG. 2, the image forming apparatus 10 of this embodiment includes a control unit 11, an operation / display unit 12, a scanner unit 13, an image processing unit 14, a printer unit 15, and a storage unit 16. And a network connection unit 17 and the like, and each unit is connected by a bus.

  The control unit 11 includes a CPU (Central Processing Unit), a ROM (Read Only Memory), a RAM (Random Access Memory), and the like. The CPU reads various programs stored in the ROM by operating the operation / display unit 12 and develops them in the RAM, and controls the operation of each unit of the image forming apparatus 10 according to the developed programs.

  The operation / display unit 12 is provided with a pressure-sensitive operation unit (touch panel) in which transparent electrodes are arranged in a grid on a display unit such as an LCD (Liquid Crystal Display). In accordance with the signal, various operation buttons, device status display, operation status of each function, etc. are displayed on the display unit. In addition, the XY coordinates of the force point pressed with a finger or a touch pen are detected by a voltage value, and the detected position signal is output to the control unit 11 as an operation signal. Here, the operation / display unit 12 is configured such that the display unit and the operation unit are integrated, but an operation unit in which various operation buttons are arranged may be provided separately from the display unit.

  The scanner unit 13 is disposed below the glass on which the printed material 30 is placed, and reads information recorded on the printed material 30. The light source that scans the printed material 30 and the light reflected by the printed material 30 are converted into electrical signals. It comprises a CCD (Charge Coupled Devices) for conversion, an A / D converter for A / D converting electrical signals, and the like.

  The image processing unit 14 performs enlargement / reduction, rotation, frequency conversion, color conversion from RGB data to YMCK data, gradation correction, and the like on an image read by the scanner unit 13 (hereinafter referred to as a scan image). Various image processes are performed and output to the printer unit 15.

  The printer unit 15 includes an image forming unit that forms an image to be printed on a transfer paper, a transfer unit that transfers the formed image to the transfer paper, a fixing unit that fixes the transferred image, a transport unit that transports the transfer paper, and an image forming unit. And a transfer unit, a cleaning unit for cleaning the fixing unit, and the like, and by electrophotography, an image is formed on a transfer sheet and output based on the input data.

  The storage unit 16 is configured by a flash memory, a hard disk, or the like, and stores various data, setting conditions, and the like.

  The network connection unit 17 includes a NIC (Network Interface Card), a modem, a LAN adapter, a router, a TA (Terminal Adapter), and the like, and performs communication control with the server 20 and the like connected via a communication network.

  FIG. 1 shows an example of the image forming apparatus 10 according to the present embodiment. The image forming apparatus 10 only needs to have a function of detecting at least a printed matter 30 and detecting a code. A multi function peripheral (MFP) including a device (ADF: Auto Document Feeder), a post-processing device (finisher), and the like may be used.

  Here, as shown in FIG. 4A, a printed matter 30 such as a magazine or a free paper (a single sheet or a plurality of sheets may be bound in a booklet form) includes a photograph, an illustration, a figure, and the like. Image information 32 and text information 33 such as hiragana, katakana, kanji, alphabets, symbols, and the like, and specific marks (cutouts arranged in the upper left, upper right, and lower left in the case of QR codes) In the case of the symbol 31a and DATA MATRIX, there are L-shaped guide cells arranged on the left and below, which will be described as a cut-out symbol 31a), and URL information and the like are encoded in an area specified by the cut-out symbol 31a. A code 31 on which the recorded pattern is arranged is printed, and further, a frame line 34a is printed around these in consideration of the design and visibility. Further, as shown in FIG. 4B, a partition line 34b is printed instead of the frame line 34a or together with the frame line 34a.

  Therefore, it is not easy to detect the code 31 from the printed material 30 on which these various types of information are printed, and the arrangement of the code 31 is not constant. Pattern matching must be performed for all areas of the scanned image. In particular, when the printed material 30 is set obliquely on the scanner unit 13, the code 31 is arranged obliquely with respect to the angle of view of the scanned image. Therefore, pattern matching must be performed from various angles. There was a problem that it took time to detect 31.

