JP5034431B2 - Behavior counting system - Google Patents

Description

  The present invention relates to a system for measuring the advertising effect of advertising content presented to a person in a place where an unspecified number of persons pass, such as a store, a station, a plaza, and a stadium.

In a conventional system for measuring the advertising effectiveness, the number of people who have seen the advertising content is counted, and a judgment index for the advertising effectiveness is calculated based on the counted number of people. For example, the advertising effectiveness of certain advertising content is defined by the amount of content exposure, which is the product of the time during which the advertising content was displayed and the number of people who have seen it. It is used as the basis for calculation.
The quantified advertising effect is structured such that the advertiser pays the price according to the number of people the advertising content sees, and it is more objective than the case where there is no quantified advertising effect. It is a judgment index.
In the prior art for quantifying the advertising effect, for example, in Patent Document 1, by using an infrared sensor to detect the face of a person, the counting accuracy of the number of people who have seen the advertisement is increased, and the advertisement is A technique for calculating the effect is disclosed. Alternatively, the advertising effect can also be calculated by using a method of detecting the face area of a person from the image of the person imaged using the camera image and counting the number of persons based on the number of detected face areas.

Republished Patent International Publication No. WO01 / 072053 (page 27, FIG. 1)

Conventionally, when an advertisement touches a person's eyes, it is considered that the content of the advertisement has been transmitted to that person, and a certain percentage of the people who have received the advertisement act to visit the store or purchase products. It is expected to cause. However, with such a method, it is only possible to capture a large change in the entire person who saw the advertisement, such as comparing changes in the number of people before and after the advertisement. On the other hand, if you want to analyze in detail whether or not the person who saw the advertisement actually took the purchase behavior, it is necessary to conduct a follow-up survey for a large number of people, and it takes a lot of trouble, There was a problem that it was not actually possible to judge whether or not there was an advertising effect.
Furthermore, whether the person who saw the advertisement does not lead to the actual purchase behavior or whether there is a problem in the place and time for displaying the advertisement or the problem in the advertisement content is captured by the macro of the exposure amount of the advertisement. However, there was a problem that it was impossible to know.

  The present invention has been made to solve the above-described problems. The present invention counts the rate at which a person who has seen an advertisement is making a purchase action, and further, To provide a behavior counting system that can evaluate advertising effectiveness in more detail and perform cost-effective advertising by acquiring detailed information every hour and making it available for purchase behavior analysis It is an object.

The action counting system according to the present invention is composed of an electronic advertisement screen and an advertisement effect evaluation device, each having the following means.
First, in the electronic advertisement screen, the advertisement display means for displaying the contents of the advertisement, and the display on the advertisement display means based on the display control information constituted by the contents of the advertisement and the display time of the contents of the advertisement. Display control means for controlling the content of the advertisement and outputting the display control information when the advertisement is displayed on the advertisement display means; and the face of the person who is facing the advertisement displayed on the advertisement display means A first face feature amount extracting means for detecting an image and extracting a first face feature amount from the face image; the display control information output by the display control means; and the face feature amount extracting means. In addition, an advertisement visual information generating unit that generates advertisement visual information that associates the first facial feature quantity with the first facial feature amount, and a transmission unit that transmits the advertisement visual information via a network are provided.
Next, the advertisement effect evaluation device detects a face image of a person installed in the store advertised by the advertisement and visiting the store, and extracts a second face feature amount from the face image. Facial feature amount extraction means, store visit information generation means for generating store visit information that associates the store information with the second face feature amount, and the advertisement visual information transmitted by the transmission means. One face feature amount and the second face feature amount of the store visit information generated by the store visit information generation means, and the first face feature amount that matches the second face feature amount First advertisement viewing information managing means for outputting the advertisement viewing information as advertisement effect information, and acquiring the advertisement effect information output by the first advertisement viewing information managing means, and directing the face to the advertisement Visit the store from among those who are People counting was, and a advertising effect calculation means for calculating an index of advertising effectiveness based on a count result.

