WO2018090447A1 - 广告质量评估方法及装置 - Google Patents

广告质量评估方法及装置 Download PDF

Info

Publication number
WO2018090447A1
WO2018090447A1 PCT/CN2016/112595 CN2016112595W WO2018090447A1 WO 2018090447 A1 WO2018090447 A1 WO 2018090447A1 CN 2016112595 W CN2016112595 W CN 2016112595W WO 2018090447 A1 WO2018090447 A1 WO 2018090447A1
Authority
WO
WIPO (PCT)
Prior art keywords
image
advertisement
time period
user
preset time
Prior art date
Application number
PCT/CN2016/112595
Other languages
English (en)
French (fr)
Inventor
张银刚
Original Assignee
深圳Tcl数字技术有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 深圳Tcl数字技术有限公司 filed Critical 深圳Tcl数字技术有限公司
Publication of WO2018090447A1 publication Critical patent/WO2018090447A1/zh

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00Commerce
    • G06Q30/02Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241Advertisements
    • G06Q30/0242Determining effectiveness of advertisements
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06QINFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q30/00Commerce
    • G06Q30/02Marketing; Price estimation or determination; Fundraising
    • G06Q30/0241Advertisements
    • G06Q30/0277Online advertisement

Definitions

  • the present invention relates to the field of network and mobile communication technologies, and in particular, to an advertisement quality evaluation method and apparatus.
  • the main object of the present invention is to provide an advertisement quality evaluation method and device, which aims to solve the technical problem that the user cannot easily and quickly know the user's satisfaction with the advertisement, and thus cannot improve the advertisement efficiency in time.
  • the present invention provides an advertisement quality evaluation method, and the method for evaluating an advertisement quality method includes:
  • the user's head behavior image is collected every preset time period, and the collected head behavior images are compared with the reference image;
  • the user characteristics of each preset time period are mapped to the advertisement quality of the preset time period for evaluation by the evaluator.
  • the method for evaluating the quality of the advertisement further includes:
  • the behavior image of the user is collected every preset time period, and the collected behavior images are compared with the reference image;
  • the user characteristics of each preset time period are mapped to the advertisement quality of the preset time period for evaluation by the evaluator.
  • the step of acquiring the difference value between the behavior image and the reference image, and marking the user feature corresponding to the preset time period according to the difference value between the behavior image and the reference image comprises:
  • the step of marking the user feature corresponding to the preset time period as the attraction class feature comprises:
  • the method for evaluating the quality of the advertisement further comprises:
  • the advertisement quality of the preset time period is mapped according to the user feature for evaluation by an evaluator.
  • the step of acquiring the difference value between the comparison image and the reference image comprises:
  • the method for evaluating the quality of the advertisement further comprises:
  • the user features of the preset number of users are obtained in each preset time period of the advertisement, and the advertisement quality of each preset time period is mapped according to the user characteristics of the preset number of users for evaluation by the evaluator.
  • the present invention also provides an advertisement quality evaluation apparatus, and the advertisement quality evaluation apparatus includes:
  • a first marking module configured to acquire a difference value between the behavior image and the reference image, and mark a user feature corresponding to the preset time period according to the difference value between the behavior image and the reference image;
  • the first marking module comprises:
  • a first marking unit configured to acquire a difference value between the behavior image and the reference image, and when the difference value between the behavior image and the reference image exceeds a first preset threshold, marking the user feature corresponding to the preset time period as a repulsion feature ;
  • a detecting module configured to mark, according to the interrupted advertisement event, a user feature corresponding to a preset time period when detecting an interrupted advertisement instruction sent by the preset input device;
  • the first marking module comprises:
  • An obtaining unit configured to acquire position information of a reference point image and a center point of the user's head in each of the acquired behavior images, and obtain the position between the comparison image and the reference image according to position information of the center point of the user head The difference value.
  • the third mapping module is configured to acquire a user feature of the preset number of users in each preset time period of the advertisement, and map the advertisement quality of the preset time period according to the user feature of the preset number of users, for the evaluator to evaluate.
  • the user quality of the preset time period is mapped according to the difference between the behavior image and the reference image and the user feature corresponding to the difference value, so that the user can quickly learn the presets.
  • the satisfaction of the time period advertisement is fed back to the advertiser, and the advertiser improves the corresponding advertisement according to the feedback, thereby improving the advertisement efficiency.
  • FIG. 1 is a schematic flow chart of a first embodiment of an advertisement quality evaluation method according to the present invention.
