JP4598799B2 - Form processing device with advertisement - Google Patents

Description

  The present invention relates to a form processing apparatus with an advertisement for processing a form in which an advertisement is placed on a paper such as a questionnaire.

  Forms such as questionnaires and test forms are distributed to the entrants, and answers to the questions are written and collected by the entrants by a method such as handwriting. In this case, if the advertisement is placed on the form, the form such as the questionnaire form or the test form will be viewed by the writer because the form person sees the entire page, so that the advertisement is highly effective. In addition, the client that conducts the questionnaire and test can be different from the client that places the advertisement, and the client that places the advertisement bears the advertising fee, so the burden on the client that conducts the questionnaire and test can be reduced. become.

In this case, when a form such as a questionnaire form or a test form is printed, it is conceivable to provide advertisement data to a client that conducts the questionnaire or test and print a form with an advertisement. However, there is a problem that an advertisement is placed on a form and cannot be printed, or an advertisement different from an advertisement designated to be placed may be placed and printed.
Therefore, an object of the present invention is to provide a form processing apparatus with advertisement that can confirm that an advertisement designated in a form is posted, printed, and distributed to a writer.

  In order to achieve the above object, the present invention analyzes image data obtained by converting a plurality of collected forms and detects whether or not an advertisement is placed on a form sheet. In this case, the main feature is that a process for detecting whether or not the advertisement image of the advertisement placed on the form paper matches the advertisement image designated to be placed is performed.

  The present invention analyzes image data obtained by converting a plurality of collected forms and detects whether or not an advertisement has been placed on a form sheet. When it is detected that an advertisement has been placed, Since the process of detecting whether the ad image of the displayed ad matches the ad image specified to be displayed, the specified ad is printed and printed on the form sheet. It will be possible to confirm that it has been distributed to the entrant.

FIG. 1 is a block diagram showing a configuration of an advertisement-added form processing system including an advertisement-added form processing apparatus according to an embodiment of the present invention.
In the form processing system with advertisement 1 shown in FIG. 1, the client that conducts the questionnaire is A, the client A has a word processor X and a printer V, and the word processor X is installed with a program Y for processing the form with advertisement. . The identification information (ID) of the client A on the service of the form processing system 1 is A1. The entrant who answers the questionnaire is B, and the advertising client who advertises the entrant B is C. For example, the advertisement client C creates an advertisement manuscript that fits within a frame of 10 cm long × 2 cm wide, and if the image data of the advertisement image U of the advertisement manuscript and the entry column are arranged on the advertisement image U, the entry is made Data for the input form UF and position information UF1 on the advertisement image U for reading the input form UF are provided in advance to the form processing apparatus D with advertisement. It is the form processing apparatus D with an advertisement concerning this invention that performs the form process with an advertisement in the form processing system 1 with an advertisement. The advertisement-added form processing apparatus D has a server Z including a database Z1, and reads an advertisement image U provided from one or more advertisement clients C and an input form UF arranged on the advertisement image U. The position information UF1 is registered in the server Z. The advertisement-added form processing apparatus D can generate a form identification symbol D1 that uniquely identifies a form for each different content created by the client A, and issues an identification sign D1 generated in response to a request from the client A can do.

  Here, the client A creates a questionnaire by the word processor X in which the program Y provided from the advertisement-added form processing apparatus D is installed. For example, a check box that can be read by image analysis processing is arranged in the questionnaire. By checking the check box by filling it in, the answer to the questionnaire question is made. When the creation of the questionnaire is completed and the word processor X is operated so as to print the questionnaire by the printer V, the program Y starts and the word processor X sends a request to the server Z via a communication network such as the Internet. This request includes the property information of the question content in the created questionnaire, and the server Z that received the request searches the advertisement registered in the database Z1 for an advertisement suitable for the property information, and the retrieved advertisement. The advertisement image U is designated and the image data and the identification information U1 are sent to the word processor X. The property information of the question content in the questionnaire may be input by the client A, or the program Y may be created by analyzing the words and sentences of the question sentence and the answer sentence in the questionnaire. When the word processor X receives the image data of the advertisement image U and the identification information U1, the program Y uses the function of the word processor X to force the advertisement image U to the blank area below the virtual paper of the questionnaire created by the client A. And paste. In this case, if the attachment position of the advertisement image U is inappropriate for printing, the printing routine is interrupted and the client A can place the advertisement image U at another position on the virtual paper surface. However, the program Y checks whether or not the advertisement image U is always placed at a location where browsing is possible on the virtual paper.

