JP5292641B2 - Voting paper reader and reading method - Google Patents

Description

  The present invention relates to a ballot paper reader and a reading method.

  As a conventionally used ballot paper reader, for example, a configuration shown in Patent Document 1 below is used. That is, image information of a description surface (front side surface) of the voting paper is generated, and based on this image information, an entry frame attached to the voting paper is detected, and for example, a candidate described inside the entry frame It reads the description contents of a person or a political party and automatically classifies them according to the description contents.

  When a voter writes on the outside (outside the description area) of the ballot paper entry frame, the write is regarded as “other matter description” and the vote becomes an invalid vote.

Japanese Patent Laid-Open No. 2004-5022

  However, the conventional ballot paper reader has a problem in that it cannot properly detect writing outside the description area (other matters are described). Election names and cautionary notes may be printed outside of the entry area on the entry side (front side) of the ballot, but these prints and matters written outside the entry area by the voter Difficult to distinguish from (other matters).

  The present invention has been made to solve the above-described problems, and provides a ballot paper reading device and a reading method capable of appropriately detecting writing outside the description area (other matters). That is.

A voting paper reading apparatus according to the present invention includes a conveyance path through which a voting paper is conveyed, an imaging unit disposed opposite to the conveyance path, and an information processing unit connected to the imaging unit. When the ballot is transported along the transport path, image information representing the surface of the ballot is generated, and the information processing means is a voting that is not filled in by the voter and does not include other matters in the learning mode. The mask area is determined by expanding the colored dots representing the color in the image information representing the surface of the paper, and the information processing means adds the area formed by the dots corresponding to the inside of the entry frame to the mask area. In the determination mode, the means determines the presence / absence of other matters based on the colored dots other than the mask area in the image information.

The expansion of the colored dots may be performed by using dots positioned around the colored dots as part of the mask area.
The dots positioned around the colored dots may be dots positioned within a predetermined range from the colored dots.
The colored dots may be dots that represent printing on ballot paper or writing with a writing instrument.
An entry frame that divides a predetermined description area may be printed on the surface of the ballot paper, and the mask area may be further determined based on the predetermined description area.

  The ballot paper reading method according to the present invention uses the above-mentioned ballot paper reading device.

  According to the voting paper reader of the present invention, the mask area is determined by expanding the printed matter in the image information of the blank voting paper in the learning mode, and the completed voting is performed based on the coloring in the area other than the mask area in the determination mode. Since the presence / absence of other matters on the paper is determined, printing or the like included in the mask area can be excluded from the determination regarding the other matters description, and the other matters can be appropriately detected.

It is a block diagram which shows the reading apparatus of the voting paper by Embodiment 1 of this invention. It is a block diagram which shows the imaging unit of FIG. It is a figure which shows the description surface of an unfilled ballot. It is a figure which shows the description surface of a completed ballot. It is a flowchart which shows the flow of a process of the reading device of the voting paper in learning mode. It is a figure which shows the dot pattern explaining the expansion | swelling of a colored dot. 5 is a diagram illustrating an example of a mask region 50. FIG. It is a flowchart which shows the flow of a process of the reading device of the voting paper in determination mode. It is a figure which shows the example of the completed ballot paper which does not contain other matter description. It is a figure which shows the example of the completed ballot paper containing other matter description.

Hereinafter, embodiments for carrying out the present invention will be described with reference to the drawings.
Embodiment 1 FIG.
FIG. 1 is a block diagram showing a ballot paper reader according to Embodiment 1 of the present invention. In the figure, a voting paper reading device is provided with a reading device main body 1 and an extension stacker unit 2. The reading device main body 1 is provided with a hopper 10, a main body side transport unit 11, first to fifth stackers 12 to 16, an imaging unit 17, and a display unit 18.