  • FIG. 2 is a schematic diagram of a refinement process of obtaining a difference value between the behavior image and the reference image in the second embodiment of the advertisement quality evaluation method according to the present invention, and marking a user feature step corresponding to the preset time period according to the difference value between the behavior image and the reference image ;
  • FIG. 3 is a refinement process of the step of marking the user feature of the preset time period as the attracting feature step when the difference between the behavior image and the reference image is lower than the first preset threshold in the third embodiment of the advertisement quality evaluation method of the present invention.
  • FIG. 4 is a schematic flow chart of a fourth embodiment of an advertisement quality evaluation method according to the present invention.
  • FIG. 7 is a schematic diagram of a refinement function module of a second marking unit in a third embodiment of an advertisement quality evaluation apparatus based on user behavior according to the present invention.
  • the method includes:
  • Step S10 capturing a behavior image of the user at the initial moment of the advertisement playing, and using the behavior image as a reference image;
  • T1 the time at this time is recorded as T1.
  • Initialize and turn on the smart device camera initialize the process to reduce the resolution of the image and increase the size of the image, which is good for detecting human displacement and changes in human body language, so as to analyze the image according to human displacement and changes in human body language.
  • each behavior image records the behavior characteristics of the user in the preset time period, and thus divides the entire advertisement playing time into n hours, and the corresponding advertising period of the preset time period mark is A1, A2, A3...
  • An comparing the collected behavior images with the reference image, the comparison is not only the displacement of the human in the image but also the change of the human body language, and the judgment of the user shaking the head, leaving, etc., when the user
  • the position information of the user on the captured image may change, such as when the user is not collecting the image or the user is at the edge position of the captured image, and when the user has a bias or the like, The position information of the head on the image also changes.
  • Step S30 acquiring a difference value between the behavior image and the reference image, and marking a user feature corresponding to the preset time period according to the difference value between the behavior image and the reference image;
  • Obtaining a difference value between the behavior image and the reference image such as acquiring location information of a user's head in the reference image, where the location information includes an abscissa and an ordinate information of a user's head center point in the image, in the reference image
  • the position of the center point of the user's head can be located in the center of the image, and the position information of the user's head in the behavior image is obtained, that is, the abscissa and ordinate information of the center point of the user's head in the picture in the behavior image is obtained.
  • Obtaining a difference value between the behavior image and the reference image includes acquiring a difference value of the position information of the user head center point in the reference image and the behavior image or a difference value of the position information of the user limb center point in the reference image and the behavior image, according to The difference value of the behavior image and the reference image marks the user feature corresponding to the preset time period, and the user feature includes the attraction class feature and the exclusion class feature.
  • the user feature is the attraction class feature
  • the user has high attention to the advertisement, for example, The user does not have an action, always looks at the advertisement, and when the user feature is a exclusion class feature, the advertisement pair is Households unattractive, such as shaking his head, leaving the user, the user using the remote control to terminate the ad playback, when the difference exceeds a preset threshold value, it is determined that user behavior has changed.
  • the user characteristics of the preset time period are mapped to the advertisement quality of the preset time period.
  • the advertisement quality is high in the preset time period, and the user applies the advertisement to the advertisement.
  • the satisfaction of the advertiser is high, and the advertiser can continue to use the advertisement of the period of time.
  • the user feature is a exclusion feature in the preset time period, the user's satisfaction with the advertisement to be placed is low, that is, each preset time period is The user feature maps the advertisement quality of the preset time period, and the advertiser accordingly improves or tailors the advertisement period that the user is not interested or has low satisfaction for the preset time period.
  • the user quality of the preset time period is mapped according to the difference between the behavior image and the reference image and the user feature corresponding to the difference value, so that the user can quickly learn the presets.
  • the satisfaction of the time period advertisement feeds the satisfaction of the user to each preset time period to the advertiser, and the advertiser improves the corresponding advertisement according to the feedback, thereby improving the advertising efficiency.
  • step S30 includes:
  • Step S31 Obtain a difference value between the behavior image and the reference image.
  • the difference value between the behavior image and the reference image exceeds a first preset threshold, marking the user feature corresponding to the preset time period as an exclusion type feature;
  • Step S32 when the behavior image and the reference image difference value are lower than the first preset threshold, marking the user feature corresponding to the preset time period as an attraction class feature.
  • the threshold is set to reduce misjudgment, such as the user blinks and the user does not think that the user feature is Exclusion class features.
  • the user feature corresponding to the preset time period is marked as an attractive feature.
  • the user feature corresponding to the preset time period is marked as a repulsion feature; when the difference between the behavior image and the reference image is lower than the first preset threshold Marking the user feature corresponding to the preset time period as the attracting feature, that is, accurately determining the user feature, and accurately determining the user feature is the basis for accurately obtaining the user's satisfaction with each preset time period, thereby facilitating Merchants further improve their advertising and improve their advertising effectiveness.