At the same time, the word processor X acquires the reference point image S from the server Z through the communication network via the program Y, and requests the form processing apparatus D with advertisement to generate the identification symbol D1. In response to this request, the form processing apparatus with advertisement D generates an identification symbol D1 for the created form and sends it to the word processor X. The word processor X performs a process for the word processor X to paste the reference point image S acquired from the server Z to at least one blank portion that is a predetermined position on the virtual paper surface, thereby providing a reference position on the virtual paper surface. . The reference point image S has a shape that can easily identify the position accurately by image recognition. Further, the program Y pastes the obtained identification symbol D1 generated by the advertisement-added form processing apparatus D on the blank portion of the virtual paper. The form identification symbol D1 is pasted on the virtual paper surface in the form of a character string H or barcode T. In this case, the client A can select whether the identification symbol D1 is displayed as the character string H or the barcode T.
On the virtual page of the questionnaire created in this way, the program Y uses the identification information U1 of the advertisement image U, the position information U2 of the advertisement image U based on the reference point position information S1 on the virtual page, and the attached form. Is transmitted to the server Z via the communication network. The server Z that has received these pieces of information stores the identification information U1, the position information U2, the form identification symbol D1 and the identification information A1 of the client A in association with the advertisement image U and the position information UF1 in the database Z1.

As described above, the questionnaire created by the program Y in the word processor X of the client A is used as the advertisement-added form Q to which the advertisement image U or the like is pasted, and is printed by the printer V. An example of the advertisement-added form Q is shown in FIG.
In the advertisement-added form Q shown in FIG. 2, the reference point images S that are double circles are printed at the four corners, and the area excluding the lower part of the page is the questionnaire area in which the questionnaire questions and the responses to the questionnaire are filled E. An answer to a question is a method of answering by filling a check box provided at the head of each item in a plurality of answers, and the check box has a white oval shape. The lower part of the form Q is an advertisement area F in which an advertisement is placed. The advertisement area F is printed with an advertisement U ′ pasted with an advertisement image U, and the writer who viewed the advertisement U ′ in the advertisement U ′. An input form UF filled in by B is printed. In the example shown in the drawing, an answer to the question printed in the advertisement U ′ is an input form UF for selecting an answer by filling an elliptical check box attached to the head of the answer. In addition, the character string H and the barcode T representing the form identification symbol D1 are printed on the right side of the advertisement U ′, but it is sufficient that either the character string H or the barcode T is printed. As shown in the figure, when both the character string H and the barcode T are printed, the identification symbol D1 of the form can be recognized even if either of them cannot be read due to dirt or the like.

  Such questionnaire forms Q are printed or copied and distributed to the entrant B. A plurality of respondents B answer the question in the questionnaire area E by filling the check box, and if the advertisement U ′ in the advertisement area F has a check box input form UF, the check box is also painted. It may answer by crushing. The entrant B returns the completed form Q ′ to the client A. The client A converts the collected multiple forms Q ′ into image data of each image. In this case, the form Q ′ is converted into image data by the scanner W using a personal computer (hereinafter referred to as “personal computer”) PC in which the program Y is installed. In this case, the personal computer PC may also serve as the word processor X. The page image group R including the converted data of the plurality of images r is transmitted from the client A to the server Z of the advertisement-added form processing apparatus D through the communication network by the program Y. The form Q ′ entered by the writer B may be converted into image data and sent to the server Z.