  The hopper 10 is a case in which a plurality of voting sheets 3 input from the outside are stored. The main body side transport means 11 is composed of a plurality of transport rollers 11a, and transports the voting paper 3 along the transport direction 110 from the hopper 10 to the first to fifth stackers 12 to 16 in the first main body transport path 111. Is forming. The main body side transport means 11 is separated from the midway position 111a of the first main body transport path 111, and forms a second main body transport path 112 for transporting the voting paper 3 toward the expansion stacker unit 2. Yes. The imaging unit 17 generates image information indicating the end face of the voting paper 3 as described later, and is disposed downstream of the hopper 10 and upstream of the midway position 111a along the transport direction 110. The display means 18 displays information for the user.

  The extension stacker unit 2 is an extension unit that can be separated from the reader main body 1. The expansion stacker unit 2 is provided with an expansion side conveying means 20 and sixth to thirteenth stackers 21 to 28. The expansion side conveyance means 20 forms an expansion side conveyance path 200 that conveys the voting paper 3 conveyed to the expansion stacker unit 2 through the second main body conveyance path 112 to the sixth to thirteenth stackers 21 to 28. Yes.

  The first to thirteenth stackers 12 to 16 and 21 to 28 are cases in which voting sheets 3 classified as described later are respectively stored. In this embodiment, an example is shown in which one expansion stacker unit 2 is connected to the reading apparatus main body 1. However, a plurality of expansion stacker units 2 can read depending on the number of stackers required. It may be connected to the apparatus main body 1. Although the extension stacker unit 2 has been described as being separable from the reader main body 1, the reader main body 1 and the extension stacker unit 2 may be provided integrally with each other.

  Next, FIG. 2 is a block diagram showing the imaging unit 17 of FIG. The imaging unit 17 includes first and second guide plates 170 and 171, first to third roller portions 172 to 174, first and second imaging means 175 and 176, and first and second opposing rollers 177. , 178.

  The first and second guide plates 170 and 171 are plate members that extend along the transport direction 110 and are disposed to face each other so as to sandwich the first in-body transport path 111. The first and second guide plates 170 and 171 guide the voting paper 3 when the voting paper 3 passes through the imaging unit 17. Although not shown in detail, the first and second guide plates 170 and 171 are each provided with an opening at each position where the first to third roller portions 172 to 174 are disposed. .

  The first roller portion 172 includes a pair of rollers 172a that are disposed with the first main body conveyance path 111 interposed therebetween so that the outer peripheral surfaces of the first roller portion 172 come into contact with each other through the openings of the first and second guide plates 170, 171. ing. Similarly to the first roller portion 172, the second and third roller portions 173 and 174 are also composed of a pair of rollers 173a and 174a. The first to third roller portions 172 to 174 are arranged to be separated from each other along the transport direction 110 and send out the voting paper 3 that has entered between the rollers 172a to 174a along the transport direction 110 ( Transported). That is, the first to third roller portions 172 to 174 constitute a part of the main body side transport unit 11.

  Here, one end surface of the first in-body conveyance path 111 on the side where the first guide plate 170 is disposed is referred to as a conveyance path surface 113, and the other end surface on the side where the second guide plate 171 is disposed. This is referred to as the conveyance path back surface 114. In other words, when the ballot paper 3 passes through the imaging unit 17, the end surface of the first in-body conveyance path 111 facing the first and second end surfaces 3 a and 3 b of the ballot paper 3 is the conveyance path surface 113 and the conveyance path. This is called the back surface 114. Depending on the orientation of the voting paper 3, the first end surface 3a is one of the front surface (description surface) and the back surface of the voting paper 3, and the second end surface 3b is the other of the front surface (description surface) and the back surface. Further, when the directions of the front and back directions and the top and bottom directions of the ballot paper 3 are not aligned with each other, four directions of the ballot paper 3 are assumed. That is, they are front-facing, front-facing, back-facing, and lining-facing.

