WO2007125285A1

WO2007125285A1 - System and method for targeting information

Info

Publication number: WO2007125285A1
Application number: PCT/GB2007/001459
Authority: WO
Inventors: David Cumming; Stefan Veeser; Katherine Goldie
Original assignee: David Cumming; Stefan Veeser; Katherine Goldie
Priority date: 2006-04-21
Filing date: 2007-04-23
Publication date: 2007-11-08
Also published as: GB0607875D0

Abstract

A system and method allow targeting of information, in a way that allows an advertiser to reach a target audience more effectively. A camera captures an image of a person or an audience viewing an advert, and the image is analysed to determine physical characteristics of the person or audience. The most appropriate adverts for that person or audience are then presented, based upon the physical characteristics of the person or audience determined by analysis of the image.

Description

System and Method for Targeting Information

The present invention relates to a system and method for targeting information and in particular to a system and method that allows an advertiser to reach a target audience more effectively.

The range of media through which an advertiser can present information to a potential customer has increased greatly in the last 20 years. As a result it has become more important to ensure that advertisements reach their target audience. Targeting of this type is easier in the case of television advertising because there is a lot of research carried out into the type of person who is likely to be watching a particular type of program e.g. sport; this allows an advertiser to select an advert for transmission during a particular program that is most likely to be watched by the target audience.

However, a large number of advertisements are carried on digital signage, billboards, and other media, which can be situated stand alone or can be incorporated into bus shelters, on the side of buildings or in any location where they are viewable by the public.

These types of advertisements normally comprise either single static images that are changed on a regular schedule that might vary, depending on the installation, from minute to minute or perhaps on a weekly or monthly basis. Digital signage that incorporates audio- visual assets are also common. These may run a continuous loop or be activated by a movement detector or other sensor that detects the proximity of a person.

However, there is little scope for targeting this type of advertisement to a particular social group or individual . For example, normally adverts for younger women are shown to a viewing public of men and women, young and old. Consequently, the advert may be the wrong advert for 75% of the audience it eventually reaches. The intended target group of young women, in this case, would only be a 25% portion of the actual audience reached.

Digital signage advertising in a bus shelter is one example of the type of information that could be delivered more efficiently, if it were possible to more accurately identify the target audience.

In addition, devices including information kiosks such as those used to provide tourist information, ATMs and point of sale devices (POS) often carry advertisements that are displayed during their use. Digital signage is becoming ubiquitous in the urban environment, with screens in many different areas, in shops, in cinema foyers, in post offices, supermarkets, and shopping malls. Information may also be delivered in the form of voice, music and other sensory modalities including olfactory release, and tactile response.

It is an object of the present invention to provide for the more efficient targeting of information. It is also an object of the present invention to enable advertisers to more efficiently focus their advertising on their target audience.

In accordance with a first aspect of the invention there is provided a system for supplying targeted information, the system comprising: a sensor for generating sensor data which defines predetermined physical characteristics of a subject at or near the sensor location; sensor data analysis means for determining the physical characteristics of the one or more person in the image; analysis means for analyzing the physical characteristics of the one or more person to target the supply of information from an information source based upon the physical characteristics; and an information output.

Preferably, the analysis means further comprises a content controller which controls the current output of the information output.

Preferably, the content controller is updatable in response to a change in the image of the one or more person as determined by the image analysis means. Preferably, the analysis means analyses more than one person and supplies the most suitable information based upon the physical characteristics of each person.

Preferably, the sensor data analysis means comprises image analysis means.

Preferably, the image analysis means analyses characteristics of the face of the one or more person.

Preferably, the image analysis means calculates the distance of the face from the camera.

Preferably, the image analysis means calculates the position of the face with respect to the camera.

Preferably, the image analysis means is adapted to distinguish between each of the faces in a series of images .

Preferably, the image analysis means is adapted to record the length of time a person has been positioned facing the information output.