The server Z in the form processing apparatus with advertisement D that has received the sheet image group R recognizes the data of the image r for each sheet image group R, thereby position information S2 of the reference point image S on the image r. Ask for. Further, the character string H or barcode T is extracted from the image r, and the form identification symbol D1 is recognized from the extracted character string H or barcode T. The server Z searches the database Z1 using the recognized identification symbol D1 as a key, and uses the identification information U1 of the advertisement image U and the position information U2 of the advertisement image U, and the position information UF1 for reading the input form UF, if any. get. Further, the server Z searches the database Z1 using the identification information U1 of the advertisement image U as a key, and acquires the image data of the advertisement image U. At this time, the number of times the form Q on which the advertisement image U is posted for the same identification information U1 is collected and recorded in the database Z1, thereby counting the number of coarse advertisements issued for the advertisement image U corresponding to the identification information U1. Can be weighed.
Then, the server Z compares the position information S2 of the reference point image S on the image r and the reference point position information S1 on the virtual paper surface of the advertisement-added form Q, whereby the image r and the virtual paper face of the advertisement-added form Q are compared. Define the conversion parameters for the coordinates above. Next, the position information U2 of the advertisement image U is converted by the conversion parameter based on the reference point position information S1 on the virtual sheet of the advertisement-added form Q, and the advertisement image U and the image r on the virtual sheet of the form Q with advertisement are converted into the advertisement image U. A candidate advertisement image U3 which is a candidate for comparing whether or not the posted advertisement U ′ is the same is cut out from the image r.

  If the position information UF1 for reading the input form UF is present, the candidate advertisement image U3 cut out using the position information UF1 for reading the input form UF is subjected to image analysis, and the input form UF Information UF2 is acquired. In this case, for example, when the input form UF of the advertisement image U is a check box input form, the position information UF1 for reading the input form is the coordinate position of the check box on the advertisement image U, Using the position information UF1 for reading the input form UF, a check box portion is extracted from the candidate advertisement image U3, and it is determined whether or not the check box is filled. As a result, the information UF2 obtained from the input form is information indicating whether or not the check box is filled. The information UF2 acquired in this way is recorded in the database Z1 in association with the identification symbol D1 of the form generated and issued by the form processing apparatus D with advertisement by the server Z.

（１）式で表される相関係数の絶対値｜Ｒｃ｜が、最大値の「１」に近づくほど広告画像Ｕと候補広告画像Ｕ３とが類似する度合いが高くなる画像であることが示され、最小値の「０」に近づくほど類似する度合いが低くなることが示される（必要ならば、高木幹雄・下田陽久著「新編 画像解析ハンドブック」 東京大学出版会 ２００４年 ｐ．１５８７，ｐ．１６７１を参照されたい。）。このことから、相関係数の絶対値｜Ｒｃ｜を類似度Ｕ４とおくことができる。また、（１）式における関数θ（ｐ，α，ｑ，β）は次に示す（２）式で表され、θ（ｐ，α，ｑ，β）はα周りのＦに関するｐ次，βまわりのＧに関するｑ次の統計モーメントの輝度値数に関する平均値であることを表している。特に、θ（１，０，０，０），θ（０，０，１，０）は、それぞれ、輝度値Ｆ（ｉ，ｊ）、輝度値Ｇ（ｉ，ｊ）の平均値となる。

In order to detect whether or not the advertisement image U sent from the server Z of the advertisement-added form processing apparatus D to the client A is correctly displayed as the advertisement U ′ in the advertisement-added form Q, the server Z Similarity U4 of candidate advertisement image U3 cut out from image r is calculated. When the calculated similarity U4 satisfies a predetermined threshold, it can be considered that the advertisement image U is correctly printed on the sheet of the advertisement-added form Q and stored until the image r is acquired.
The similarity U4 is calculated by, for example, a method of calculating a correlation coefficient between the two images by looking at the correlation of luminance values of corresponding pixels of the advertisement image U and the candidate advertisement image U3 for all pixels. For simplicity, a gray scale image will be described, but a color image can also be converted into a gray scale image, or the similarity can be confirmed by applying the following test method for each of the three color component values. In calculating the correlation coefficient, the pixels in the region are excluded from the portion corresponding to the input form UF. Here, the number of pixels of the advertisement image U is M × N, and the grayscale luminance value of each pixel is expressed as F (i, j) as a function of the coordinate value. In addition, since the number of pixels of the candidate advertisement image U3 may be different from the number of pixels of the advertisement image U, for example, the bi-cubic interpolation method (if necessary, Mikio Takagi and Yoshihisa Shimoda “New Image Analysis Handbook” Tokyo University Press 2004 (p. 1366-1370)), the candidate advertisement image U3 is scaled up to M × N. The grayscale luminance value of each pixel of the enlarged / reduced candidate advertisement image U3 is expressed as G (i, j) as a function of the coordinate value. At this time, the absolute value | Rc | of the correlation coefficient between the advertisement image U and the candidate advertisement image U3 is expressed by the following equation (1).