  According to the present invention, it is possible to count the number of people who may have seen the advertisement and to count the number of people who may have caused purchase behavior by the advertisement. By calculating the ratio to the whole number, it becomes possible to analyze and evaluate the advertisement effect in more detail.

Embodiment 1 FIG.
FIG. 1 is a configuration diagram of an action counting system according to the first embodiment.
In FIG. 1, reference numeral 100 denotes an electronic advertisement screen, which displays advertisement content on a display. Reference numeral 101 denotes an advertisement effect evaluation apparatus which is installed in a store advertised by an advertisement displayed on the electronic advertisement screen 100 or a store selling a advertised product and displayed on the electronic advertisement screen 100. Of the people who may have seen the advertisement, the number of people who visited the store actually advertised by the advertisement is counted to calculate a judgment index of the advertisement effect. Reference numeral 7 denotes a network, which is an information network such as the Internet or a dedicated line that connects the electronic advertisement screen 100 and the store 101 and can exchange information with each other.

  In the electronic advertisement screen 100, reference numeral 1 denotes a passerby, who indicates a person who has passed in front of the advertisement display means 2, such as a liquid crystal display, a plasma display, or a rear projection display. 3 is a display control means for controlling the schedule management such as the name of the advertisement content to be displayed and the start time and end time of the display so that the advertisement content is displayed on the advertisement display means 2 and an advertisement visual information generating means to be described later 5, control information relating to schedule management such as a time zone including the displayed advertisement content name and the displayed start time and end time is output. Reference numeral 4 denotes face feature amount extraction means, which detects the face area of the passerby 1 from the captured image by the camera and the image processing device, and extracts the face feature amount from the detected face area. Reference numeral 5 denotes advertisement visual information generation means, which processes information obtained by the display control means 3 and the face feature quantity extraction means 4 to display the face feature quantity, advertisement content name, and display of a person who has watched certain advertisement contents. Information such as a time zone is associated and generated as advertisement viewing information. Reference numeral 6 denotes a transmission means, which is a communication device such as a LAN (Local Area Network) that transmits information via the network 7.

  In the advertisement effect evaluation apparatus 101, reference numeral 8 denotes transmission means, which is the same means as the transmission means 6 provided in the electronic advertisement screen 100. Reference numeral 9 denotes advertisement viewing information management means, which accumulates advertisement viewing information sent from the electronic advertisement screen 100 and constructs a database of people who viewed the advertisement. Also, advertisement visual information having a face feature amount that matches the face feature amount that has been inquired is retrieved and output. Reference numeral 13 denotes a visitor who shows a person who has visited the store 101. Reference numeral 12 denotes a face feature amount extraction means, which is the same means as the face feature amount extraction means 4 provided in the electronic advertisement screen 100 and extracts the face feature amount of the visitor 13. Reference numeral 11 denotes store visit information generation means, which generates store visit information by associating the face feature amount of the visitor 13 extracted by the face feature amount extraction means 12 with information such as the store name and product name. Reference numeral 10 denotes an advertisement effect calculation unit, which inquires of the advertisement viewing information management unit 9 about the store visit information passed from the store visit information generation unit 11 and whether or not the advertisement content related to the store is viewed on the electronic advertisement screen 100. To calculate the advertising effectiveness.

  Next, the operation of the behavior counting system in the first embodiment will be described with reference to FIGS. 1 and 2 as appropriate. FIG. 2 is a flowchart of the operation of the behavior counting system in the first embodiment.

First, in step S201, the display control means 3 creates display control information related to schedule management such as the name of the advertisement content to be displayed and the display start time and end time. The display control information is control command information for controlling the display of the advertisement display means 2, and an advertisement display that is set in advance after considering beforehand the contents, location, time zone, etc. of the advertisement before performing the advertisement. Based on a calendar. For example, the advertising display schedule determines the advertising content according to the result of analyzing the consumer image of the advertising target, and sets the location and time zone where the advertising target consumer is likely to go through and expect the advertising effect It is a thing. The display control means 3 creates display control information based on the advertisement display schedule set in this way, and controls the name of the advertisement content displayed on the advertisement display means 2, the display start time and end time, and the like. .
At the same time, the display control information is output to the advertisement visual information generating unit 5. The display control information output here is the content of the advertisement when displayed on the advertisement display means 2 and the display time of the content of this advertisement.

