JP6268888B2 - Information processing apparatus and information processing program - Google Patents

Description

  The present invention relates to an information processing apparatus and an information processing program.

  Patent Document 1 uses a handwriting input system as an object to accurately determine a correction process for an input character on the surface of an entry sheet on which a character is written. The handwriting input system uses a dot pattern on a form. A digital pen that generates handwriting information by reading a handwriting, a handwriting image generation unit that generates a handwriting image obtained by converting the handwriting of the digital pen into image data from the handwriting information, and a character recognition unit that performs character recognition by inputting the handwriting image Based on the detection result of the correction instruction detection unit that detects from the handwriting image the cancellation character specification line that specifies the character to be deleted from the character recognition result and the deletion character number specification line that specifies the number of characters to be deleted. And a character recognition result correction unit that corrects a character recognition result by the character recognition unit.

  In Patent Document 2, in a writing means using a tablet or an electronic pen, changing work such as deletion, addition, substitution, and emphasis of written characters is performed as much as possible with a normal writing action using ordinary paper and a pen. The task is to provide a means that can be performed on a computer, and a writer uses a double line or fill that means deletion, a mountain or valley symbol that means addition, an arrow symbol, or emphasis that is used in normal writing. Corresponding symbols such as encircled lines and underlines are associated with the start of the modification process and control commands for specifying the target character, and these commands are automatically detected from the written information and the process is automatically performed. When a control symbol such as a double line or an enclosing line is entered using the character cutout information generated in the character recognition method for the detection of the character to be changed at that time The calculated degree of overlap of the character segmentation information and their control symbols, by utilizing the determination of the change target characters, it is disclosed that to realize highly accurate detection.

  Patent Document 3 aims to prevent leakage of writing information after performing processing based on writing information on a medium, writing on a printed document with an electronic pen, and connecting to a communication device. Electronic document indicating stroke information obtained by digitizing writing to the terminal device, the terminal device authenticating the electronic pen using the pen authentication server, and inquiring the identification information server to associate the stroke information with the electronic document Information is obtained, and registration is requested by transmitting electronic document information and stroke information to the document server, whereby the document server transmits result information indicating success / failure of registration to the terminal device, Upon receiving the result information, the communication device deletes the stroke information in the electronic pen, and outputs a message prompting re-input if the result information indicates registration failure. Door has been disclosed.

JP 2008-040759 A JP 2004-152040 A JP 2008-077753 A

  An object of the present invention is to provide an information processing apparatus and an information processing program that determine a deleted character from a stroke group that may include a deletion symbol.

The gist of the present invention for achieving the object lies in the inventions of the following items.
The invention of claim 1 is a receiving means for receiving a stroke group, a stroke extracting means for extracting a stroke group for each unit divided according to a predetermined rule from the stroke group received by the receiving means, and for each unit. There is a rectangle extracting means for extracting a rectangle surrounding the stroke group, an area dividing means for dividing the rectangle extracted by the rectangle extracting means into regions, and the stroke exists for each stroke extracted by the stroke extracting means. Counting means for counting the number of areas divided by the area dividing means, and determination for determining whether or not the stroke group in the rectangle has been deleted based on the number of areas counted by the counting means. An information processing apparatus comprising: means.

  The invention of claim 2 further comprises a calculation means for calculating an intersection point between the stroke and the boundary line of the region, and the counting means has a stroke in a region in contact with the intersection point calculated by the calculation means. The information processing apparatus according to claim 1, wherein the information processing apparatus is an area.

  The invention of claim 3 is characterized in that the rectangle extracting means extracts the rectangle with the height of the row of units as the height of the rectangle, or with the width of the row of units as the width of the rectangle. An information processing apparatus according to claim 1 or 2.

The invention of claim 4 is a receiving means for receiving a stroke group, a stroke extracting means for extracting a stroke group for each unit divided according to a predetermined rule from the stroke group received by the receiving means, and for each unit. There is a rectangle extracting means for extracting a rectangle surrounding the stroke group, an area dividing means for dividing the rectangle extracted by the rectangle extracting means into regions, and the stroke exists for each stroke extracted by the stroke extracting means. Counting means for counting the number of areas divided by the area dividing means, measuring means for measuring the length of each stroke extracted by the stroke extracting means, and stroke measured by the measuring means The length of either the height or width of the rectangle containing the stroke The stroke group in the rectangle is deleted based on the normalizing means that normalizes using one or both, the number of areas counted by the counting means and the length of the stroke normalized by the normalizing means It is an information processing apparatus characterized by comprising a judging means for judging whether or not it has been done.

The invention of claim 5 includes a computer that receives a stroke group, a stroke extracting unit that extracts a stroke group for each unit divided according to a predetermined rule from the stroke group received by the receiving unit, Rectangle extraction means for extracting a rectangle surrounding the stroke group for each unit, area dividing means for dividing the rectangle extracted by the rectangle extraction means into areas, and for each stroke extracted by the stroke extraction means, the stroke Counting means for counting the number of areas divided by the area dividing means, and whether or not the stroke group in the rectangle is deleted based on the number of areas counted by the counting means. It is an information processing program for functioning as a determination means for determining.

The invention of claim 6 includes a computer that receives a stroke group, a stroke extracting unit that extracts a stroke group for each unit divided according to a predetermined rule from the stroke group received by the receiving unit, Rectangle extraction means for extracting a rectangle surrounding the stroke group for each unit, area dividing means for dividing the rectangle extracted by the rectangle extraction means into areas, and for each stroke extracted by the stroke extraction means, the stroke Counting means for counting the number of areas divided by the area dividing means, measuring means for measuring the length of each stroke extracted by the stroke extracting means, and measuring by the measuring means The length of the stroke made is the height of the rectangle that contains the stroke. Normalizing means using one or both of the widths, and the number of areas counted by the counting means and the length of the stroke normalized by the normalizing means. The stroke group is an information processing program for functioning as a determination means for determining whether or not the stroke group has been deleted.

  According to the information processing apparatus of the first aspect, it is possible to determine the deleted stroke group from the stroke group that may include the deletion symbol.

  According to the information processing apparatus of the second aspect, it is possible to determine a region in contact with the intersection of the stroke and the boundary line of the region as a region where the stroke exists.

  According to the information processing apparatus of the third aspect, the rectangle can be extracted using the height of the unit column or the width of the unit column.

According to the information processing apparatus of the fourth aspect , the deleted stroke group is determined from the stroke group that may include the deletion symbol with higher accuracy than in the case where the present configuration is not provided. be able to.