  The first imaging means 175 is disposed between the first roller portion 172 and the second roller portion 173 so as to face the transport path surface 113 through the opening of the first guide plate 170. The first imaging unit 175 includes a light source unit 175a and an imaging unit 175b, and generates first image information 175c indicating the first end surface 3a of the voting paper 3 when the voting paper 3 passes through the imaging unit 17. To do. Specifically, when the voting paper 3 passes through the imaging unit 17, the imaging light is irradiated from the light source unit 175a to the first end surface 3a of the voting paper 3, and the imaging light reflected by the first end surface 3a is reflected. The first image information 175c is generated by receiving light at the imaging unit 175b.

  The first opposing roller 177 is disposed between the first roller portion 172 and the second roller portion 173 so as to face the imaging side end surface of the first imaging means 175 through the opening of the second guide plate 171. The first counter roller 177 is configured to send the voting paper 3 along the transport direction 110 when the voting paper 3 enters between the outer peripheral surface and the imaging side end surface of the first imaging means 175. It is arranged close to the imaging side end surface of the imaging means 175.

The second imaging unit 176 is disposed between the second roller unit 173 and the third roller unit 174 so as to face the conveyance path back surface 114 through the opening of the second guide plate 171. That is, the second imaging unit 176 is disposed downstream of the first imaging unit 175 along the transport direction 110 and is not opposed to the first imaging unit 175. The second image pickup unit 176 includes a light source unit 176a and an image pickup unit 176b similarly to the first image pickup unit 175 described above, and shows the second end surface 3b of the voting paper 3 on the side opposite to the first end surface 3a. The second image information 176c is generated.
Each of the first image information 175c and the second image information 176c represents an image by a set of two-dimensionally arranged dots.

  The second facing roller 178 is disposed between the second roller portion 173 and the third roller portion 174 so as to face the second imaging unit 176 through the opening of the first guide plate 170. Similar to the first counter roller 177, the second counter roller 178 is disposed close to the imaging side end face of the second imaging means 176.

  The information processing means 4 is connected to the first and second imaging means 175 and 176. The information processing means 4 is composed of a computer including a storage unit that stores information such as a program and an arithmetic unit that performs a processing operation based on the information stored in the storage unit. Specifically, the information processing means 4 is configured by a computer built in the reader main body 1 or a combination of a computer provided separately from the reader main body 1 and a computer built in the reader main body 1. ing. As will be described later, the information processing means 4 processes the first and second image information 175c and 176c from the first and second imaging means 175 and 176.

  FIG. 3 shows a description surface of the voting paper 3. This ballot 3 is an unfilled ballot 31 that has not yet been filled in by the voter. Predetermined print items are printed on the description surface of the voting paper 3. This print item includes a print character 301 and an entry frame 302. The print character 301 represents, for example, an election name or a note, and the entry box 302 defines a description area in which a voter should write a predetermined item (for example, a candidate name or a party name). It is.

  The background color of the voting paper 3 is, for example, white, but may be another specific color. The printed matter is, for example, black-colored printing. If this is a color that can be distinguished from the ground color of the voting paper 3 in the first and second image information 175c and 176c, it is colored in a color other than black. May be.

  FIG. 4 shows a description surface of the ballot paper 3 in a state different from that in FIG. This ballot 3 is a completed ballot 32 filled in by the voter. On the description surface of the completed ballot paper 32, the description contents 303 by the voter are entered in addition to the print matters (print characters 301 and the entry frame 302) of FIG. The description content 303 is entered by the voter in the entry box 302 using a writing tool, and represents a predetermined item (for example, a candidate name or a party name) that the voter intends to designate. The description content 303 is colored by a writing tool in a color that can be distinguished from the ground color of the voting paper 3 in the first and second image information 175c and 176c.