It is shown that the degree of similarity between the advertisement image U and the candidate advertisement image U3 increases as the absolute value | Rc | of the correlation coefficient expressed by the equation (1) approaches the maximum value “1”. It is shown that the closer to the minimum value “0”, the lower the degree of similarity (if necessary, Mikio Takagi and Yoshihisa Shimoda “New Image Analysis Handbook”, University of Tokyo Press 2004, p. 1587, p. 1671). From this, the absolute value | Rc | of the correlation coefficient can be set as the similarity U4. Further, the function θ (p, α, q, β) in the equation (1) is expressed by the following equation (2), and θ (p, α, q, β) is a p-th order for β around α, β This represents an average value regarding the number of luminance values of the q-th order statistical moment regarding the surrounding G. In particular, θ (1, 0, 0, 0) and θ (0, 0, 1, 0) are average values of the luminance value F (i, j) and the luminance value G (i, j), respectively.

  When the calculated degree of similarity U4 satisfies a predetermined threshold, it can be considered that the advertisement image U is correctly placed in the advertisement-added form Q and collected. On the contrary, when the similarity U4 does not satisfy the predetermined threshold, it is considered that the advertisement image U was not correctly placed in the advertisement-added form Q. The server Z checks whether or not the similarity U4 acquired every time the image r is analyzed satisfies a predetermined threshold value, so that the number of times the advertisement image U is correctly placed in the form Q with advertisement or is not correctly placed. Is recorded in the database Z1. The predetermined threshold here refers to, for example, a test for analyzing the image r obtained by converting the advertisement-added form Q on which the advertisement image U has been printed in advance into image data of the image under various conditions such as paper stains, If it is normally printed, it is a value that will hardly fall below. The calculated similarity U4 is recorded in the database Z1 in association with the form identification symbol D1 generated and issued by the form processing apparatus D with advertisement and the identification information A1 of the client A.

  The server Z calculates and acquires statistical values U5 of a plurality of similarities U4 related to the client A obtained so far and recorded in the database Z1. The statistical value U5 is a value representing the reliability of the client A as an advertisement publisher. If the acquired statistical value U5 satisfies a predetermined threshold, the client A is considered to be highly reliable and the processing operation for analyzing the paper image group R is continued. On the contrary, when the statistical value U5 does not satisfy the predetermined threshold value, the client A has low reliability and the processing operation for analyzing the paper image group R is stopped. However, since a certain amount of similarity U4 information is required to obtain the statistical value U5, a significant result may not be obtained until a certain amount of similarity U4 information is collected. For this reason, it is preferable not to perform the process determination until a certain amount of similarity U4 information is collected. For example, the average value or median value of the similarity U4 related to the client A is used as the statistical value U5 calculated here. Further, it may be determined to continue the processing operation for analyzing the paper image group R using a plurality of statistical values.

Then, when it is determined from the calculation result of the server Z that the processing operation for analyzing the paper image group R in the form processing apparatus D with advertisement is determined, the server Z performs image processing on the check box area in the questionnaire area E of the image r. And the result of the questionnaire is totaled by the server Z. When the aggregation of the questionnaire is completed, the advertisement-added form processing apparatus D notifies the client A of an email to the effect that the questionnaire has been added from the advertisement-added form processing apparatus D. The client A can obtain the questionnaire count result by accessing the server Z of the advertisement-added form processing apparatus D. Further, the advertisement-added form processing apparatus D acquires from the database Z1 how many times the advertisement image U of the identification information U1 has been placed in the advertisement-added form Q, and depending on the number of times, the advertisement fee is sent to the advertisement client C. Make a claim. If there is information UF2 obtained from the input form UF in the advertisement area F in the advertisement-added form Q, the information UF2 is delivered to the advertisement client C.
If it is determined that the processing of analyzing the paper image group R in the form processing unit with advertisement D is to be stopped, the form processing unit with advertisement D sends an email to the client A indicating that the continuation has been stopped. Notice. The client A can receive the mail and pay the questionnaire analysis processing fee separately to allow the advertisement-added form processing apparatus D to continue the questionnaire counting process.