Preferably, the image analysis means comprises means for determining the gender of the one or more person.

Preferably, the image analysis means comprises means for determining the age of the one or more person.

Preferably, the image analysis means comprises means for determining the ethnicity of the one or more person. Preferably, the image analysis means comprises means for determining the fat-to-lean ratio of the one or more person.

Other characteristics of the person (s) may be analyzed, for example, the colour of their clothes. This feature is suitable for identifying the allegiance of sport fans

Preferably, the sensor is an imaging means for creating an image of a field of view in which one or more person may be located

Preferably, the imaging means is a video camera.

Optionally, the imaging means comprises a still camera.

Optionally, the sensor detects smells.

Optionally, the sensor detects sound.

Optionally, the sensor is touch sensitive.

Preferably , the inf ormation source is a database .

Preferably, the information in the information source is tagged with physical characteristics labels to enable the most suitable information to be identified and output.

Optionally, the information source is at a location remote from the information output. Preferably the system further comprises a module that provides statistical information on the information output .

Preferably, the module also provides characteristics of the Advert Viewer (s) .

Preferably, said statistical information is accessible via a secure web site. The statistical information can also be made available at the site of the advert, and/or on a non-secure website.

The proprietor of the output information, such as an advertiser, can access the statistical information and can compare the extent to which his information is being output at different sites where the system is in place.

The proprietor of the output information, such as an advertiser, can access the information on the person (s) currently viewing the displayed information.

Preferably, the analysis means is adapted to combine information on the person (s) currently viewing with data on the information that have been output to give a complete history of the type of information viewed.

Preferably the analysis means further provides information on the viewers that viewed a particular advert .

In accordance with a second aspect of the invention there is provided a method for supplying targeted information, the method comprising the steps of: detecting predetermined physical characteristics of a subject at or near the sensor location; determining the physical characteristics of the one or more person from sensor data; analyzing the physical characteristics of the one or more person to target the supply of information from an information source based upon the physical characteristics; and an information output.

The information output may be delivered in the form of still images, video, voice, music and other sensory modalities including olfactory release, and tactile response.

Preferably, the step of analyzing the physical characteristics comprises a content controller which controls the current output of the information output.

Preferably, the content controller is updatable in response to a change in the image of the one or more person as determined by the image analysis means.

Preferably, the step of analyzing the physical characteristics comprises analyzing sensor data derived from an image .

Preferably, the step of analyzing the physical characteristics comprises analyzing sensor data derived from a sound.

Preferably, the step of analyzing the physical characteristics comprises analyzing sensor data derived from a smell . Preferably, the step of analyzing the physical characteristics comprises analyzing sensor data derived from a touch sensor.

Preferably, the step of analyzing the physical characteristics comprises analyzing more than one person to supply the most suitable information based upon the physical characteristics of each person.

Preferably, the step of analyzing the physical characteristics comprises analyzing characteristics of the face of the one or more person.

Preferably, the step of analyzing the physical characteristics calculates the distance of the face from the camera.

Preferably, the step of analyzing the physical characteristics calculates the position of the face with respect to the camera.

Preferably, the step of analyzing the physical characteristics comprises analyzing distinguishes between each of the faces in a series of images.

Preferably, the step of analyzing the physical characteristics calculates the length of time a person has been positioned facing the information output.

Preferably, the step of analyzing the physical characteristics comprises determining the gender of the one or more person. Preferably, the step of analyzing the physical characteristics comprises determining the age of the one or more person.

Preferably, the step of analyzing the physical characteristics comprises determining the ethnicity of the one or more person.

Preferably, the step of analyzing the physical characteristics determines the fat-to-lean ratio of the one or more person.

Preferably, the information is stored in a database.

Optionally, the information source is at a location remote from the information output.

Preferably the method further comprises providing statistical information on the information output.

Preferably, said statistical information is accessible via a secure web site.

Optionally, said statistical information is accessible at the advert site. Optionally, said statistical information is accessible via a non-secure website.