The ballot paper reader according to the present embodiment can operate in a plurality of operation modes including a learning mode and a determination mode. The learning mode is a mode that is selected in correspondence with the blank ballot 31 of FIG. 3, and is a mode in which a mask region is determined by performing learning using the blank ballot 31. The determination mode is a mode that is selected corresponding to the completed ballot paper 32 of FIG. 4, and is a mode for determining whether or not there is any other description by the voter using the mask area.
Although not shown, input means capable of communicating with the information processing means 4 may be provided, and the user operates the input means to issue an instruction to the information processing means 4 so that the operation of the ballot paper reader is performed. The mode can be set as appropriate. Such an input means can be realized by a known configuration including buttons and the like.

FIG. 5 is a flowchart showing the flow of processing of the ballot paper reader in the learning mode. The user sets the operation mode of the ballot paper reader to the learning mode, and then puts the blank ballot 31 into the hopper 10.
First, the first and second imaging units 175 and 176 generate first and second image information 175c and 176c representing the corresponding surfaces in the blank blank sheet 31 (step S1).

  Next, the information processing means 4 identifies the front and back sides of the blank ballot 31 based on the first and second image information 175c and 176c from the first and second imaging means 175 and 176 (step S2). Here, the information processing means 4 is one of the first and second end faces 3a, 3b based on the first and second image information 175c, 176c from the first and second imaging means 175, 176. The entry frame 302 is detected from the surface. Then, the front and back sides are identified by determining from which of the first end surface 3a and the second end surface 3b the entry frame 302 is detected. As a result, it is identified which of the first and second image information 175c and 176c corresponds to the description surface of the unfilled voting paper 31.

In the following example, a case where the first image information 175c corresponds to a description surface will be described. The same applies to the case where the second image information 176c corresponds to the description surface.
The information processing means 4 identifies the top / bottom direction of the blank ballot 31 based on the first image information 175c (step S3). This is performed based on the orientation of the print character 301, for example. At this time, it is identified whether the direction of the blank ballot 31 is front-facing, front-facing, back-facing, or lining-facing.

  The information processing means 4 expands dots (colored dots) representing coloring in the first image information 175c, and determines a mask area (step S4). Here, since the first image information 175c is the image information of the unfilled voting paper 31, the coloring is only due to the printed matters (printed characters 301 and filled frames 302). Whether a certain dot is a dot (ground color dot) representing the background color of the unfilled voting paper 31 or a colored dot can be determined by a known technique using the color value of the dot, for example.

  FIG. 6 shows an example of a dot pattern 40 that explains the expansion of colored dots. The dot pattern 40 is a part of the first image information 175c. In this example, the dot pattern 40 is 11 dots wide × 14 dots long. The dot pattern 40 includes colored dots (shown using a grid-like tile pattern) and ground-color dots (shown using a plain tile pattern and a parallel diagonal tile pattern).

For all the colored dots, the dots located around the colored dots are defined as expansion dots. Colored dots and expanded dots are dots that constitute the mask region. “Ambient” can be defined, for example, as an area within 2 dots in each of up, down, left, and right. As an example, when looking at the colored dot 41 located at the top in FIG. 6 among the colored dots, all the dots within the range of the upper, lower, left, and right two dots from the colored dot 41 are all the expanded dots 41a (in the figure, the colored dots and Excluding overlapping ones).
A region constituted by the colored dots and the expanded dots is set as a mask region.

Further, the information processing means 4 adds an area formed by dots corresponding to the inside of the entry frame 302 to the mask area.
FIG. 7 shows an example of the mask region 50 configured as described above.
In the example of FIG. 6, the range of 2 dots in each of the upper, lower, left, and right directions is “periphery”, but the definition of the periphery may be different from this. For example, the sizes may be different (within 3 dots or the like), or the shapes may be different (circular etc.).

Returning to FIG. 5, next, the information processing means 4 stores the determined mask area in the storage unit (step S5). This can be stored, for example, as binary image format data indicating whether each dot in the image is a mask region or a non-mask region. Further, the mask area includes information indicating the vertical direction. For example, the direction in which the coordinates in one direction (for example, the vertical direction) of the image are smaller can be set to the sky.
In this way, the operation in the learning mode is completed.