  Therefore, in the present embodiment, the frame line 34a and the partition line 34b are provided so that the code 31 can be easily and reliably detected even when the printed matter 30 on which various information is printed is scanned obliquely. Is specified as a reference line, and pattern matching is performed only in a specific direction such as parallel and / or perpendicular to the reference line, or a specific orientation such as parallel and / or perpendicular to the reference line A region including a predetermined number or more of straight lines is specified as a candidate region, and pattern matching is performed only on the candidate region.

  Specifically, as shown in FIG. 3, the control unit 11 uses the above-described binomial wavelet transform and Hough transform to identify a reference line from the scan image, and a plurality of scan images. Candidate regions that are divided into regions and include a predetermined number or more of straight lines in a specific direction such as substantially parallel or substantially perpendicular to the reference line (hereinafter simply referred to as parallel and / or right angle). Pattern matching is performed on the candidate area specifying unit 11b specified as the area and the candidate area of the scan image, or pattern matching is performed on the scan image in a specific direction such as parallel and / or perpendicular to the reference line. And a code detection unit 11c that detects the cut-out symbol 31a and reads information recorded in the code 31.

  The reference line specifying unit 11a, the candidate area specifying unit 11b, and the code detecting unit 11c may be configured as hardware, or the computer may include a reference line specifying unit 11a, the candidate area specifying unit 11b, and the code detecting unit 11c, Alternatively, it may be configured as a code detection program that functions as the reference line specifying unit 11a and the code detection unit 11c, and the code detection program is operated on the control unit 11.

  Next, a procedure from reading the printed matter 30 to printing a discount ticket or the like using the image forming apparatus 10 having the above configuration will be described with reference to the flowcharts of FIGS. 5 and 6 and the schematic diagram of FIG. .

  First, in step S101 in FIG. 5, when the printed material 30 on which the code 31 is printed is set in the image forming apparatus 10 and the operation / display unit 12 is operated to instruct scanning, the scanner unit 13 causes the printed material 30 to be scanned with a light source. The light that is scanned and reflected by the printed material 30 is received by the CCD and converted into an electrical signal, and the electrical signal is A / D converted to create a scanned image in which information of the printed material 30 is recorded. This scan may be either monochrome or color, but the frame line 34a and partition line 34b are usually printed in black, and the processing of the scanned image is simpler in monochrome, so the code 31 is color. If it is not a two-dimensional code, it is preferable to scan in monochrome.

  Next, in step S102, the control unit 11 detects a candidate area where it is determined that the code 31 is printed. This candidate area detection procedure will be described in detail with reference to FIGS.

  First, in step S <b> 102 a, the control unit 11 acquires a scanned image from the scanner unit 13. The scanned image is not particularly limited as long as it is an image representing brightness, such as luminance, density, and brightness data, and may be an image obtained by the scanner unit 13 or an image processed by the image processing unit 14.

Next, in step S102b, the reference line specifying unit 11a performs a one-level binomial wavelet transform on the scanned image to obtain a horizontal high-frequency component (Wx) and a vertical high-frequency component (Wy). the formula (8), according to (9), a high frequency intensity (i.e., magnitude _M n of the change vector _{V n),} the edge angle (i.e., deflection angle _{a n)} for calculating a.

  More specifically, FIG. 7A schematically shows the configuration of a scan image, and when binomial wavelet transform is performed on this scan image, an edge is detected and shown in FIG. 7B. Such an edge image is generated. At that time, the frame line 34a, the partition line 34b, and the code 31 hardly change, but the text information 33 becomes a straight line group with various inclinations, and the image information 32 becomes a gentle curve.

  Next, in step S102c, the reference line specifying unit 11a selects an edge whose high-frequency intensity (that is, density fluctuation) is equal to or greater than a predetermined value from edges obtained by the binomial wavelet transform. Here, the image information is usually a gentle curve, and the high-frequency intensity of the gentle curve is small in the one-level binomial wavelet transform due to the characteristics of the above-described binomial wavelet transform. As described above, the image information 32 is excluded from the scanned image.