Preferably, the step of analyzing the physical characteristics comprises combining information on the person (s) currently viewing with data on the information that have been output to give a complete history of the type of information viewed and the viewers.

The present invention will now be described by way of example only with reference to the accompanying drawings in which:

Figure 1 is a block diagram that shows one embodiment of the present invention;

Figure 2 is a block diagram that shows the imaging means, image analysis means and analysis means of the embodiment of the present invention of figure 1; and

Figure 3 is a block diagram of the statistical module of the present invention.

The example of the invention shown in figures 1 to 3 relates to the use of the invention in an advertising bill board that can be found in a bus shelter or at the side of the road, for example and which has multimedia capability. .The system may be used in taxis where it can be incorporated into a screen positioned behind the driver; in bars or restaurants where it can be mounted on the wall or in a table top; in information kiosks for example tourist information kiosks; in car showrooms; in ATMs where advertisements are displayed to customers undertaking or waiting to undertake a financial transaction; in shops; in cinema foyers; in post offices; supermarkets; and shopping malls.

The example of the system 1 of the present invention shown in figure 1 comprises: an observer module which contains the sensor which in this example comprises an imaging means (camera 33, figure 2) and image analysis means (including a face detector 37 and means for determining a range of physical characteristics 41, 43, 45, 47, and 51); a display manager 13, an information output (screen 23), an advert database 19 and a statistics collection system 7.

The Observer Module 3 (shown in more detail in figure 2) provides a set of values derived from analysis of the image stream showing the field of view of the camera as it looks to the front of the advertising display. The camera is positioned above or to the side of the screen that is displaying the advert. The camera may be pointed in any suitable direction, for example it may be pointed down a queue of people, or to capture an image of a person approaching the advert.

The value that is derived from an analysis of the physical characteristics of the person (s) in the field of view (also known as the Advert Viewer (s) State) in this example, the face position, face distance from the screen (as measured by the software looking through the camera) , viewing time (ms) , cumulative number of detects for this viewer, and a current estimate of the viewer (s) ' gender, age, ethnicity, fatness/thinness and ID (face recognition) , iris recognition, ear recognition and other physical characteristics including voice recognition, olfactory sensing, and tactile and chemical sensing.

The set of current values is passed from the Observer Module 3 to the Display Manager 13. The Display Manager 13 consists of the Content Controller and the Media Player.

The Content Controller is updated by the set of current values from the Observer Module 3 and uses these values to calculate what advert to show the assembled viewer (s) so that the information is targeted to the viewer (s) based upon their physical characteristics.

In this example, the advert is not changed on a millisecond to millisecond basis but an intelligent and aesthetically pleasing shift is made at an opportune moment to match the advert as closely as possible to the current audience. Fast switching is, of course, also possible if required.

The adverts are sourced by the Content Controller which accesses the Advert Database 19 to obtain the required appropriate advert. The adverts in the Advert Database 19 are all tagged with gender, age, ethnicity labels to enable the most suitable advert to be notified to the Content Controller. This appropriate advert is then given to the Media Player which then displays the advert on the screen.

The Statistics Collection System 7 (figure 1) draws information on the current viewer (s) from the Observer Module 3 and combines this information with data on the adverts that have been played to give a complete history of the type of advert viewed and the viewers. This information is then made available to the advertiser who is able to have a current and historical view of the effectiveness of the adverts in terms of number of viewers, characteristics of the viewers i.e. gender, age, ethnicity etc., length of viewing time, and other information. The information is based on the available statistics from the Observer and Display Manager Modules at the site(s) of either one or many adverts. This information can be used to improve the effectiveness of the system or to change its location

Figure 2 shows a video camera 33 which delivers digital images of the scene in front of a bill board or advertising screen, to the Face Detector module 37. The Face Detector module 37 finds any faces in the image, crops them, measures the position of the face(s) and calculates their estimated distance from the screen. This information is passed to a set of detectors.