  FIG. 8 is a flowchart showing a process flow of the ballot paper reader in the determination mode. The user sets the operation mode of the ballot paper reader to the determination mode, and then inserts the completed ballot paper 32 into the hopper 10. There may be a plurality of completed ballots 32, and the process of FIG.

First, as in the learning mode (FIG. 5), the first and second imaging units 175 and 176 generate first and second image information 175c and 176c representing the corresponding surfaces in the completed ballot paper 32, respectively ( Step S11).
Next, as in the learning mode (FIG. 5), the information processing means 4 identifies the front and back of the completed ballot paper 32 (step S12).
In the following example, a case where the first image information 175c corresponds to a description surface will be described. The same applies to the case where the second image information 176c corresponds to the description surface.

The information processing means 4 determines whether or not the description content 303 in the entry frame 302 detected in the first image information 175c can be read (step S13). If it is determined that it can be read, the description 303 is read (step S14), and the vertical direction of the completed ballot paper 32 is identified (step S15). The identification of the vertical direction is performed based on, for example, the direction of the print character 301 and the direction of the character in the description content 303. At this time, it is identified whether the direction of the completed voting paper 32 is the front side, the front side, the back side, or the back side.
Since the first image information 175c is the image information of the completed ballot paper 32, the colored dots represent the printed matters (printed characters 301 and entry frames 302) and the writing by the voter's writing tool (description contents) 303) are mixed.

  Next, the information processing means 4 applies a mask to the first image information 175c to remove some dots (step S16). That is, the data of the mask area (stored in step S5 in FIG. 5) is read from the storage unit, and the first image information 175c and the mask area are overlapped with the top-to-bottom direction aligned, and are included in the mask area. The dots are removed from the first image information 175c.

9 and 10 show a specific example of the process in step S16.
FIG. 9 shows an example of a completed ballot 32 that does not include other matters. FIG. 9A is a diagram in which the completed ballot 32 and the mask area 50 are overlapped, and FIG. 9B is a diagram in which the portion included in the mask area is deleted from FIG. 9A by the process of step S16. FIG. As shown in FIG. 9B, none of the print characters 301, the entry frame 302, and the description content 303 remain in the image obtained as a result of the process of step S16, and there are no colored dots.

  FIG. 10 shows an example of a completed ballot 33 including other matters. FIG. 10A is a diagram in which the completed ballot 33 and the mask area 50 are overlapped. In addition to the description content 303, the voter has entered a line 304 that may correspond to other matters described. FIG. 10B is a diagram in which the portion included in the mask area is deleted from FIG. 10A by the process of step S16. In the image obtained as a result of the processing in step S16, the print character 301, the entry frame 302, and the description content 303 do not remain, but the portion 304a of the line 304 that is not included in the mask area 50 remains. That is, colored dots remain. This portion 304a corresponds to other matters described.

  In this way, the printed characters 301 and the entry frame 302 that already exist on the unfilled ballot 31 and the description content 303 in the entry frame 302 can be appropriately masked by learning. On the other hand, writing that does not exist in the blank voting form 31 and that exists outside the entry frame 302, for example, a part 304a of the line 304, can be appropriately detected as other matters.

Returning to FIG. 8, next, the information processing means 4 determines whether or not there is other information (step S17). This determination is performed independently for each other description of the description side and other description of the back side, and when any side includes the other description, it is determined that the ballot 3 includes the other description. .
The determination about the writing surface is performed based on the colored dots included in the image obtained in step S16. The determination for the back surface is performed based on the colored dots included in the image information (second image information 176c in this example) generated in step S11. For example, if the number of colored dots is equal to or greater than a predetermined threshold value, it is determined that there is other matter description, otherwise it is determined that there is no other matter description. This threshold value may be appropriately changed by the user.