  Next, in step S102d, the reference line specifying unit 11a performs Hough transform on an edge having a specific intensity or more. In this Hough transform, when an edge with a small number of votes is cut and then a reverse line is extracted to extract a straight line, the frame line 34a and the partition line 34b are extracted as long straight lines as shown in FIG. Among the edges of 33 and the code 31, an edge having a vote count equal to or more than a predetermined value is extracted as a short straight line. Then, by selecting the longest straight line (that is, the straight line with the largest number of votes) from these straight lines, the frame line 34 a and the partition line 34 b are specified as the reference line 35.

  Next, in step S102e, the candidate area specifying unit 11b selects a specific direction (here, the reference line 35) with respect to the reference line 35 specified in the above step from the plurality of straight lines extracted by the Hough transform. Only straight lines (parallel and / or perpendicular) are extracted. By this process, the oblique line in the text information 33 is excluded and the scanned image becomes as shown in FIG.

  Next, in step S102f, the candidate area specifying unit 11b divides the scanned image into a plurality of areas by vertical and horizontal lines, and in step S102g, selects an area having a predetermined number of edges or more from the divided areas. Then, the area is specified as a candidate area. Specifically, as shown in FIG. 7F, among the areas divided by the vertical and horizontal lines, the area indicated by shading is a candidate that includes a predetermined number or more of straight lines parallel to and / or perpendicular to the reference line 35. It becomes an area. Since the candidate area is a set of several areas, the cut-out symbol 31a is searched for at the boundary between the set and the non-candidate area.

  Next, returning to FIG. 5, in step S <b> 103, the control unit 11 determines whether a candidate area is detected from the scanned image. If there is no candidate area, the user has printed the code 31 on the printed material 30. Then, it is checked whether or not the printed material 30 is properly set in the scanner unit 13. When rescan is instructed in step S104, the process returns to step S102 and the same processing is repeated.

  On the other hand, if there is a candidate region, in step S105, the code detection unit 31c of the control unit 11 performs pattern matching in the vicinity of the candidate region of the scan image using a known method, and sets the preset figure and the cutout symbol 31a. The cut-out symbol 31a is detected by comparing. In the case of DATA MATRIX, an L-type guide cell may be detected, and in the case of a code using another mark, the mark may be detected.

  In the flowcharts of FIGS. 5 and 6, the candidate area is specified and pattern matching is performed on the candidate area of the scanned image. However, after the reference line 35 is specified in step S102d of FIG. A pattern matching may be performed on the scanned image in a specific direction (for example, parallel and / or perpendicular to the reference line 35) with respect to the line 35, and by performing pattern matching only in a specific direction, The code 31 can be detected easily and quickly.

  In step S106, the code detection unit 31c determines the pattern matching result. If the cutout symbol 31a cannot be detected by pattern matching, the user has printed the code 31 on the printed material 30 or the printed material 30 is the scanner unit. 13 is checked and if rescan is instructed in step S104, the process returns to step S102 and the same processing is repeated.

  On the other hand, if the cutout symbol 31a can be detected by pattern matching, the code detection unit 31c reads the information recorded in the code 31 in step S107. If the URL information is recorded in the code 31, in step S108, the control unit 11 connects to the communication network using the network connection unit 17, accesses the server 20 specified by the URL information, and Information such as discount tickets is acquired from the server 20. If information such as a discount ticket is recorded in the code 31, the process of step S108 can be omitted.

  In step S109, the control unit 11 waits for the user's instruction, and if an instruction to end without printing a discount ticket or the like is given (for example, if the user does not desire a discount ticket or the like), the series of processing ends. If an instruction to print a discount ticket or the like is given, the information recorded in the code 31 or the information acquired from the server 20 is transferred to the printer unit 15 and the printer unit 15 prints the discount ticket or the like in step S110. To finish the process.