According to the information processing program of the fifth aspect , it is possible to determine the deleted stroke group from the stroke group that may include the deletion symbol.

According to the information processing program of the sixth aspect , the deleted stroke group is determined from the stroke group that may include the deletion symbol with higher accuracy than in the case of not having this configuration. be able to.

It is a conceptual module block diagram about the structural example of 1st Embodiment. It is a flowchart which shows the process example by 1st Embodiment. It is a flowchart which shows the process example by 1st Embodiment. It is explanatory drawing which shows the data structure example of stroke information. It is explanatory drawing which shows the process example by 1st Embodiment. It is explanatory drawing which shows the process example by 1st Embodiment. It is explanatory drawing which shows the process example by 2nd Embodiment. It is a conceptual module block diagram about the structural example of 3rd Embodiment. In prior art, it is explanatory drawing which shows the example of correction of a character string. In prior art, it is explanatory drawing which shows the example of correction of a character string. In prior art, it is explanatory drawing which shows the example of correction of a character string. It is explanatory drawing which shows the example of a system in the case of implement | achieving this Embodiment. It is explanatory drawing which shows the example of the electronic pen paper on which the information image was printed. It is explanatory drawing which shows the structural example in an electronic pen. It is a flowchart which shows the process example by an electronic pen. It is explanatory drawing which shows the example of the information image (code pattern image) handled by this Embodiment. It is explanatory drawing which shows the example of the encoding process of the information in this Embodiment, and the production | generation process example of an information image (dot code image). It is a block diagram which shows the hardware structural example of the computer which implement | achieves this Embodiment.

First, before explaining the present embodiment, the presupposed technology will be described with reference to FIGS. This description is intended to facilitate understanding of the present embodiment.
In an apparatus that performs character recognition using an electronic pen, there are cases where it is desired to delete (erase) a character once entered. As shown in the example of FIG. 9, after writing the character string “ABC”, there is a case where it is actually desired to correct the character string “AC”.
When using electronic information on a computer such as a word processor, the character “B” can be selected and deleted.
When the electronic pen is used, “ABC” is already written on the paper with ink. It is desirable to delete on the notation in ink, and the deletion on the notation is directly interpreted as deletion of characters as electronic data.

The following example shows an example of correcting the character string “ABC” to the character string “AC” in the online character recognition using the electronic pen (an example of deleting “C” of the character string “ABC”).
FIG. 10 is an explanatory diagram showing a modification example of a character string in the technique described in Patent Document 1. In FIG. As shown in the example of FIG. 10, a deletion character is specified using a character string strikethrough composed of a horizontal line specifying the deletion character string and a diagonal line specifying the deletion character.
FIG. 11 is an explanatory diagram showing a modification example of a character string in the technique described in Patent Document 2. As shown in the example of FIG. 11, when multiple lines (including a single line) and fill are described, an example of handling as deletion is shown.
When deleting with horizontal lines and diagonal lines as in Patent Document 1, there may be characters of the same shape, so especially when only one character is deleted, the deletion symbol and the character are confused. There is a possibility.
When using multiple lines or fills as in Patent Document 2, it is possible to distinguish between characters and deletion symbols.
However, in Patent Document 2, it is necessary to register “in the storage unit with a combination of the shape of each control symbol and the corresponding operation”. That is, the operator needs to write the same deletion symbol as that determined in advance.

Hereinafter, examples of various preferred embodiments for realizing the present invention will be described with reference to the drawings.
FIG. 1 is a conceptual module configuration diagram of a configuration example according to the first embodiment.
The module generally refers to components such as software (computer program) and hardware that can be logically separated. Therefore, the module in the present embodiment indicates not only a module in a computer program but also a module in a hardware configuration. Therefore, the present embodiment is a computer program for causing these modules to function (a program for causing a computer to execute each procedure, a program for causing a computer to function as each means, and a function for each computer. This also serves as an explanation of the program and system and method for realizing the above. However, for the sake of explanation, the words “store”, “store”, and equivalents thereof are used. However, when the embodiment is a computer program, these words are stored in a storage device or stored in memory. It is the control to be stored in the device. Modules may correspond to functions one-to-one, but in mounting, one module may be configured by one program, or a plurality of modules may be configured by one program, and conversely, one module May be composed of a plurality of programs. The plurality of modules may be executed by one computer, or one module may be executed by a plurality of computers in a distributed or parallel environment. Note that one module may include other modules. Hereinafter, “connection” is used not only for physical connection but also for logical connection (data exchange, instruction, reference relationship between data, etc.). “Predetermined” means that the process is determined before the target process, and not only before the process according to this embodiment starts but also after the process according to this embodiment starts. In addition, if it is before the target processing, it is used in accordance with the situation / state at that time or with the intention to be decided according to the situation / state up to that point. When there are a plurality of “predetermined values”, they may be different values, or two or more values (of course, including all values) may be the same. In addition, the description having the meaning of “do B when it is A” is used in the meaning of “determine whether or not it is A and do B when it is judged as A”. However, the case where it is not necessary to determine whether or not A is excluded.
In addition, the system or device is configured by connecting a plurality of computers, hardware, devices, and the like by communication means such as a network (including one-to-one correspondence communication connection), etc., and one computer, hardware, device. The case where it implement | achieves by etc. is included. “Apparatus” and “system” are used as synonymous terms. Of course, the “system” does not include a social “mechanism” (social system) that is an artificial arrangement.
In addition, when performing a plurality of processes in each module or in each module, the target information is read from the storage device for each process, and the processing result is written to the storage device after performing the processing. is there. Therefore, description of reading from the storage device before processing and writing to the storage device after processing may be omitted. Here, the storage device may include a hard disk, a RAM (Random Access Memory), an external storage medium, a storage device via a communication line, a register in a CPU (Central Processing Unit), and the like.

  The information processing apparatus according to the present embodiment determines a deleted stroke group from a stroke group that may include a deletion symbol. As shown in the example of FIG. It has a recognition module 110, a stroke scratch determination module 120, and a character removal module 130. Note that characters are mainly shown as examples as “units delimited according to a predetermined rule”. The rectangle extraction performed by the information processing apparatus is often performed for each character. However, the character is not accurately determined. In addition, since the boundary between characters may be unknown, the stroke group may not be extracted “by character”. The rectangle described here has a meaning of “a circumscribed rectangle of a stroke group extracted as a unit like a character”. The “predetermined rule” refers to, for example, a rule for extracting a stroke group as a character by the online character recognition module 110, and is specifically described within a predetermined size range. That it is performed, that it is described within a predetermined time, that there is an interval more than a predetermined time between the preceding and following strokes, or a combination thereof. The “stroke group” is a set of strokes composed of one or more strokes. Therefore, the stroke group may be composed of one stroke.