Next, FIG. 3 shows a configuration of the server Z included in the advertisement-added form processing apparatus D.
In the server Z shown in FIG. 3, the CPU 10 is a central processing unit that controls the overall operation of the server Z. The ROM 11 is an operation program such as a program related to a form processing with advertisement according to the present invention executed by the CPU 10 and a ROM (Read Only Memory) in which various data are stored. The RAM 12 is read from the ROM 11 or the storage device 13. A RAM (Random Access Memory) in which a storage area for temporarily storing various data such as the output form data and a work area of the CPU 10 is set. The storage device 13 is a large-capacity storage means having storage areas for various application programs and the database Z1, and includes an HDD (Hard Disk Drive), an FDD (Floppy (registered trademark) Disk Drive), a CD (Compact Disk) -ROM, An MO (Magneto Optical Disk), a DVD (Digital Versatile Disk), or the like can be used.

The I / F 14 is an interface that exchanges data with a peripheral device 15 such as a scanner or a printer provided outside. The communication interface 16 is a communication interface for connecting to another computer via a communication network 17 such as a LAN (Local Area Network) such as Ethernet (registered trademark), the Internet, or a telephone line. Data can be exchanged with the word processor X of the client A and the personal computer PC. The operation element 20 is an operation element such as a keyboard or a pointing device. The operation element information when the operation element 20 is operated is detected by the detection circuit 19 and sent to the bus 23. The display 22 is a display device such as a liquid crystal display on which an advertisement image U and an image of a virtual paper surface of the advertisement-added form Q are displayed. The display 22 is driven by the display circuit 21 to display a desired screen.
The hardware configuration of the server Z described above is the same as that of the personal computer PC. When the word processor X has the same hardware configuration and the word processor X is installed by installing a program for the word processor, the personal computer PC Thus, the word processor X can also be used.

Next, FIG. 4 shows a flowchart of the form processing with advertisement executed in the form processing system 1 with advertisement according to the present invention.
In the form process with advertisement shown in FIG. 4, the advertisement client C registers the advertisement image U that the client C wants to place on the form processing apparatus D with advertisement in step S10. The advertisement-added form processing apparatus D associates the form with the identification symbol D1 for uniquely identifying the form, and uses the database Z1 as the identification information U1 of the advertisement image U, the image data of the advertisement image U, and the position information UF1 for reading the input form UF. Save to. Next, the client A creates a questionnaire using the word processor X in step S11. At this time, the advertisement-added form processing apparatus D specifies the advertisement image U suitable for the property information of the questionnaire to the client A, and the identification information U1, the reference point image S and the form identification symbol D1 are received from the advertisement-added form processing apparatus D. Provided. Thereby, the virtual paper surface of the form Q with an advertisement used as the questionnaire on which the advertisement is posted is created. A character string H or a barcode T representing the identification symbol D1 that can uniquely recognize the form is also attached to the form Q with advertisement.

  Next, in step S12, the identification information U1, the position information U2 of the advertisement image U, and the identification symbol D1 are registered in the server Z of the advertisement-added form processing apparatus D for the created advertisement-added form Q, and the advertisement-added form Q is printed by the printer. Print by V. The printed advertisement-added form Q is distributed to the writer B in step S13 and filled in the questionnaire. The advertisement-added form Q 'entered by the writer B is collected and converted into image data by the scanner W in step S14. The sheet image group R composed of the converted images r of each form Q ′ is transmitted to the form processing apparatus D with advertisement. The server Z of the advertisement-added form processing apparatus D that has received the sheet image group R performs image analysis of each image r, and the advertisement image U designated to be displayed on the client A from the advertisement-added form processing apparatus D is correctly displayed. It is detected whether or not the advertisement U 'was posted in Q. Next, the advertisement-added form processing apparatus D reads the check box area in the questionnaire area E of the image r by image processing in step S16, totals the results of the questionnaire, and notifies the client A of the totalization. Further, in step S17, the advertisement-added form processing apparatus D acquires how many times the advertisement image U of the identification information U1 has been placed in the advertisement-added form Q, and in accordance with the number of times, the advertisement fee is paid to the advertisement client C. When there is information UF2 read from the input form UF in the advertisement area F, the information UF2 is delivered to the advertisement client C. Thereby, the form process with advertisement ends.