  Next, in step S <b> 202, the advertisement display unit 2 displays the specified advertisement content in the specified time zone according to the display control information sent from the display control unit 3. The advertisement display means 2 is a liquid crystal display, a plasma display, a rear projection display, or the like, but may be any means as long as it can present video and audio, and is not limited to the above display.

  Next, in step S <b> 203, the advertisement visual information generation unit 5 first obtains display control information related to the advertisement content schedule management from the display control unit 3. Next, the face of the passerby 1 watching the advertisement display means 2 displayed in a certain time zone based on the advertising content name included in the display control information and the displayed time zone information In order to cause the face feature amount extraction unit 4 to extract the feature amount, an extraction execution command is issued to the face feature amount extraction unit 4.

Next, in step S <b> 204, the face feature amount extraction unit 4 captures the passerby 1 who is looking at the advertisement display unit 2 with a camera in accordance with the extraction execution command from the advertisement visual information generation unit 5. An image area corresponding to a face is detected as a face area, and a face feature amount is extracted from the detected face area. At this time, the passerby 1 from which facial feature amounts are extracted may be one person or a plurality of passersby.
As the face feature amount, any of various feature amounts used in known face recognition technology and face authentication technology may be used, or a plurality of feature amounts may be used in combination. Further, by using a set of coefficient information calculated in the middle of the face detection process for detecting the face area, the processing amount in the face feature amount extraction process, which is a subsequent process, can be reduced.

  Next, in step S205, the advertisement visual information generating unit 5 groups the face feature amounts of the passerby 1 extracted by the face feature amount extracting unit 4 in association with the displayed time zone and the advertisement content name, It is generated as advertisement visual recognition information including the identification number of the electronic advertisement screen 100.

  Next, in step S206, the transmission means 6 sends the advertisement visual information generated by the advertisement visual information generation means 5 to the store advertised by the advertising content via the network 7.

Next, in step S207, at the store side that has received the advertisement visual information about the own store by the transmission means 8, the advertisement visual information is stored in the database of the advertisement visual information management means 9 in chronological order. As will be described later, the accumulated advertisement viewing information is searched in response to an inquiry from the advertisement effect calculation unit 10, and the advertisement effect information as a search result is output to the advertisement effect calculation unit 10.
When a plurality of electronic advertisement screens 100 are installed, the advertisement viewing information is stored in the database of the advertisement viewing information management unit 8 separately for each electronic advertisement screen 100. At this time, instead of dividing the advertisement viewing information for each electronic advertisement screen 100, a method of storing the identification information of the electronic advertisement screen 100 in association with the advertisement viewing information may be used.

  Next, in step S208, first, the store visit information generation unit 11 issues an extraction execution command to the face feature amount extraction unit 12 when the store visitor 13 is detected, and determines the face feature amount of the store visitor 13. Let it be extracted. Next, the store visit information generation means 11 generates store visit information for each visitor 13 by associating the extracted facial feature quantity of the store visitor 13 with store information such as the store name and product name, and the advertisement effect. Store visit information is sent to the calculation means 10.

  Next, in step S209, the advertisement effect calculation means 10 inquires the advertisement visual information management means 9 about the face feature amount associated with the store visit information of the store visitor 13. At this time, in addition to the facial feature amount, accompanying information such as the location where the electronic advertising screen 100 is installed and the display time zone are added to the inquiry command, and the facial feature in the advertisement visual information management means 8 is added. It is also possible to limit the amount of processing for face feature amount matching by limiting the advertisement viewing information to be subject to amount matching.