The online character recognition module 110 is connected to the stroke scratch determination module 120. The online character recognition module 110 receives the stroke information 105. The stroke information 105 (group) received by the online character recognition module 110 includes a plurality of strokes, and may include a stroke indicating a deletion symbol in addition to a stroke indicating a character string. The deletion symbol is a symbol (scratch) that fills the character. Specifically, it will be described later using the example of FIG.
The stroke information 105 includes, for example, stroke information 400. FIG. 4 is an explanatory diagram showing an example of the data structure of the stroke information 400. The stroke information 400 has a time column 410, a coordinate position column 420, and a pen up / down column 430. The time column 410 stores the time. The time here may be information indicating time series (information indicating the order in which coordinate positions are generated). The coordinate position column 420 stores coordinate positions. The coordinate position indicates the position of the electronic pen on the paper (for example, XY coordinates). The pen up / down column 430 stores information indicating pen up and pen down. Pen-up indicates that the electronic pen has been released from the paper, and pen-down indicates that the electronic pen has been pressed against the paper.
For example, the stroke information 400 is generated by using the coordinate position of the electronic pen obtained at a frequency of about 70 to 100 times per second and the up / down information of the electronic pen.
A state where the electronic pen is down can be regarded as a state where characters are written on paper. That is, it can be regarded as one stroke (a line of characters written with a single stroke) from when the electronic pen is down to when it is up.
Using this stroke information 105, the online character recognition module 110 recognizes characters written on paper. There are various methods for recognizing characters using the stroke information 105.
As a result of character recognition by the online character recognition module 110, one or more characters (recognition result (with stroke information) 115) are output. As a result of character recognition, it is possible to determine which character the stroke corresponds to for each of a plurality of strokes.
Furthermore, a rectangle surrounding each character can be formed. As a rectangle surrounding a character, for example, there is a circumscribed rectangle of the character. A circumscribed rectangle can be calculated using the maximum and minimum values of the X-axis and Y-axis of the coordinate information of the stroke corresponding to a certain character.

The stroke scratch determination module 120 is connected to the online character recognition module 110 and the character removal module 130. The stroke scratch determination module 120 receives a recognition result (with stroke information) 115 from the online character recognition module 110. As described above, the recognition result (with stroke information) 115 includes a stroke corresponding to each character. The stroke scratch determination module 120 extracts a stroke group in the received recognition result (with stroke information) 115 for each character. That is, for each recognized character, a stroke used for recognizing the character is extracted.
Next, the stroke scratch determination module 120 extracts a rectangle surrounding the character for each character. A circumscribed rectangle as described above may be used. Alternatively, the rectangle may be extracted with the height of the character string as the height of the rectangle or the width of the character string as the width of the rectangle. The character string here is a character string including the target character. In the case of horizontal writing, “the height of the character string is set as the height of the rectangle”, and the rectangle is extracted. In the case of vertical writing, “the width of the character string is set as the width of the rectangle” and the rectangle is extracted. For example, in horizontal writing, punctuation marks such as “.”, Urgent sounds such as “tsu”, “yo”, etc., and small characters indicating stuttering are rectangles in which the circumscribed rectangle is expanded vertically.

Next, the stroke scratch determination module 120 divides the extracted rectangle into regions (hereinafter also referred to as meshes). The region is a rectangle smaller than the rectangle because it divides the rectangle. The number of rectangular divisions may be determined in advance, such as K in the vertical direction and L in the horizontal direction. For example, K = 6, L = 4, etc.
Next, for each stroke extracted for each character, the stroke scratch determination module 120 counts the number of regions in which the stroke exists. For example, it is only necessary to follow a stroke and extract a region where the coordinates exist. Alternatively, the intersection of the stroke and the boundary line of the area may be calculated, and the area in contact with the calculated intersection may be counted as the area where the stroke exists.
Next, the stroke scratch determination module 120 determines whether or not the characters in the rectangle are deleted based on the counted number of areas. If the target stroke is a scratch, the characters in the rectangle are deleted. Let a be the number of regions where strokes exist.
For example,
P = a / (K × L)
Is calculated. In this way, the number of stroke existing areas normalized by the total number of areas is defined as a stroke existence ratio (P).
If P is larger than a predetermined threshold T, it is determined that the stroke is a deletion symbol (scratch). T is, for example, 0.7.
Then, the stroke scratch determination module 120 passes the determination result 125 as the processing result to the character removal module 130. The determination result 125 includes a recognition result and information indicating the deleted character (for example, information indicating that the Xth character is a deleted character).

  The character removal module 130 is connected to the stroke scratch determination module 120. The character removal module 130 removes characters overwritten with strokes that are deletion symbols, and outputs a character string of the recognition result 135.

FIG. 2 is a flowchart illustrating a processing example according to the first exemplary embodiment.
In step S202, the online character recognition module 110 receives the stroke information 105. The example of FIG. 5A is a character string (target stroke group 500) before erasure. In the example of FIG. 5B, only B (deleted character (B) 510) of ABC (target stroke group 500) is erased by scratch. That is, as shown in the example of FIG. 5A, after handwriting “ABC” as the target stroke group 500 using the electronic pen, the electronic pen is used as shown in the example of FIG. 5B. When the scratch is applied on the character “B”, the deleted character (B) 510 is designated by the operator's writing.
In step S204, the online character recognition module 110 recognizes the stroke information 105 as characters. As a result of character recognition, a circumscribed rectangle (circumscribed rectangle 520, circumscribed rectangle 530, circumscribed rectangle 540) of each character can be set for each character as shown in FIG. A character string “ABC” is obtained as a character recognition result. Note that although there is a scratch stroke in the circumscribed rectangle 530, a portion where characters can be recognized may be extracted, or after writing as in the example of FIG. 5A, FIG. 5B. When the scratch is written as in the example (that is, when the scratch is written on “B” after “C” is written), the character string “ABC” is obtained from the time series information.
In step S206, the stroke scratch determination module 120 determines whether the stroke is a scratch from the stroke information 105, and identifies the deleted character. The process of step S206 will be described later using the example of FIG.
In step S208, the character removal module 130 deletes the deleted character specified from the character string that is the character recognition result.
In step S210, the character removal module 130 outputs the recognition result 135.