Next, a flowchart of the image analysis process executed in step S15 of the advertisement-added form process is shown in FIG.
The image analysis process is executed by the advertisement-added form processing apparatus D, and when the image analysis process is started, it is determined whether all the images r in the page image group R have been processed in step S20. Immediately after the image analysis process is started, the image r has not been processed. Therefore, it is determined No here, and the process proceeds to step S21, where one image r is acquired from the sheet image group R. Next, the position information S2 of the reference point image S on the image r and the character string H or bar code T representing the form identification symbol D1 are recognized in step S22 by image recognition from the acquired image r. Then, the database Z1 is searched based on the recognized identification symbol D1, and the identification information U1 of the advertisement image U and the position information U2 on the paper surface are acquired in step S23, and the image data of the advertisement image U is obtained from the identification information U1. Is obtained from the database Z1. Further, the fact that the advertisement image U corresponding to the identification information U1 is placed in the image r being processed, that is, the advertisement-added form Q is recorded in the database Z1 in step S24.

  Next, in the advertisement information extraction process in step S23, if position information UF1 for reading the input form UF in the advertisement image U is also acquired, it is determined as “present” in step S25, and the process proceeds to step S26. In step S26, the position information S2 of the reference point image S recognized in step S22 is compared with the reference point position information S1 on the virtual paper surface of the advertisement-added form Q obtained based on the identification symbol D1. Based on the position information U2 converted according to the result, the region of the candidate advertisement image U3 on which the advertisement is posted is cut out from the image r. Next, in step S27, the input form UF arranged on the advertisement image U is read based on the position information UF1 (image recognition), and the read information UF2 is associated with the identification symbol D1 and recorded in the database Z1. Here, it is determined whether or not the similarity U4 needs to be checked in step S28. If it is determined that the similarity needs to be checked, the process proceeds to step S30, where the similarity between the advertisement image U and the candidate advertisement image U3 is determined. The degree U4 is detected by, for example, a method of calculating the correlation coefficient between the two images by looking at the correlation of the luminance values of the corresponding pixels as described above, and the detection result is associated with the identification symbol D1. Is recorded in the database Z1 in step S30.

  Next, in step S31, it is determined whether or not the statistical value so far of the similarity U4 satisfies a predetermined threshold value. Here, each time the image r is analyzed, the statistical value U5 of the similarity U4 obtained is calculated to obtain the statistical value, and in this case, it is determined that the acquired statistical value U5 satisfies a predetermined threshold value. In this case, the process proceeds to step S32, where image recognition of the questionnaire area in the image r is performed, and a process of counting the questionnaire results is performed. If it is determined that the acquired statistical value U5 does not satisfy the predetermined threshold value, the process branches to step S33 to notify the client A that the image analysis process has been stopped by e-mail. finish. If it is determined in step S25 that the position information UF1 does not exist and if it is determined in step S29 that the similarity check is not required, the process jumps to step S32 and the above-described questionnaire totalization process is performed. Is called. Note that if it is determined in step S29 that the similarity check is not necessary, for example, an image analysis of the advertisement-added form Q created by the client A determined to be very reliable is performed. . When the questionnaire totalization process in step S32 is completed, the process returns to step S20, and the processes in steps S20 to S33 described above are repeated, so that the images r in the sheet image group R are sequentially processed. Then, when all the image analysis processing of the image r of the paper image group R is completed, YES is determined in step S20 and the image analysis processing is ended.

In the above-described report processing apparatus with advertisement, the advertisement-added form in which the advertisement is placed in the questionnaire is used. However, the present invention is not limited to this, and the form with the advertisement in which the advertisement is placed in the test or the like may be used.
In addition, regarding the calculation of the similarity U4 between the advertisement image U and the candidate advertisement image U3, the above description is made based on the correlation coefficient between the advertisement image U and the candidate advertisement image U3. However, the present invention is not limited to this. , Spearman's rank correlation coefficient, Kendall's tau (if necessary, see Numerical Recipes in C Japanese version 1993 Technical Review, p. 474-480), or the central moment vector of both image luminance values Similarity based on various pattern matching methods such as matching of distances and feature points (see New Image Analysis Handbook 2004, Tokyo University Press, Takagi & Shimoda p. 1723, p. 1686-1694, if necessary) The degree U4 may be calculated.
In addition, when the form identification symbol D1 is issued for each form Q with advertisement, the writer B can be specified, and a service suitable for each writer can be provided from the answer contents of the questionnaire. .