  • step S32 includes:
  • step S34 the collected facial expression image is compared with the system preset expression.
  • the user feature corresponding to the preset time period is marked as an attractive feature.
  • the facial expression image corresponding to the behavior image of each preset time period is collected; and the collected facial expression image is compared with the system preset expression.
  • the user feature corresponding to the preset time period is marked as an attraction class feature. Since the user's satisfaction with the advertisement is not necessarily high when the difference value between the behavior image and the reference image is lower than the preset threshold, for example, the user may have a closed eye or a disgusting expression on the face, etc.
  • the system preset expressions are the expressions that the user is interested in the advertisement or attracted by the advertisement, thus When the user's expression matches the preset expression, that is, both expressions are happy, or the eyes are staring at the screen, and the user features corresponding to the preset time period are marked as attractive features.
  • the facial expression image corresponding to each preset time period behavior image is collected; the collected facial expression image and system
  • the preset expressions are compared.
  • the user feature corresponding to the preset time period is marked as an attractive feature. Since the user characteristics are judged on the basis of obtaining the expression, the user's satisfaction with the advertisement can be more accurately known, which is beneficial to the merchant to further improve the advertisement and improve the advertisement efficiency.
  • a fourth embodiment of the method for evaluating the quality of the advertisement further includes:
  • Step S50 when detecting an interrupted advertisement instruction sent by the preset input device, marking, according to the interrupted advertisement event, a user feature corresponding to the preset time period;
  • Step S60 Mapping the advertisement quality of the preset time period according to the user feature, for the evaluator to perform the evaluation.
  • the user When the interrupted advertisement instruction sent by the preset input device is detected, the user is generally dissatisfied or uninterested in the corresponding time period advertisement, and the user feature corresponding to the preset time period is marked according to the interrupted advertisement event, according to the user.
  • the feature maps the quality of the advertisement for the preset time period for evaluation by the evaluator.
  • the user feature corresponding to the preset time period is marked according to the interrupted advertisement event; and the preset time period is mapped according to the user feature.
  • the quality of the advertisement for evaluation by the evaluator. Ad quality is assessed more fully by considering the disruption of advertising sent by the input device.
  • a fifth embodiment of the advertisement quality evaluation method is proposed.
  • the difference between the comparison image and the reference image is acquired.
  • Value steps include:
  • the difference value between the comparison image and the reference image is obtained according to the position information of the center point of the user's head.
  • the position information of the center point of the user's head in the reference image and the collected behavior images is obtained, and the comparison image and the reference image are obtained according to the position information of the center point of the user's head.
  • the difference between the values Since the difference value between the comparison image and the reference image is obtained according to the location information of the center point of the user's head, the user feature corresponding to the preset time period can be accurately obtained, thereby providing a basis for accurately obtaining the user's satisfaction with the advertisement. .
  • the method for evaluating the quality of the advertisement further includes:
  • the present invention also provides an advertisement quality evaluation method apparatus.
  • the apparatus includes:
  • the shooting module 10 is configured to capture a behavior image of the user at the initial moment of the advertisement playing, and use the behavior image as a reference image;
  • T1 the time at this time is recorded as T1.
  • Initialize and turn on the smart device camera initialize the process to reduce the resolution of the image and increase the size of the image, which is good for detecting human displacement and changes in human body language, so as to analyze the image according to human displacement and changes in human body language.
  • the user Before the advertisement is played, the user generally views the target content that is of interest to the user or opens the corresponding application of the smart terminal to view the target content, so that the user makes an interest in the advertisement content at the initial moment of the advertisement playing, such as near the smart
  • the terminal captures the behavior image of the user at the initial moment of the advertisement playing, and thus the captured image may be the user's head or the limb at the center of the picture, and the image is used as a reference template P for judging the user behavior, that is, the reference image, and then the user is started to capture.
  • the image when watching the animation.
  • each behavior image records the behavior characteristics of the user in the preset time period, and thus divides the entire advertisement playing time into n hours, and the corresponding advertising period of the preset time period mark is A1, A2, A3...
  • An comparing the collected behavior images with the reference image, the comparison is not only the displacement of the human in the image but also the change of the human body language, and the judgment of the user shaking the head, leaving, etc., when the user
  • the position information of the user on the captured image may change, such as when the user is not collecting the image or the user is at the edge position of the captured image, and when the user has a bias or the like, The position information of the head on the image also changes.