  As described above, the control unit 11 of the image forming apparatus 10 is provided with the reference line specifying unit 11a, the candidate area specifying unit 11b, and the code detecting unit 11c as hardware or software, and obtained by scanning the printed material 30 with the scanner unit 13. A reference line 35 is specified by performing binomial wavelet transform and Hough transform on the scanned image, and an area including a predetermined number or more of straight lines in a specific direction such as parallel and / or perpendicular to the reference line 35 Compared to the conventional method in which pattern matching is performed on all regions at various angles, because the pattern 31 is identified as a candidate region and pattern matching is performed on the candidate region of the scanned image to detect the code 31, the time required for pattern matching And the code 31 can be detected easily and reliably.

  Further, the control unit 11 of the image forming apparatus 10 is provided with a reference line specifying unit 11a and a code detecting unit 11c as hardware or software, and two items are obtained with respect to a scanned image obtained by scanning the printed matter 30 with the scanner unit 13. By detecting the code 31 by performing wavelet transform and Hough transform to identify the reference line 35 and performing pattern matching on the scanned image in a specific direction such as parallel and / or perpendicular to the reference line 35 However, the time required for pattern matching can be significantly shortened, and the code 31 can be detected easily and reliably.

  In the present embodiment, the configuration in which the printed material 30 includes the frame 34a and the partition line 34b is shown. However, even when the printed material 30 is not printed, when the printed material 30 is scanned by the scanner unit 13, Since the edge can be extracted, the edge of the sheet can be used as the reference line 35.

  In the above-described embodiment, the case where the code 31 printed on the printed material 30 is detected has been described. However, the present invention is not limited to the above-described embodiment, and the scan includes image information 32, text information 33, and the like. The present invention can be similarly applied to the case of detecting a figure or the like containing many components parallel and / or perpendicular to the reference line 35 from the image.

  The present invention is applicable to a program, a method, and an image forming apparatus for detecting a code such as a one-dimensional barcode or a two-dimensional code from a printed material.

It is a figure which shows typically the structure of the information provision system which concerns on one Example of this invention. 1 is a block diagram illustrating a configuration of an image forming apparatus according to an embodiment of the present invention. FIG. 3 is a block diagram illustrating a configuration of a control unit of the image forming apparatus according to an embodiment of the present invention. It is a figure which shows the structural example of the printed matter which concerns on one Example of this invention. It is a flowchart figure which shows the information provision method using the information provision system which concerns on one Example of this invention. It is a flowchart figure which shows the identification procedure of the candidate area | region in the code | cord | chord detection method which concerns on one Example of this invention. It is a figure which shows the specific example of the identification procedure of the candidate area | region in the code | cord | chord detection method which concerns on one Example of this invention. It is a figure which shows a wavelet transformation function.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Image forming apparatus 11 Control part 11a Base line specific | specification part 11b Candidate area | region specific part 11c Code detection part 12 Operation / display part 13 Scanner part 14 Image processing part 15 Printer part 16 Storage part 17 Network connection part 20 Server 30 Printed matter 31 Code 31a Cutout symbol 32 Image information 33 Text information 34a Frame line 34b Partition line 35 Reference line

Claims (12)