  Next, in step S210, when the advertisement viewing information management means 9 receives an inquiry about the store visit information from the advertisement effect calculation means 10, the advertisement view information management means 9 searches the stored advertisement view information and visits associated with the store visit information. The face feature amount of the person 13 and the face feature amount of the passerby 1 associated with the accumulated advertisement viewing information are compared and collated, and the degree of coincidence between the face feature amounts is calculated. Next, when the highest facial feature amount in the calculated degree of coincidence exceeds a predetermined threshold, it is determined that the visitor 13 and the passerby 1 are the same person, and the facial feature amount of the passerby 1 The advertisement viewing information associated with is determined and output to the advertisement effect calculation means 9 as advertisement effect information. If the highest facial feature quantity in the calculated degree of coincidence does not exceed a predetermined threshold, it is determined that there is no passerby 1 who is the same person as the store visitor 13, and a notification that it is not applicable is calculated as an advertising effect Output to means 10.

  Thus, when it is judged that the store visitor 13 is the same as the passerby 1 in the advertisement viewing information stored in the advertisement viewing information management means 9, the visitor 13 and the passerby 1 who has watched the advertisement content The association between the two is now complete. By determining the presence or absence of this relevance for all the store visitors 13 who have been able to generate store visit information, it becomes possible to know whether or not the display of the advertisement content has led to the purchase behavior of an actual store visit. The effect can be calculated.

  Finally, in step S211, the advertisement effect calculation means 10 acquires the advertisement effect information from the advertisement visual information management means 9, and counts the number of visitors 13 that are considered to match the passerby 1 who viewed the advertisement content. Then, an advertisement effect index is calculated based on the counting result. As an index of the advertising effect, a ratio with the total number of people who viewed the advertising content, a ratio with the total number of visitors to the store advertised with the advertising content, and the like can be calculated.

  The advertisement viewing information determined to be advertisement effect information includes information on the identification number of the electronic advertisement screen 100 corresponding to the place where the advertisement is made, in addition to the advertisement content name and the time period during which the advertisement content is displayed. Since they are associated with each other, it is possible to analyze in detail which advertising content has an advertising effect at what time zone and in which place.

  Based on the analysis result, the advertising effect of the electronic advertising screen 100 installed in various places is compared, and the advertising effect is obtained for each of a plurality of advertising contents, for each displayed time zone, and for each displayed place. In addition, it is possible to formulate an advertising strategy that enables more effective advertising by analyzing information in detail and determining information such as the advertising content that has a higher advertising effect, the displayed time zone, and the displayed location. it can.

  As described above, according to the first embodiment, it is possible to grasp the number of people who have visited the store by seeing the advertisement content displayed on the electronic advertisement screen 100. By counting the number of visits to the store at a certain rate, and further acquiring detailed information for each place and time where the advertisement is performed and making it available for analysis of purchase behavior, advertising You will be able to analyze and evaluate the effectiveness and perform cost-effective advertising.

  In addition, when a plurality of electronic advertisement screens 100 are installed, the advertisement effect of each screen can be individually evaluated, so that a more effective place can be specified.

  Furthermore, the advertising effect for each of a plurality of advertising contents, for each displayed time zone, and for each displayed place can be analyzed and evaluated in detail.

Embodiment 2. FIG.
As described above, in the first embodiment, the configuration in which the electronic advertisement screen 100 directly transfers the advertisement viewing information to the store is shown. Next, the advertisement viewing information management means 9 is connected to the network 7. Embodiment 2 having a configuration on the server 102 is shown.

FIG. 3 is a configuration diagram of the behavior counting system in the second embodiment.
In FIG. 3, the server 102 includes transmission means 14 and advertisement visual information management means 9. Only one server 102 may be configured in the entire behavior counting system, or the server 102 may be installed at each advertiser store. Further, the behavior counting system and the store may each be constituted by a plurality of servers 102. On the other hand, in the store, the advertisement effect evaluation device 103 includes a transmission unit 8, an advertisement effect calculation unit 9, a store visit information generation unit 11, and a face feature amount extraction unit 12.

  The advertisement viewing information generated on the electronic advertisement screen 100 is sent to the server 102 via the network 7. In the server 102, the advertisement visual information management means 9 accumulates the advertisement visual information for each time zone and advertising content to construct a database.