The Face Tracker 41 homes in more accurately onto the detected faces to give more accurate face positions and distances from the screen. The Face Tracker also keeps track of which face is where in order to enable the accumulative analysis of the various characteristics of the faces .

The Gender Detector 43 analyses the face images to estimate the gender of each face.

The Age Detector 45 analyses the face images to estimate the age of each viewer as judged from their facial image.

The Ethnicity Detector 47 analyses the face images to estimate the racial group with which the viewer is associated.

The Fat-to-Lean ratio detector 51 analyses the face images to estimate how overweight the viewer is.

Face recognition can be used to target adverts to specific individuals in some cases.

All these detectors deliver updates to the Advert Viewer (s) State 53 where each viewer is associated with a set of values for Face Position, Face Distance (from screen), Advert Viewing time in milliseconds, cumulative number of detects for this viewer, and current estimates for gender, age, ethnicity, fatness/thinness etc.

In figure 3 , the Observer Module updates the Advert Viewings Logger 71 with the Advert Viewer (s) State 53. The Display Manager 13 (figure 1) updates the Advert Viewing (s) Logger with data on what adverts played at what times. The Advert Viewings Logger 71 uses the information from both these modules to calculate Who Viewed, When, How Long, and What advert was viewed. This information is passed to the Advert Viewings Database 75.

The Advertiser is then able via the Viewing Statistics Processor 79 to query the Advert Viewings Database either on-site or via a secure website across the internet. The Advertiser is then able to obtain a variety of data pertaining to the actual audience that viewed a particular advert at a particular site. Hence direct objective comparisons become possible between advert sites. In this way, the present invention not only focuses the advertising budget on the right people but can also report on exactly who saw what when and for how long.

In another embodiment of the present invention, the physical characteristics of a subject are sensed by an audio sensor which can for example estimate the age and sex of the subject.

In another embodiment of the present invention smell sensors are used to detect whether a subject wears perfume.

A further embodiment of the invention incorporates a touch sensitive detector . In addition , a camera , audio sensor , smell sensor and touch sensor may be incorporated in a single embodiment of the invention .

The present invention has a number of advantages for advertisers and consumers. The advertiser gains because advertising is more precisely targeted at the intended consumer .

Use of the present invention allows the advertiser to reach more of the intended audience. For example, normally adverts for younger women are shown to a viewing public of men and women, young and old. Consequently, the advert may be the wrong advert for 75% of the audience it eventually reaches . The intended target group of young women, in this case, would only be a 25% portion of the actual audience reached. When older women and men are the viewers the present invention switches to a more appropriate advert for that grouping.

Ideally an advert that is displayed using the present invention will 100% correctly advertise to each person when viewers are viewing the advert singly. This percentage, of course, changes subject to the accuracy of the system that depends on head pose, illumination, and facial expression in any image stream being analysed. When varied groups are present, the IAS has to present the best possible advert for a mixed group, consequently, the accurate targeting percentage will reduce but will be maximized by the present invention.

The consumer gains greatly from the present invention. For example, a teenage male surfer will no longer have to watch an advert for female menopause treatments or insurance for the elderly. Now he can watch an advert for a computer game or for sports equipment with a background of contemporary music. And that same 'great' music will not be inflicted on an older man who will instead be shown an advert for a golfing holiday with a background of Frank Sinatra music for example.

Ethnicity detection is also quite important and much requested by advertisers as different racial groups tend to have very different consumer interests in some retail areas. Of course, the consumer is also helped by only being targeted by adverts that may be of interest.

Improvements and modifications may be incorporated herein without deviating from the scope of the invention as defined by the appended claims.

Claims

CLAIMS ;

1. A system for supplying targeted information, the system comprising: a sensor for generating sensor data which defines predetermined physical characteristics of a subject at or near the sensor location; sensor data analysis means for determining the physical characteristics of the one or more person in the image; analysis means for analyzing the physical characteristics of the one or more person to target the supply of information from an information source based upon the physical characteristics; and an information output.