Computer
Edges are extracted from the scanned image obtained by scanning a printed material on which at least one-dimensional barcode or two-dimensional code is printed, straight lines are extracted from the extracted edges, and the longest straight line among the extracted straight lines is the reference line. Reference line identification part, identified as
A code detection program that performs pattern matching on the scanned image in a specific direction with respect to the reference line and functions as a code detection unit that detects the one-dimensional barcode or the two-dimensional code. Computer
Edges are extracted from the scanned image obtained by scanning a printed material on which at least one-dimensional barcode or two-dimensional code is printed, straight lines are extracted from the extracted edges, and the longest straight line among the extracted straight lines is the reference line. Reference line identification part, identified as
A candidate area specifying unit that specifies, as a candidate area, an area that includes a predetermined number or more of straight lines in a specific direction with respect to the reference line among the areas obtained by dividing the scan image;
A code detection program that performs pattern matching on the candidate area of the scanned image and functions as a code detection unit that detects the one-dimensional barcode or the two-dimensional code.   The code detection program according to claim 1 or 2, wherein the specific direction is a direction substantially parallel and / or substantially perpendicular to the reference line.   The reference line specifying unit performs one-level binomial wavelet transform on the scanned image to extract edges, selects an edge having a high frequency intensity equal to or higher than a predetermined value from the extracted edges, and converts the selected edge to a Hough transform The code detection program according to any one of claims 1 to 3, wherein a straight line having the largest number of Hough transform votes is specified as the reference line from the extracted straight lines. . A code detection method for detecting a one-dimensional barcode or two-dimensional code printed on a printed material,
Scanning the printed matter to obtain a scanned image;
Extracting an edge from the scanned image;
Extracting a straight line from the extracted edges;
A step of identifying the longest straight line among the extracted straight lines as a reference line;
A code detection method comprising: performing pattern matching on the scanned image in a specific direction with respect to the reference line, and extracting the one-dimensional barcode or the two-dimensional code. A code detection method for detecting a one-dimensional barcode or two-dimensional code printed on a printed material,
Scanning the printed matter to obtain a scanned image;
Extracting an edge from the scanned image;
Extracting a straight line from the extracted edges;
Identifying the longest straight line of the extracted straight lines as a reference line;
Dividing the scanned image into a plurality of regions;
Identifying a region including a predetermined number or more of straight lines in a specific direction with respect to the reference line among the plurality of regions as a candidate region;
A coded information detection method comprising: performing pattern matching on the candidate area of the scanned image and extracting the one-dimensional barcode or the two-dimensional code.   The code detection method according to claim 5 or 6, wherein the specific direction is a direction substantially parallel and / or substantially perpendicular to the reference line. In the step of extracting the edge, an edge is extracted by performing a one-level binomial wavelet transform on the scanned image;
In the step of extracting the straight line, an edge having a high frequency intensity equal to or higher than a predetermined value is selected from the extracted edges, a straight line is extracted by performing a Hough transform on the selected edge,
The code detection according to any one of claims 5 to 7, wherein, in the step of specifying the reference line, a straight line having the largest number of Hough transform votes is specified as the reference line from the extracted straight lines. Method. An image forming apparatus including at least a scanner unit,
Edges are extracted from the scanned image obtained by scanning a printed material on which at least one-dimensional barcode or two-dimensional code is printed, straight lines are extracted from the extracted edges, and the longest straight line among the extracted straight lines is the reference line. A reference line specifying part to be specified as
An image forming apparatus comprising: a code detection unit configured to perform pattern matching on the scanned image in a specific direction with respect to the reference line and detect the one-dimensional barcode or the two-dimensional code. An image forming apparatus including at least a scanner unit,
Edges are extracted from the scanned image obtained by scanning a printed material on which at least one-dimensional barcode or two-dimensional code is printed, straight lines are extracted from the extracted edges, and the longest straight line among the extracted straight lines is the reference line. A reference line specifying part to be specified as
A candidate area specifying unit for specifying, as a candidate area, an area that includes a predetermined number or more of straight lines in a specific direction with respect to the reference line among the areas obtained by dividing the scan image;
An image forming apparatus comprising: a code detection unit configured to perform pattern matching on the candidate area of the scan image and detect the one-dimensional barcode or the two-dimensional code.   The image forming apparatus according to claim 9, wherein the specific direction is a direction substantially parallel and / or substantially perpendicular to the reference line.   The reference line specifying unit performs one-level binomial wavelet transform on the scanned image to extract edges, selects an edge having a high frequency intensity equal to or higher than a predetermined value from the extracted edges, and converts the selected edge to a Hough transform 12. The image forming apparatus according to claim 9, wherein a straight line having the largest number of votes for Hough transform is identified as the reference line from the extracted straight lines. .

Images