It is a block diagram which shows the structure of the form processing system with an advertisement provided with the form processing apparatus with an advertisement of the Example of this invention. It is a figure which shows the example of the form with an advertisement which the form processing apparatus with an advertisement of this invention processes. It is a figure which shows the structure of the server with which the form processing apparatus with an advertisement of this invention is provided. It is a flowchart of the form processing with an advertisement performed in the form processing system with an advertisement of this invention. It is a flowchart of the image analysis process performed in the form process with an advertisement which the form process system with an advertisement of this invention performs.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Form processing system with an advertisement, 10 CPU, 11 ROM, 12 RAM, 13 Storage device, 14 I / F, 15 Peripheral device, 16 Communication interface, 17 Communication network, 19 Detection circuit, 20 Operator, 21 Display circuit, 22 Display, 23 buses, A client, A1 identification information, B writer, C advertisement client, D advertisement form processing device, D1 identification symbol, Dc correlation coefficient, E questionnaire area, F advertisement area, H character string, PC PC , Q advertising form, R paper image group, S reference point image, S1 reference point position information, S2 position information, T barcode, U ′ advertisement, U advertisement image, U1 identification information, U2 position information, U3 candidate advertisement image , U4 similarity, U5 statistics, UF input form, UF1 position information, UF2 Information, V printer, W scanner, X word processor, Y program, Z server, Z1 database, r image

Claims (4)

A form processing apparatus with advertisement that processes image data obtained by converting a plurality of forms of advertisement with a form distributed to and collected from the form person,
A form for each different content having an entry area having a check box for the answerer to fill in an answer to the question and an advertisement area in which an advertisement designated in accordance with property information of the content of the question is placed Processing means for performing image processing for each region of the image data of the advertisement-added form on which an identification symbol for uniquely identifying
A database in which at least the identification information of the advertisement image, the position information of the advertisement image, and the identification symbol of the form are stored in a storage unit in association with a plurality of types of advertisement images that are candidates for placement in the advertisement area. ,
The processing means includes
An acquisition means for acquiring a corresponding advertisement image and position information of the advertisement image by searching the database based on the identification symbol of the form recognized from the image data of the advertisement-added form;
By reading the area of the check box in the entry area of the image data by image processing , tabulation processing means for obtaining the fill information entered by the writer and performing tabulation processing of answers to the question ;
Calculation means for calculating a similarity between the advertisement image acquired by the acquisition means and candidate advertisement image data cut out from the image data based on position information of the advertisement image;
When the similarity calculated by the calculating means is within a predetermined threshold, the advertisement image of the advertisement that has been posted on the form sheet is posted among the advertisement images stored in the database. and detection means for detecting that match the specified advertising image as,
The form processing apparatus with an advertisement characterized by comprising.   The advertisement-attached form processing according to claim 1, wherein an advertisement suitable for property information of the content of the question is searched from the database, and the searched advertisement is designated as an advertisement to be placed on the advertisement-attached form. apparatus. When there is a second entry area in which the writer enters an answer to the question about the advertisement in the advertisement area in the advertisement-added form, in the processing means , the acquisition means includes the acquisition form of the advertisement-added form. By searching the database based on the identification symbol of the form extracted from the image data, the position information of the second entry area is also obtained, and the position information of the second entry area obtained by the obtaining means The second entry information entered by the writer is obtained by performing image processing on the image data corresponding to the second entry area extracted from the image data based on the image data. Item 1. A form processing apparatus with advertisement according to Item 1. The similarity information obtained each time the processing means processes the image data obtained by converting the form of the advertisement-added form and the identification information of the client of the advertisement-added form are recorded in the database in association with each other. The processing means calculates and obtains a statistical value of the similarity associated with the client in a plurality of pieces of similarity information recorded in the database, and the obtained statistical value is attribute information of the client The form processing apparatus with an advertisement according to claim 1 .

Images