FIG. 3 is a flowchart illustrating a processing example according to the first exemplary embodiment.
In step S302, a circumscribed rectangle of the character is extracted. The example shown in FIG. 5B is a circumscribed rectangle 520, a circumscribed rectangle 530, and a circumscribed rectangle 540, as shown in the example of FIG. 5C.
Here, although the example of FIG. 5C is delimited by the circumscribed rectangle of the character, as described above, since it may not be delimited by the circumscribed rectangle of the character, it is exactly “character”. Note that this is not a requirement.
In step S304, the circumscribed rectangle is divided into a predetermined number of meshes. For example, FIG. 5D shows an example in which the circumscribed rectangle 530 is divided into meshes. The example shown in FIG. 5D is a diagram in which only the scratch stroke of the character B (deleted character (B) 510) is extracted and enlarged, and is shown in B (deleted character (B) 510). The circumscribed rectangle 530 is divided by a mesh as shown in the example of FIG. Of course, the circumscribed rectangle 520 and the circumscribed rectangle 540 are similarly divided into meshes, but it is determined that there is no scratch (step S314).
In step S306, one stroke in the circumscribed rectangle is extracted.

In step S308, the mesh in which the stroke extracted in step S306 exists is counted. That is, it is determined whether or not a scratch stroke exists in each mesh. For example, FIG. 5E shows an example in which a mesh having a stroke in the circumscribed rectangle 530 is indicated by hatching. Let a be the number of meshes with N meshes and strokes inside. In the example of FIG. 5D, since the number of meshes is 4 in the horizontal direction and 6 in the vertical direction, N is 4 × 6 = 24. The mesh through which the stroke passes is a portion indicated by hatching in the example of FIG. 5E, and a = 22.
In step S310, it is determined whether or not the existing density of the stroke> the threshold value. If the existing density of the stroke> the threshold value, the process proceeds to step S316. Otherwise, the process proceeds to step S312. If it demonstrates using the above-mentioned example, when a / N is larger than a threshold value, it will judge that the stroke is a scratch. For example, the threshold value is 0.7, and in the example of FIG. 5E, 0.7 <a / N, so this stroke is a scratch.

In step S312, it is determined whether or not there is a stroke to be determined next in the target character. If there is a stroke, the process returns to step S306. Otherwise, the process proceeds to step S314. That is, the process from step S306 to step S310 is performed for all strokes in the circumscribed rectangle.
In step S314, it is determined that there is no scratch (not a deleted character) in the circumscribed rectangle.
In step S316, it is determined that the stroke is a scratch (a deleted character).
In step S318, it is determined whether or not there is a character to be processed next. If there is, the process returns to step S302, and otherwise, the process ends (step S399).

FIG. 6 is an explanatory diagram illustrating a processing example according to the first exemplary embodiment. This is an example of the process of step S308, and shows an example using the intersection of the stroke and the boundary line of the mesh described above.
A method for calculating the number m of meshes having strokes will be specifically described.
First, as shown in the example of FIG. 6, the X axis and the Y axis are set. In this setting method, any Cartesian coordinates that can specify the position of a point in two dimensions may be used. The lines in the X-axis direction that divide the circumscribed rectangle of a character into meshes are X1 to Xn. Similarly, the lines in the Y-axis direction are Y1 to Ym. Also, a memory (array) indicating whether or not a stroke exists in each mesh is prepared. This array has two types of values: ON (there is a stroke) and OFF (there is no stroke). The initial values are all OFF.
One stroke from the start point 610 to the end point 620 is composed of a plurality of point sequences.
Target two consecutive points. A line segment connecting these two points is the object.
First, the mesh where the end point of the line segment exists is turned ON.
Next, the X coordinate maximum value Xmax and minimum value Xmin, the Y coordinate maximum value Ymax, and the minimum value Ymin of the two end points of the line segment are obtained.
A line Xi (i = 1 to n) existing between Xmax and Xmin is extracted.
The position where the line segment intersects with Xi (i = 1 to n) is calculated. By using the intersecting line and the X coordinate of the intersecting position, it is possible to know in which mesh the line segment exists.
For example, the X coordinate at which the line segment intersects with Xi in FIG. 6 is the position indicated by the arrow. Among the meshes present at this position, the line segment passes through a mesh having Xi as a boundary. The array corresponding to the mesh indicated by the diagonal lines is set to ON.
This is done for all extracted Xi.
Furthermore, the above is performed with a line segment made of all two consecutive points.
What is necessary is just to count the number of meshes in which the array is finally turned ON.

<Second Embodiment>
FIG. 7 is an explanatory diagram illustrating a processing example according to the second exemplary embodiment. In the second embodiment, the stroke scratch determination module 120 accepts a stroke group, extracts the accepted stroke group for each character, extracts a rectangle surrounding the character for each character, and extracts the stroke for each stroke. Based on the normalized stroke length, normalize the measured stroke length using one or both of the height and width of the rectangle that contains the stroke. Thus, it is determined whether or not the characters in the rectangle have been deleted. Note that “accepting a stroke group, extracting the accepted stroke group for each character, and extracting a rectangle surrounding the character for each character” is a process equivalent to the first embodiment.

In the second embodiment, the stroke scratch determination module 120 evaluates the stroke length.
First, the height H of the rectangle surrounding the character is obtained. This may be the height of the rectangle of the character or the height of the entire character string.
Next, the stroke length is measured. The length of the stroke can be measured, for example, as the sum of the distances between all two consecutive points. Let Q be the length of the stroke. A threshold value S is prepared in advance, and if (Q / H)> S, it is determined that the stroke is a scratch. The stroke length normalized with the size H of the rectangle surrounding the character in this way is taken as the stroke length ratio.
In this example, the height H of the rectangle surrounding the character is used, but the width W of the rectangle surrounding the character may be used. Alternatively, a number calculated from the height of the rectangle surrounding the character and the width of the rectangle may be used.
For example, the following equation may be used.
(H + W) / 2
sqrt (H × W) (where sqrt () is a function for obtaining the square root)