2. A system as claimed in claim 1 wherein, the analysis means further comprises a content controller which controls the current output of the information output.

3. A system as claimed in claim 2 wherein, the content controller is updatable in response to a change in the image of the one or more person as determined by the image analysis means.

4. A system as claimed in any of claims 1 to 3 , the analysis means analyses more than one person and supplies the most suitable information based upon the physical characteristics of each person.

5. A system as claimed άn any preceding claim wherein, the sensor data analysis means comprises image analysis means .

6. A system as claimed in claim 5 wherein, the image analysis means analyses characteristics of the face of the one or more person.

7. A system as claimed in claim 6 wherein, the image analysis means calculates the distance of the face from the camera.

8. A system as claimed in claim 6 or claim 7 wherein, the image analysis means calculates the position of the face with respect to the camera.

9. A system as claimed in any of claims 5 to 8 wherein, the image analysis means is adapted to distinguish between each of the faces in a series of images.

10. A system as claimed in any of claims 5 to 9 wherein, the image analysis means is adapted to record the length of time a person has been positioned facing the information output.

11. A system as claimed in any of claims 5 to 10 wherein, the image analysis means comprises means for determining the gender of the one or more person.

12. A system as claimed in any of claims 5 to 11 wherein, the image analysis means comprises means for determining the age of the one or more person.

13. A system as claimed in any of claims 5 to 12 wherein, the image analysis means comprises means for determining the ethnicity of the one or more person.

14. A system as claimed in any of claims 5 to 13 wherein, the image analysis means comprises means for determining the fat-to-lean ratio of the one or more person.

15. A system as claimed in any of claims 5 to 14 wherein, the image analysis means comprises means for determining a colour of an item of clothing of the one or more person.

16. A system as claimed in any preceding claim wherein, the sensor is an imaging means for creating an image of a field of view in which one or more person may be located.

17. A system as claimed in claim 16 wherein, the imaging means is a video camera.

18. A system as claimed in claim 16 wherein, the imaging means comprises a still camera.

19. A system as claimed in any preceding claim wherein, the sensor detects smells.

20. A system as claimed in any preceding claim wherein, the sensor detects sound.

21. A system as claimed in any preceding claim wherein, the sensor is touch sensitive.

22. A system as claimed in any preceding claim wherein, the information source is a database.

23. A system as claimed in any preceding claim wherein, the information in the information source is tagged with physical characteristics labels to enable the most suitable information to be identified and output .

24. A system as claimed in any preceding claim wherein, the information source is at a location remote from the information output.

25. A system as claimed in any preceding claim wherein, the system further comprises a module that provides statistical information on the information output.

26. A system as claimed in claim 25 wherein, the module also provides characteristics of the one or more person in the image.

27. A system as claimed in claim 25 or claim 26 wherein, the statistical information is accessible via a web site.

28. A system as claimed in claim 27 wherein, the web site is a secure website.

29. A system as claimed in any of claims 25 to 28 wherein, the statistical information is made available at a site of the targeted information.

30. The system as claimed in any of claims 25 to 29 wherein, a proprietor of the output information can access the statistical information so as to compare the extent to which the information is being output at different sites where the system is in place.

31. The system as claimed in any of claims 25 to 30 wherein, the proprietor of the output information can access the information on one or more persons currently viewing the targeted information.

32. The system as claimed in any of claims 1 to 31 wherein, the analysis means is adapted to combine information on one or more persons currently viewing the targeted information with data on the information that have been output to give a complete history of the type of information viewed.

33. The system as claimed in any of claims 1 to 32 wherein, the analysis means further provides information on one or more persons that viewed a particular advert.