  On the other hand, in the advertisement effect evaluation apparatus 103, the store visit information generating unit 11 generates store visit information of the store visitor 13 and sends it to the advertisement effect calculating unit 10. Next, the advertisement effect calculation means 10 inquires of the advertisement viewing information management means 9 about the store visit information of the store visitor 13 via the network 7 and stores the advertisement viewing information stored in the advertisement viewing information management means 9. Thus, the advertisement viewing information of the passerby 1 determined to be the same as the visitor 13 is acquired, and the advertisement effect is calculated in the same manner as in the first embodiment.

  As described above, according to the second embodiment, advertisement viewing information relating to a plurality of stores can be centrally managed by the server 102, so that the entire advertisements of various stores can be analyzed, and the consumer It is possible to grasp the cross-sectional trends of the best-selling products from the whole and the advertising target layer for each place and time zone. Thereby, in addition to the advertising effect of the own store, the preference tendency of the entire consumer can be distributed to each store, and the advertising effect of the own store can be analyzed and evaluated from a wider field of view. .

  Furthermore, since a large amount of advertisement viewing information can be centrally managed by the server 102, the apparatus configuration on the store side can be simplified, and the cost of the advertisement effect evaluation apparatus 103 can be reduced.

Embodiment 3 FIG.
As described above, in the first and second embodiments, the case where the store visitor 13 is associated with the advertisement viewing information that is the viewing information of the advertisement content has been described. Embodiment 3 regarding an action counting system in the case of associating with advertisement visual information that is viewing information of advertisement content will be described.

FIG. 4 is a configuration diagram of the behavior counting system in the third embodiment.
In FIG. 4, the advertisement effect evaluation device 104 includes a transmission unit 8, an advertisement visual information management unit 9, an advertisement effect calculation unit 10, a store visit information generation unit 11, and a facial feature amount similar to the configuration in the first embodiment. In addition to the extracting means 12, a sales information collecting means 15 is newly provided.

  The sales information collecting means 15 is connected to a cash register used in a store, a POS (Point Of Sales) system, and the like, and collects and acquires information on products purchased by customers before the cash register. Further, the sales information collecting unit 15 transmits the acquired product information and information on the time zone when the product was purchased to the store visit information generating unit 11 as purchased product information. At this time, when a plurality of products are purchased, the purchased product information is set in a unit of the plurality of purchased products.

  Next, when the store visit information generation means 11 receives the purchase product information from the purchase product information acquisition means 15, the store visitor regards the store visitor as a purchaser, controls the facial feature quantity extraction means 12, and visits the store as a purchaser. The facial feature amount of the person 13 is extracted and acquired. At this time, the camera used in the face feature amount extraction unit 12 is installed, for example, around the sales information collection unit 16 so that the face of the visitor 13 as a purchaser falls within the imaging range of the camera. Further, the store visit information generation unit 11 associates the purchase product information acquired by the purchase product information acquisition unit 15 with the face feature amount of the visitor 13 extracted and acquired by the face feature amount extraction unit 12 to obtain purchase information. Generate and send purchase information to the advertising effect calculation means 10.

  Next, based on the purchase information received from the store visit information generation unit 11, the advertisement effect calculation unit 10 inquires the advertisement visual information management unit 9 about the facial feature amount of the store visitor 13 as a purchaser. The operation of the advertisement visual information management means 9 at this time is the same as in the case of the first embodiment, and the advertisement visual information whose facial feature quantity matches the store visitor 13 is output to the advertising effect calculation means 9 as the advertising effect information. To do.

  Next, the advertisement effect calculation means 10 counts the number of visitors 13 who are buyers who coincide with the passers-by 1 who viewed the advertisement content, based on the advertisement effect information acquired from the advertisement visual information management means 9. Based on the counting result, an advertisement effect index is calculated. In this case, as an index of the advertising effect, a ratio between the number of shoppers 13 who are purchasers and the total number of people who have viewed the advertising content is used. Alternatively, the ratio between the number of shoppers 13 who are purchasers and the total number of shoppers advertised in the advertising content is used. Further, the above advertising effect index can be calculated for each purchased product.

  As described above, according to the third embodiment, since the advertising content and the purchasing behavior are associated with each other at the product level, among those who have actually viewed the advertising content and have visited the store, Since it is possible to count even the number of people who have purchased a specific product, it is possible to analyze and evaluate the advertising effect in more detail.