<Third Embodiment>
FIG. 8 is a conceptual module configuration diagram of an exemplary configuration according to the third embodiment.
In the first embodiment, the scratch is determined using the abundance ratio of strokes existing in the mesh. In the second embodiment, scratch determination is made using the stroke length ratio. In the third embodiment, a combination of these is determined.
The example of FIG. 8 shows a module configuration example in the stroke scratch determination module 120 in the first embodiment (second embodiment).
The stroke scratch determination module 120 includes a stroke existence ratio A calculation module 810, a stroke length ratio B calculation module 820, and a determination module 830.
The stroke presence ratio A calculation module 810 is connected to the determination module 830. The stroke existence ratio A calculation module 810 receives the recognition result (with stroke information) 115 from the online character recognition module 110, and performs the process of the stroke scratch determination module 120 in the first embodiment to calculate the stroke existence ratio A. To do.
The stroke length ratio B calculation module 820 is connected to the determination module 830. The stroke length ratio B calculation module 820 receives the recognition result (with stroke information) 115 from the online character recognition module 110, performs the processing of the stroke scratch determination module 120 in the second embodiment, and performs the stroke length ratio B. Is calculated.
The determination module 830 is connected to the stroke existence ratio A calculation module 810 and the stroke length ratio B calculation module 820. The determination module 830 includes the number of areas counted by the stroke existence ratio A calculation module 810 and the stroke length normalized by the stroke length ratio B calculation module 820 (including the stroke existence ratio A and the stroke length ratio B). ) To determine whether the characters in the rectangle have been deleted. Then, the determination result 125 is passed to the character removal module 130.

When “the stroke existence ratio A is larger than the threshold value T1” and “the stroke length ratio B is larger than the threshold value T2,” it may be determined that the scratch is present, or “the stroke existence ratio A is larger than the threshold value T1. Or “the stroke length ratio B is greater than the threshold value T2” may be determined as a scratch. Further, an order may be provided to determine whether or not “the stroke length ratio B is greater than the threshold value T2” when “the stroke existence ratio A is greater than the threshold value T1”. When “the stroke length ratio B is greater than the threshold value T2”, it may be determined whether or not “the stroke existence ratio A is greater than the threshold value T1”. When the order is provided, the determination module 830 determines whether or not to perform the other process after the one process of the stroke existence ratio A calculation module 810 or the stroke length ratio B calculation module 820 is completed. Also good.
Further, instead of preparing a threshold value for each individual determination in advance, the determination may be performed using a general two-class classification machine learning method. For example, various machine learning methods such as neural network, adaboost, and SVM can be used. For example, the determination module 830 may be a two-class classifier configured using a machine learning device such as a neural network, an adaboost, or an SVM. The determination module 830 learns the correspondence between the stroke existence ratio A and the stroke length ratio B and the scratch / non-scratch determination result in advance, and includes the learning result. The determination module 830 receives the values of the stroke existence ratio A and the stroke length ratio B calculation module 820 from the stroke existence ratio A calculation module 810 and outputs the determination result 125.
In the example of FIG. 8, only the stroke existence ratio A and the stroke length ratio B are accepted, but other values may be input. However, when accepting another value, it is necessary to learn in advance using that value.

FIG. 12 is an explanatory diagram showing a system example in the case of realizing the present embodiment.
The electronic pen paper printing system 1220 and the writing information processing system 1230 are connected via a communication line 1299 (wired, wireless, or a mixed line). A printing device 1225 is connected to the electronic pen paper printing system 1220, and an electronic pen 1235 is connected to the writing information processing system 1230. The module configuration illustrated in FIGS. 1, 8, etc. is mainly constructed in the writing information processing system 1230.
The electronic pen paper printing system 1220 is a system that uses a printing device 1225 to print a document on which information images (hereinafter also referred to as dot code images) are superimposed using a paper ID. The writing information processing system 1230 is a system for superimposing writing information on a document when writing is performed using the electronic pen 1235 on a sheet on which an information image is printed by the electronic pen sheet printing system 1220. . When the operator uses the electronic pen 1235 to detect a deletion symbol, the character string (corrected character string) from which the deletion symbol is superimposed is superimposed on the document.

FIG. 13 is an explanatory diagram illustrating an example of electronic pen paper 1310 on which an information image is printed by the electronic pen paper printing system 1220. The electronic pen paper 1310 is printed by the electronic pen paper printing system 1220 using the printing device 1225. A dot code image is printed on the electronic pen paper 1310. For example, in a region 1320 in the electronic pen paper 1310, a dot code image as shown in FIG. 13B, which is an enlarged view of the region 1320, is printed. The dot code image represents the paper ID assigned to each electronic pen paper 1310 and the position information (X, Y coordinate values) on the paper.
For example, the paper ID is a numerical value in a 32-bit space. In the case of character string notation, it is represented by a hexadecimal character string. Therefore, the range of the paper ID is “00000000” to “FFFFFFFF”.

FIG. 14 is an explanatory diagram illustrating a configuration example in the electronic pen 1235.
An outline will be described. When writing with the electronic pen 1235 on the electronic pen paper 1499, when the pressure sensor is turned on (the above-described pen down), a dot code image on the electronic pen paper 1499 is imaged, decoded, and the paper ID of the electronic pen paper 1499. The position information (X, Y coordinate values) on the electronic pen paper 1499 is taken out and stored in the memory. Then, the information stored in the memory is transmitted to the writing information processing system 1230 via the communication circuit. In addition, like the stroke information 400 shown in the example of FIG. 4, the above-described writing pressure ON information (the above-mentioned pen down) is the information that the pressure sensor is turned on, and the above-mentioned writing pressure is the information that the pressure sensor is turned off The OFF information (the pen-up described above) may also be stored in the memory. Moreover, you may make it include the information regarding the time when each information generate | occur | produced (it may be a year, a month, a day, a second, below second, or these combination).

Next, this will be described in detail. As illustrated, the electronic pen 1235 includes a control circuit 1401 that controls the operation of the entire pen. The control circuit 1401 includes an image processing unit 1401a that processes a dot code image detected from an input image, and a data processing unit 1401b that extracts a paper ID and position information from the processing result there.
The control circuit 1401 is connected to a pressure sensor 1402 that detects a writing operation by the electronic pen 1235 by a pressure applied to the pen tip 1409. In addition, an infrared LED 1403 for irradiating infrared light on the paper and an infrared CMOS 1404 for inputting an image are also connected. Further, an information memory 1405 for storing paper ID and position information, a communication circuit 1406 for communicating with an external device, a battery 1407 for driving the electronic pen 1235, and identification information for the electronic pen 1235 (pen A pen ID memory 1408 for storing (ID) is also connected.