34. A method for supplying targeted information, the method comprising the steps of: (a) detecting predetermined physical characteristics of a subject at or near the sensor location; (b) determining the physical characteristics of the one or more person from sensor data; (c) analyzing the physical characteristics of the one or more person to target the supply of information from an information source based upon the physical characteristics; and (d) providing an information output.

35. A method as claimed in claim 34 wherein, the information output is delivered in one or more forms selected from the group of; still images, video, voice, music and other sensory modalities including olfactory release, and tactile response.

36. A method as claimed in claim 34 or claim 35 wherein, the step of analyzing the physical characteristics comprises a content controller which controls the current output of the information output.

37. A method as claimed in claim 36 wherein, the content controller is updatable in response to a change in the image of the one or more person as determined by the image analysis means.

38. A method as claimed in any of claims 34 to 37 wherein, the step of analyzing the physical characteristics comprises analyzing sensor data derived from an image.

39. A method as claimed in any of claims 34 to 38 wherein, the step of analyzing the physical characteristics comprises analyzing sensor data derived from a sound.

40. A method as claimed in any of claims 34 to 39 wherein, the step of analyzing the physical characteristics comprises analyzing sensor data derived from a smell.

41. A method as claimed in any of claims 34 to 40 wherein, the step of analyzing the physical characteristics comprises analyzing sensor data derived from a touch sensor.

42. A method as claimed in any of claims 34 to 41 wherein, the step of analyzing the physical characteristics comprises analyzing more than one person to provide the most suitable information based upon the physical characteristics of each of the one or more person.

43. A method as claimed in any of claims 34 to 42 wherein, the step of analyzing the physical characteristics comprises analyzing characteristics of the face of the one or more person.

44. A method as claimed in claim 43 wherein, the step of analyzing the physical characteristics calculates the distance of the face from the camera.

45. A method as claimed in claim 43 or claim 44 wherein, the step of analyzing the physical characteristics calculates the position of the face with respect to the camera.

46. A method as claimed in any of claims 43 to 45 wherein, the step of analyzing the physical characteristics comprises distinguishing between each of the faces in a series of images.

47. A method as claimed in any of claims 34 to 46, the step of analyzing the physical characteristics calculates the length of time a person has been positioned facing the information output.

48. A method as claimed in any of claims 34 to 47 wherein, the step of analyzing the physical characteristics comprises determining the gender of the one or more person.

49. A method as claimed in any of claims 34 to 48 wherein, the step of analyzing the physical characteristics comprises determining the age of the one or more person.

50. A method as claimed in any of claims 34 to 49 wherein, the step of analyzing the physical characteristics comprises determining the ethnicity of the one or more person.

51. A method as claimed in any of claims 34 to 50 wherein, the step of analyzing the physical characteristics determines the fat-to-lean ratio of the one or more person.

52. A method as claimed in any of claims 34 to 51 wherein, the step of analyzing the physical characteristics determines a colour of an item of clothing of the one or more person.

53. A method as claimed in any of claims 34 to 52 wherein, the information is stored in a database.

54. A method as claimed in any of claims 34 to 53 wherein, the information in the information source is tagged with physical characteristics labels to enable the most suitable information to be identified and output .

55 . A method as claimed in any of claims 34 to 54 wherein , the information source is at a location remote from the information output .

56. A method as claimed in any of claims 34 to 54 further comprising the step of providing statistical information on the information output.

57. A method as claimed in claim 56 wherein, the statistical information is accessible via a web site.

58. A method as claimed in claim 57 wherein, the web site is secure.

59. A method as claimed in any of claims 56 to 57 wherein, the statistical information is accessible at an advert site.

60. A method as claimed in any of claims 56 to 59 wherein, a proprietor of the output information can access the statistical information so as to compare the extent to which the information is being output at different sites where the system is in place.

61. A method as claimed in any of claims 34 to 60 wherein, the step of analyzing the physical characteristics comprises combining information on the one or more person currently viewing with data on the information that have been output to give a complete history of the type of information viewed and the viewers .