Embodiment 4 FIG.
From the first embodiment to the third embodiment, the advertisement viewing information accumulated in the advertisement viewing information management means 9 increases with the passage of time. Next, the advertisement viewing information accumulated is enlarged. Embodiment 4 regarding the action counting system in the case of realizing the improvement of the efficiency of the calculation of the advertisement effect by preventing the conversion will be described.

  In the fourth embodiment, in consideration of the fact that the influence of the advertisement content on the consumer becomes smaller with the passage of time, the time that the influence of the advertisement content is expected to last is set to be finite, and the advertisement viewing information Set the validity period. The effective period may be set by the advertisement viewing information management means 9 storing a predetermined effective period together with each advertisement viewing information, or the importance of the advertising content input by the advertiser and the display time. Accordingly, it may be automatically set for each advertising content. The advertisement visual information management means 9 periodically checks the expiration of the effective period for the effective period set in this way, and deletes the advertisement visual information that has passed the effective period.

  As a result, among the advertisement viewing information accumulated in the advertisement viewing information management means 9, old advertisement viewing information that has exceeded the effective period is deleted, so the scale of the storage device necessary for storing advertisement viewing information is reduced. Can be limited.

  Further, in the advertisement visual information management means 9, in consideration of the distance between the electronic advertisement screen 100 and the store, the time required for the passerby 1 who viewed the advertisement content to move to the store is estimated as the assumed travel time. The advertisement viewing information used for calculating the advertisement effect index is selected using the estimated travel time. For example, the expected travel time is classified as less than 30 minutes, 30 minutes or more, less than 1 hour, 1 hour or more, but less than 2 hours, etc., and the advertising effect is to be analyzed for stores located near the electronic advertisement screen 100 In this case, only the advertisement viewing information having an estimated travel time of less than 30 minutes is selected and used for calculating the advertisement effect index. Alternatively, an effective period corresponding to the estimated travel time for each store may be set, and the advertisement visual information that greatly exceeds the estimated travel time may be deleted from the advertisement visual information management means 9.

  This makes it possible to calculate the index of advertising effectiveness by focusing on the advertisement viewing information that matches the analysis target, and to delete old advertisement viewing information that has passed the effective period. And a substantial accuracy of determining the advertisement effect can be maintained.

  Furthermore, the actual travel time is calculated from the difference between the time when the facial feature quantity is extracted between the electronic advertisement screen 100 and the store, and the advertisement effect information is only obtained when the actual travel time is within a predetermined range of the estimated travel time. Is used to calculate the advertising effectiveness index. First, the advertisement visual information generating unit 9 adds the time when the face feature amount of the passerby 1 is extracted as the first time to the advertisement visual information. On the other hand, the store visit information generating means 11 adds the time when the facial feature amount of the store visitor 13 is extracted as the second time to the store visit information. Next, the advertisement effect calculation means 10 obtains the actual travel time from the difference between the first time added to the advertisement viewing information and the second time added to the store visit information, and this actual travel time is assumed to travel. When the time is within a predetermined range, the advertisement viewing information is used for calculating the advertisement effect index. Alternatively, an effective period corresponding to the estimated travel time for each store may be set, and advertisement visual information that is not within a predetermined range of the actual travel time may be deleted from the advertisement visual information management means 9.

  This eliminates advertisement viewing information for actual travel times that deviate significantly from the estimated travel time, thus improving the efficiency of facial feature amount matching processing and maintaining substantial accuracy in determining ad effectiveness. Will be able to.

  As described above, according to the fourth embodiment, the effective period is set in the advertisement viewing information, and the advertisement viewing information is selected using the estimated travel time for the passerby 1 who has viewed the advertisement content to move to the store. In addition, since the advertisement effect information when the actual travel time from the electronic advertisement screen 100 to the store greatly deviates from the estimated travel time is excluded, the scale of the storage device necessary for accumulating advertisement viewing information is reduced. In addition, the calculation process of the degree of coincidence of the facial feature amounts can be made more efficient, and the substantial accuracy of determining the advertisement effect can be maintained.