Here, an outline of the operation of the electronic pen 1235 will be described.
When writing with the electronic pen 1235 is performed, the pressure sensor 1402 connected to the pen tip 1409 detects the writing operation. As a result, the infrared LED 1403 is turned on, and the infrared CMOS 1404 captures an image on the sheet by the CMOS sensor.
Note that the infrared LED 1403 is pulse-lit in synchronization with the shutter timing of the CMOS sensor in order to suppress power consumption.
The infrared CMOS 1404 uses a global shutter type CMOS sensor that can simultaneously transfer captured images. A CMOS sensor having sensitivity in the infrared region is used. In order to reduce the influence of disturbance, a visible light cut filter is disposed on the entire surface of the CMOS sensor. The CMOS sensor captures an image with a period of about 70 fps to 100 fps (frame per second). The image sensor is not limited to a CMOS sensor, and other image sensors such as a CCD may be used.

When the captured image is input to the control circuit 1401, the control circuit 1401 acquires a dot code image from the captured image. Then, it is decoded to obtain the paper ID and position information embedded in the dot code image.
Hereinafter, the operation of the control circuit 1401 at this time will be described.
FIG. 15 is a flowchart illustrating an example of processing performed by the electronic pen 1235 (control circuit 1401).
In step S1501, the image processing unit 1401a inputs an image.
In step S1502, a process for removing noise included in the image is performed. Here, the noise includes variations in CMOS sensitivity, noise generated by an electronic circuit, and the like. What processing is performed to remove noise should be determined according to the characteristics of the imaging system of the electronic pen 1235. For example, sharpening processing such as blurring processing or unsharp masking can be applied.
In step S1503, the image processing unit 1401a detects a dot pattern (dot image position) from the image. For example, the dot pattern portion and the background portion are separated by binarization processing, and the dot pattern can be detected from each binarized image position. When a binarized image contains a lot of noise components, for example, it is necessary to combine a filter process for determining a dot pattern based on the area and shape of the binarized image.

In step S1504, the image processing unit 1401a converts the detected dot pattern into digital data on a two-dimensional array. For example, on a two-dimensional array, a position where there is a dot is converted to “1”, and a position where there is no dot is converted to “0”. The digital data on the two-dimensional array is transferred from the image processing unit 1401a to the data processing unit 1401b.
In step S1505, the data processing unit 1401b detects a bit pattern including a combination of two dots shown in FIG. 16A from the received digital data. For example, the bit pattern can be detected by moving the boundary position of the block corresponding to the bit pattern on the two-dimensional array and detecting the boundary position so that the number of dots included in the block is two. .
When the bit pattern is detected in this way, in step S1506, the data processing unit 1401b detects the synchronization code by referring to the type of the bit pattern.
In step S1507, the identification code and the position code are detected based on the positional relationship from the synchronization code.
Thereafter, in step S1508, the data processing unit 1401b decodes the identification code to obtain the paper ID, and decodes the position code to obtain position information. About identification code, paper ID is obtained by performing RS decoding processing. On the other hand, for the position code, position information is obtained by comparing the position of the read partial series with the M series used at the time of image generation.

Next, the electronic document for storing written information will be described. Hereinafter, when simply referred to as an electronic document, it refers to an electronic document for storing written information.
The writing information storing electronic document is data in which the contents written on the electronic pen paper 1499 by the electronic pen 1235 are collected together with the field definition. Consists of:
(1) Paper ID: Paper ID assigned to the electronic pen paper associated with the writing information storing electronic document
(2) Field definition: Field definition used for processing writing to the electronic document for storing the writing information.
(3) Form image: Form image printed on the electronic document for storing written information

Next, a code pattern that is the basis of a dot code image generated by the electronic pen paper printing system 1220 will be described.
FIG. 16 is an explanatory diagram illustrating an example of an information image (code pattern image) handled by the electronic pen paper printing system 1220.
First, the bit pattern constituting the code pattern will be described.
FIG. 16A shows an example of bit pattern arrangement.
A bit pattern is the minimum unit of information embedding. Here, as shown in FIG. 16A, bits are arranged at two locations selected from nine locations. In the figure, black squares indicate positions where bits are arranged, and hatched squares indicate positions where bits are not arranged. There are 36 (= 9C2) combinations for selecting 2 locations out of 9 locations. Therefore, 36 kinds (about 5.2 bits) of information can be expressed by such an arrangement method.
However, the paper ID and the position information are expressed using 32 (5 bits) of these 36 patterns.
Incidentally, the minimum square shown in FIG. 16A has a size of 2 dots × 2 dots at 600 dpi. Since the size of one dot at 600 dpi is 0.0423 mm, one side of this minimum square is 84.6 μm (= 0.0423 mm × 2). The larger the dots that make up the code pattern, the more likely it is to be noticeable. However, if it is too small, printing with a printer becomes impossible. Therefore, the above-described value larger than 50 μm and smaller than 100 μm is adopted as the dot size. Thereby, it is possible to form dots of an optimum size that can be printed by the printer. That is, 84.6 μm × 84.6 μm is the minimum size that can be stably formed by the printer.
In addition, by making the dot such a size, one side of one bit pattern becomes about 0.5 (= 0.0423 × 2 × 6) mm.
A code pattern composed of such bit patterns will be described.
FIG. 16B shows an example of the arrangement of code patterns.
Here, the minimum square shown in FIG. 16B corresponds to the bit pattern shown in FIG. That is, the identification code obtained by encoding the paper ID is embedded using 16 (= 4 × 4) bit patterns. Also, the X position code obtained by encoding the position information in the X direction and the Y position code obtained by encoding the position information in the Y direction are each embedded using four bit patterns. Further, a synchronization code for detecting the position and rotation of the code pattern is embedded in the upper left corner using one bit pattern.
Since the size of one code pattern is equal to the width of five bit patterns, it is about 2.5 mm. In the electronic pen paper printing system 1220, a code pattern image obtained by imaging the code pattern generated in this way is arranged on the entire surface of the paper.