  When it is determined in step S17 that there is no other description, the ballot paper 3 is conveyed to the stackers 12 to 15 and 21 to 28 according to the description content 303 and the direction (step S18). Here, the information processing means 4 controls the operations of the main body side conveying means 11 and the expansion side conveying means 20, and the stackers 12 to 15 and 21 to 28, which differ depending on the description contents 303, of the voting paper 3 (see FIG. 1) To transport.

On the other hand, when it is determined that the description content 303 is read (step S13), for example, when it is determined that reading is impossible due to, for example, a character extraction failure, or when other matters are described (step S17). If it is determined, the ballot 3 is rejected as invalid or needing re-examination, and conveyed to the rejecter or the unidentifiable stacker 16 (step S19). That is, the information processing means 4 conveys the ballot paper 3 including the other event description to the stacker 16 different from the ballot paper 3 not including the other event description. In this embodiment, the fifth stacker 16 is assigned for rejection or indistinguishability. In this way, the information processing means 4 causes the voting sheets to be conveyed to different stackers depending on whether or not other matters are described.
In this way, the operation in the determination mode ends.

  As described above, according to the reading device for the ballot paper 3 according to the present embodiment, the information processing unit 4 expands the print items in the image information of the blank blank ballot 31 to display the mask area 50 in the learning mode. decide. Further, in the determination mode, the presence / absence of other information on the completed ballots 32 and 32 is determined based on coloring in areas other than the mask area. For this reason, the printed matter 301, the entry frame 302, and the description content 303 included in the mask area 50 can be excluded from the determination regarding other matters description, and the matters (other matters description) written outside the description region by the voter are excluded. It can be clearly identified and properly detected.

  In addition, since the ballot 3 including other matters and the ballot 3 not including other matters are conveyed to different stackers 16, the invalid vote should be checked simultaneously with the classification for each description 303 of the ballot 3. And the labor related to the management of the ballot 3 can be greatly reduced. In other words, in the conventional device, after classifying the voting paper 3 into the description contents 303, it was necessary to inspect whether there was any other description. However, this post-process can be simplified and the vote counting operation can be completed more quickly. Can be made.

  In the first embodiment described above, the first and second imaging means 175 and 176 are provided as the imaging means, but the imaging means may be independent. In this case, the user aligns the front and back of the completed ballot paper 32 and puts it into the hopper 10.

  3 ballot (3a first end face, 3b second end face, 31 blank voting paper, 32 and 33 filled voting paper), 4 information processing means, 111 first main body conveyance path, 112 second main body conveyance path, 40 dot pattern, 41 colored dot, 41a expansion dot, 50 mask area, 175c first image information, 176c second image information, 175 first imaging means, 176 second imaging means, 301 printed characters, 302 entry frame, 303 Matters, 303 description, 304 line, 304a line part (others description).

Claims (6)

A conveyance path through which voting paper is conveyed;
Imaging means disposed opposite the transport path;
Information processing means connected to the imaging means,
The imaging means generates image information representing a surface of the voting paper when the voting paper is conveyed along the conveyance path,
In the learning mode, the information processing means expands the colored dots representing the coloring in the image information representing the surface of the voting paper that is not filled in by the voter and does not include other matters, and determines the mask area And
The information processing means adds an area formed by dots corresponding to the inside of the entry frame to the mask area,
In the determination mode, the information processing means is a voting paper reading device that determines presence / absence of other matters based on colored dots other than the mask area in the image information.   2. The ballot paper reading device according to claim 1, wherein the expansion of the colored dots is performed by using dots positioned around the colored dots as a part of the mask region. 3.   3. The ballot paper reader according to claim 2, wherein the dots positioned around the colored dots are dots positioned within a predetermined range from the colored dots.   The voting paper reading apparatus according to claim 1, wherein the colored dots are dots representing printing on the voting paper or writing with a writing instrument. On the face of the voting form, an entry frame is printed that divides a predetermined description area,
The voting paper reading apparatus according to claim 1, wherein the mask area is further determined based on the predetermined description area.   A voting paper reading method using the voting paper reading device according to claim 1.

Images