1 is a configuration diagram of an action counting system in Embodiment 1. FIG. 3 is a flowchart of the operation of the behavior counting system in the first embodiment. It is a block diagram of the action counting system in Embodiment 2. FIG. 10 is a configuration diagram of an action counting system in a third embodiment.

Explanation of symbols

1 passer-by, 2 advertisement display means, 3 display control means, 4 facial feature quantity extraction means, 5 advertisement visual information generation means, 6 transmission means, 7 network, 8 transmission means, 9 advertisement visual information management means, 10 advertisement effect calculation Means 11 visit information generation means 12 face feature amount extraction means 13 visitor 14 transmission means 15 sales information collection means 100 electronic advertisement screen 101 advertisement effect evaluation apparatus 102 server 103 advertisement effect evaluation apparatus 104 Advertising effect evaluation device.

Claims (5)

An advertisement display means for displaying the contents of the advertisement;
The content of the advertisement displayed on the advertisement display means is controlled based on the display control information configured by the content of the advertisement and the display time of the advertisement content, and the advertisement is displayed on the advertisement display means. Display control means for outputting the display control information;
First face feature amount extraction means for detecting a face image of a person who is facing the advertisement displayed on the advertisement display means, and extracting a first face feature amount from the face image;
Advertisement visual information generation means for generating advertisement visual information that associates the display control information output by the display control means with the first face feature quantity extracted by the face feature quantity extraction means;
An electronic advertisement screen having transmission means for transmitting the advertisement viewing information via a network;
A second facial feature quantity extraction unit that is installed in a store advertised by the advertisement, detects a facial image of a person who has visited the store, and extracts a second facial feature quantity from the facial image;
Store visit information generating means for generating store visit information that associates the store information with the second facial feature amount;
Comparing the first facial feature quantity of the advertisement visual information transmitted by the transmission means with the second facial feature quantity of the store visit information generated by the store visit information generating means, and advertisement viewing information management means you output the advertisement viewing information as the advertisement effect information having the first facial feature quantity that matches the face feature quantity,
To obtain the advertisement effect information outputted by the previous Kihiro tell visual information management unit, the ad counts the person who visits the store from among the persons are directed face, advertisement effect based on the counting result An advertising effectiveness evaluation device having advertising effectiveness calculation means for calculating an index of
Action counting system comprising the. The advertising effectiveness evaluation device includes sales information collection means for collecting product sales information of the store,
The store visit information generating means generates purchase information in which the store visit information and the product sales information are associated with each other,
The advertisement effect calculating means acquires the advertisement viewing information having the first face feature amount that matches the second face feature amount included in the store visit information of the purchase information, and directs the face to the advertisement. 2. The behavior counting system according to claim 1, wherein a person who visits the store and purchases merchandise is counted from among the persons who are present, and an index of the advertising effect is calculated based on the counting result. Before Kihiro tell visual information management means, wherein the setting the validity period the advertisement viewing information, claim 1, characterized in that to remove the advertisement viewing information has passed the expiration time to claim 2 The behavior counting system described. Before Kihiro tell visual information management means, the electronic advertisement screen and on the basis of ad store distance between the store and calculates an assumed travel time a person moves the distance between the advertisement store the assumed travel time and selecting the advertisement viewing information used for calculating the index of the advertisement effect depending on the length, the behavior of any one of claims 1 to 3, and outputs the advertisement effect calculating means Counting system. The advertisement visual information generation means adds the time when the first facial feature amount is extracted as the first time to the advertisement visual information,
The store visit information generating means adds the time when the second facial feature amount is extracted to the store visit information as a second time,
The advertisement effect calculating means, said first time added to the advertisement viewing information outputted by the previous Kihiro tell visual information management means, added to the visit information generated by the visit information generating means Obtaining the actual travel time from the difference from the second time, and using the advertisement effect information for calculating the advertisement effect index when the actual travel time is within a predetermined range of the estimated travel time. The behavior counting system according to claim 4 , wherein the behavior counting system is characterized in that:

Images