FIG. 17 is an explanatory diagram showing an example of information encoding processing and an example of information image (dot code image) generation processing in the electronic pen paper printing system 1220.
First, paper ID encoding will be described.
For encoding the paper ID, an RS (Reed Solomon) code of a block encoding method is used. As described in the example of FIG. 16, the electronic pen paper printing system 1220 embeds information using a bit pattern that can represent 5-bit information. Therefore, since an information error also occurs in units of 5 bits, an RS code having good coding efficiency is used in the block coding method. However, the encoding method is not limited to the RS code, and other encoding methods such as a BCH code can also be used.
As described above, in the electronic pen paper printing system 1220, information is embedded using a bit pattern having a 5-bit information amount. Therefore, the block length of the RS code needs to be 5 bits. For this reason, the paper ID is divided into blocks of 5 bits. In FIG. 17, a first block “00111” and a second block “01101” are cut out from the paper ID “0011101101001...”.
Then, RS encoding processing is performed on the blocked paper ID. In FIG. 17, “blk1”, “blk2”, “blk3”, “blk4”,... Are blocked and then RS-encoded.
By the way, in the electronic pen paper printing system 1220, the paper ID is divided into 16 (= 4 × 4) blocks. Therefore, the number of code blocks in the RS code can be 16.
Further, the number of information blocks can be designed according to an error occurrence state. For example, if the number of information blocks is 8, RS (16, 8) code is obtained. This code can correct even if an error of 4 blocks (= (16−8) / 2) occurs in the encoded information. Moreover, if the position of the error can be specified, the correction capability can be further improved. In this case, the amount of information stored in the information block is 40 bits (= 5 bits × 8 blocks), of which 32 bits are used.

Next, encoding of position information will be described.
For encoding the position information, an M-sequence code, which is a kind of pseudo-random sequence, is used. Here, the M sequence is a sequence of the maximum period that can be generated by a K-stage linear shift register, and has a sequence length of 2K-1. Arbitrary consecutive K bits taken out from the M sequence have a property that they do not appear at other positions in the same M sequence. Therefore, the position information can be encoded by using this property.
By the way, in the electronic pen paper printing system 1220, the required M-sequence order is obtained from the length of the position information to be encoded, and the M-sequence is generated. However, if the length of the position information to be encoded is known in advance, it is not necessary to generate the M sequence each time. That is, a fixed M sequence may be generated in advance and stored in a memory or the like.
For example, it is assumed that an M sequence (K = 13) having a sequence length of 8191 is used.
In this case, since the position information is also embedded in units of 5 bits, 5 bits are extracted from the M series having a sequence length of 8191 and blocked. In FIG. 17, the M sequence “11010011011010...” Is blocked by 5 bits.

  Thus, the electronic pen paper printing system 1220 uses different encoding methods for position information and paper ID. This is because it is necessary to set the paper ID detection capability to be higher than the position information detection capability. That is, since the position information is information for acquiring the position of the paper surface, even if there is a part that cannot be decoded due to noise or the like, the part is lost and does not affect the other part. On the other hand, if the paper ID fails to be decrypted, it is impossible to detect the target reflecting the written information. Further, with such a configuration, it is possible to minimize the image reading range when the position information and the paper ID are decoded. That is, if an encoding method having a boundary such as an RS code is used for position information, it is necessary to read the code between the boundaries when decoding it, so the range for reading the image is shown in FIG. The area needs to be twice as large as the area. However, by using the M series, it is possible to have a configuration in which an area having the same size as the area shown in FIG. This is because the position information can be decoded from an arbitrary partial sequence of the M sequence due to the nature of the M sequence. That is, when decoding the paper ID and the position information, it is necessary to read the area having the size shown in FIG. 16B, but the read position matches the boundary shown in FIG. There is no need. The position information can be decoded from a partial series at an arbitrary position of the M series. As for the paper ID, since the same information is arranged on the entire surface of the paper, even if the reading position is shifted from the boundary illustrated in FIG. 16B, the original information is restored by rearranging the pieces of the read information. be able to.

As described above, after the paper ID is divided into blocks, it is encoded with the RS code, and when the position information is encoded with the M series and then divided into blocks, the blocks are synthesized as shown in the figure. The That is, these blocks are developed on a two-dimensional plane in the format shown in the figure. The format shown in FIG. 17 corresponds to the format shown in FIG. That is, a black square means a synchronization code.
Further, “1”, “2”, “3”, “4”,... Arranged in the horizontal direction represent X position codes, and “1”, “2”, “3”, “ 4 ”,... Mean Y position codes. The position code is indicated by a number corresponding to the coordinate position because different information is arranged if the position of the paper is different. On the other hand, the shaded squares represent the identification codes. Since the same information is arranged even if the position of the sheet is different, the identification codes are all indicated by the same mark.
By the way, as can be seen from the figure, there are four bit patterns between two synchronization codes. Therefore, 20 (= 5 × 4) -bit M-sequence partial sequences can be arranged. If a 13-bit partial sequence is extracted from the 20-bit partial sequence, it is possible to specify which partial sequence in the whole (8191) the 13 bits are. As described above, when 13 bits out of 20 bits are used for specifying the position, it is possible to detect or correct the extracted 13-bit error using the remaining 7 bits. That is, it is possible to detect and correct errors by confirming 20-bit consistency using the same generator polynomial as that used when generating the M-sequence.
Thereafter, the bit pattern in each block is imaged by referring to the dot image. Then, an output image representing information with dots as shown on the rightmost side of FIG. 17 is generated.

  Note that the hardware configuration of the computer on which the program according to the present embodiment is executed is a general computer, specifically, a personal computer, a computer that can be a server, or the like, as illustrated in FIG. That is, as a specific example, the CPU 1801 is used as a processing unit (calculation unit), and the RAM 1802, the ROM 1803, and the HD 1804 are used as storage devices. For example, a hard disk may be used as the HD 1804. CPU 1801 for executing programs such as the online character recognition module 110, the stroke scratch determination module 120, the character removal module 130, the stroke existence ratio A calculation module 810, the stroke length ratio B calculation module 820, the determination module 830, and the programs and data A RAM 1802 that stores a program, a ROM 1803 that stores a program for starting up the computer, an HD 1804 that is an auxiliary storage device (may be a flash memory or the like), and a user for a keyboard, a mouse, a touch panel A communication device for connecting to a receiving device 1806 for receiving data based on the operation of the device, an output device 1805 such as a CRT or liquid crystal display, and a communication network such as a network interface card Interface 1807, and, and a bus 1808 for exchanging data by connecting them. A plurality of these computers may be connected to each other via a network.

Among the above-described embodiments, the computer program is a computer program that reads the computer program, which is software, in the hardware configuration system, and the software and hardware resources cooperate with each other. Is realized.
Note that the hardware configuration illustrated in FIG. 18 illustrates one configuration example, and the present embodiment is not limited to the configuration illustrated in FIG. 18, and is a configuration capable of executing the modules described in the present embodiment. I just need it. For example, some modules may be configured by dedicated hardware (for example, ASIC), and some modules may be in an external system and connected via a communication line. A plurality of systems shown in FIG. 5 may be connected to each other via communication lines so as to cooperate with each other. In particular, in addition to personal computers, information appliances, copiers, fax machines, scanners, printers, and multifunction machines (image processing apparatuses having two or more functions of scanners, printers, copiers, fax machines, etc.) Etc. may be incorporated.

In the above-described embodiment, it is determined whether or not the stroke is a scratch using the stroke existence ratio or the stroke length ratio. The threshold used for this determination may be changed according to the character size (rectangular size). For example, in the case of a rectangle whose number of vertical and horizontal pixels is smaller than a predetermined value, it may be difficult to superimpose a scratch. Considering such a case, processing such as lowering the threshold value is performed. Alternatively, the number of vertical and horizontal pixels per se may be received and learned as a feature quantity of a discriminator realized by a machine learning device such as a neural network, an adaboost, or an SVM.
Further, in the description of the above-described embodiment, “more than”, “less than”, “greater than”, and “less than (less than)” in a comparison with a predetermined value contradicts the combination. As long as the above does not occur, “larger”, “smaller (less than)”, “more than”, and “less than” may be used.

The program described above may be provided by being stored in a recording medium, or the program may be provided by communication means. In that case, for example, the above-described program may be regarded as an invention of a “computer-readable recording medium recording the program”.
The “computer-readable recording medium on which a program is recorded” refers to a computer-readable recording medium on which a program is recorded, which is used for program installation, execution, program distribution, and the like.
The recording medium is, for example, a digital versatile disc (DVD), which is a standard established by the DVD Forum, such as “DVD-R, DVD-RW, DVD-RAM,” and DVD + RW. Standard “DVD + R, DVD + RW, etc.”, compact disc (CD), read-only memory (CD-ROM), CD recordable (CD-R), CD rewritable (CD-RW), Blu-ray disc ( Blu-ray (registered trademark) Disc), magneto-optical disk (MO), flexible disk (FD), magnetic tape, hard disk, read-only memory (ROM), electrically erasable and rewritable read-only memory (EEPROM (registered trademark)) )), Flash memory, Random access memory (RAM) SD (Secure Digital) memory card and the like.
The program or a part of the program may be recorded on the recording medium for storage or distribution. Also, by communication, for example, a local area network (LAN), a metropolitan area network (MAN), a wide area network (WAN), a wired network used for the Internet, an intranet, an extranet, etc., or wireless communication It may be transmitted using a transmission medium such as a network or a combination of these, or may be carried on a carrier wave.
Furthermore, the program may be a part of another program, or may be recorded on a recording medium together with a separate program. Moreover, it may be divided and recorded on a plurality of recording media. Further, it may be recorded in any manner as long as it can be restored, such as compression or encryption.

105 ... Stroke information 110 ... Online character recognition module 115 ... Recognition result (with stroke information)
DESCRIPTION OF SYMBOLS 120 ... Stroke scratch determination module 125 ... Determination result 130 ... Character removal module 135 ... Recognition result 810 ... Stroke existence ratio A calculation module 820 ... Stroke length ratio B calculation module 830 ... Determination module

Claims (6)

Receiving means for receiving a stroke group;
Stroke extracting means for extracting a stroke group for each unit divided according to a predetermined rule from the stroke group received by the receiving means;
Rectangle extracting means for extracting a rectangle surrounding the stroke group for each unit;
Area dividing means for dividing the rectangle extracted by the rectangle extracting means into areas;
For each stroke extracted by the stroke extraction means, a counting means for counting the number of areas divided by the area dividing means in which the stroke exists;
An information processing apparatus comprising: determination means for determining whether or not the stroke group in the rectangle has been deleted based on the number of areas counted by the counting means. A calculation means for calculating an intersection between the stroke and the boundary line of the region;
The information processing apparatus according to claim 1, wherein the counting unit sets a region in contact with the intersection calculated by the calculation unit as a region where a stroke exists. The rectangle extraction means extracts the rectangle by using the height of the row of units as the height of the rectangle, or the width of the row of units as the width of the rectangle. The information processing apparatus described in 1. Receiving means for receiving a stroke group;
Stroke extracting means for extracting a stroke group for each unit divided according to a predetermined rule from the stroke group received by the receiving means;
Rectangle extracting means for extracting a rectangle surrounding the stroke group for each unit;
Area dividing means for dividing the rectangle extracted by the rectangle extracting means into areas;
For each stroke extracted by the stroke extraction means, a counting means for counting the number of areas divided by the area dividing means in which the stroke exists;
For each stroke extracted by the stroke extracting means, a measuring means for measuring the length of the stroke;
Normalizing means for normalizing the length of the stroke measured by the measuring means using either one or both of the height and width of the rectangle including the stroke;
Judgment means for judging whether or not the stroke group in the rectangle has been deleted based on the number of areas counted by the counting means and the length of the stroke normalized by the normalizing means. An information processing apparatus characterized by that. Computer
Receiving means for receiving a stroke group;
Stroke extracting means for extracting a stroke group for each unit divided according to a predetermined rule from the stroke group received by the receiving means;
Rectangle extracting means for extracting a rectangle surrounding the stroke group for each unit;
Area dividing means for dividing the rectangle extracted by the rectangle extracting means into areas;
For each stroke extracted by the stroke extraction means, a counting means for counting the number of areas divided by the area dividing means in which the stroke exists;
An information processing program for functioning as a judging means for judging whether or not the stroke group in the rectangle is deleted based on the number of areas counted by the counting means. Computer
Receiving means for receiving a stroke group;
Stroke extracting means for extracting a stroke group for each unit divided according to a predetermined rule from the stroke group received by the receiving means;
Rectangle extracting means for extracting a rectangle surrounding the stroke group for each unit;
Area dividing means for dividing the rectangle extracted by the rectangle extracting means into areas;
For each stroke extracted by the stroke extraction means, a counting means for counting the number of areas divided by the area dividing means in which the stroke exists;
For each stroke extracted by the stroke extracting means, a measuring means for measuring the length of the stroke;
Normalizing means for normalizing the length of the stroke measured by the measuring means using either one or both of the height and width of the rectangle including the stroke;
Based on the number of areas counted by the counting means and the length of the stroke normalized by the normalizing means, it functions as a judging means for judging whether or not the stroke group in the rectangle has been deleted. Information processing program.

Images