  • a first marking module 30 configured to acquire a difference value between the behavior image and the reference image, and mark a user feature corresponding to the preset time period according to the difference value of the behavior image and the reference image;
  • Obtaining a difference value between the behavior image and the reference image such as acquiring location information of a user's head in the reference image, where the location information includes an abscissa and an ordinate information of a user's head center point in the image, in the reference image
  • the position of the center point of the user's head can be located in the center of the image, and the position information of the user's head in the behavior image is obtained, that is, the abscissa and ordinate information of the center point of the user's head in the picture in the behavior image is obtained.
  • Obtaining a difference value between the behavior image and the reference image includes acquiring a difference value of the position information of the user head center point in the reference image and the behavior image or a difference value of the position information of the user limb center point in the reference image and the behavior image, according to The difference value of the behavior image and the reference image marks the user feature corresponding to the preset time period, and the user feature includes the attraction class feature and the exclusion class feature.
  • the user feature is the attraction class feature
  • the user has high attention to the advertisement, for example, The user does not have an action, always looks at the advertisement, and when the user feature is a exclusion class feature, the advertisement pair is Households unattractive, such as shaking his head, leaving the user, the user using the remote control to terminate the ad playback, when the difference exceeds a preset threshold value, it is determined that user behavior has changed.
  • the first mapping module 40 is configured to map user characteristics of each preset time period to the advertisement quality of the preset time period for evaluation by the evaluator.
  • the user characteristics of the preset time period are mapped to the advertisement quality of the preset time period.
  • the advertisement quality is high in the preset time period, and the user applies the advertisement to the advertisement.
  • the satisfaction of the advertiser is high, and the advertiser can continue to use the advertisement of the period of time.
  • the user feature is a exclusion feature in the preset time period, the user's satisfaction with the advertisement to be placed is low, that is, each preset time period is The user feature maps the advertisement quality of the preset time period, and the advertiser accordingly improves or tailors the advertisement period that the user is not interested or has low satisfaction for the preset time period.
  • the shooting module 10 is configured to capture a behavior image of the user at the initial moment of the advertisement playing, and use the behavior image as a reference image; and the collecting module 20 is configured to preset a preset time period during the advertisement playing process. Collecting a behavior image of the user, and comparing the collected behavior images with the reference image; the first marking module 30 is configured to acquire a difference value between the behavior image and the reference image, according to the difference between the behavior image and the reference image The value tag corresponds to the user feature of the preset time period; the first mapping module 40 is configured to map the user features of the preset time segments to the advertisement quality of the preset time period for the evaluator to perform the evaluation.
  • the user quality of the preset time period is mapped according to the difference between the behavior image and the reference image and the user feature corresponding to the difference value, so that the user can quickly learn the presets.
  • the satisfaction of the time period advertisement feeds the satisfaction of the user to each preset time period to the advertiser, and the advertiser improves the corresponding advertisement according to the feedback, thereby improving the advertising efficiency.
  • the first marking module includes:
  • the first marking unit 31 is configured to acquire a difference value between the behavior image and the reference image. When the difference value between the behavior image and the reference image exceeds a first preset threshold, marking the user feature corresponding to the preset time period as a exclusion class feature;
  • the second marking unit 32 is configured to mark the user feature corresponding to the preset time period as an attraction class feature when the behavior image and the reference image difference value are lower than the first preset threshold.
  • the user feature corresponding to the preset time period is a exclusion type feature, such as a difference between the picture P and the Pt. If the difference value is greater than 40%, the user is considered to have left, and the time of the advertisement is used to take a break. At this time, the user feature corresponding to the preset time period is the exclusion type feature, and if the user does not have an action, the difference is less than the threshold value.
  • the threshold is set to reduce misjudgment, such as the user blinks and the user does not think that the user feature is Exclusion class features.
  • the first marking unit 31 is configured to acquire a difference value between the behavior image and the reference image, and when the difference value between the behavior image and the reference image exceeds the first preset threshold, the marking corresponds to the preset time.
  • the user feature of the segment is a repulsion class feature; the second tagging unit 32 is configured to mark the user feature corresponding to the preset time period as an attraction class feature when the behavior image and the reference image difference value are lower than the first preset threshold.
  • the user feature corresponding to the preset time period is marked as a repulsion feature; when the difference between the behavior image and the reference image is lower than the first preset threshold Marking the user feature corresponding to the preset time period as the attracting feature, that is, accurately determining the user feature, and accurately determining the user feature is the basis for accurately obtaining the user's satisfaction with each preset time period, thereby facilitating Merchants further improve their advertising and improve their advertising effectiveness.
  • the second marking unit includes:
  • the collecting sub-unit 33 is configured to collect a facial expression image corresponding to each preset time period behavior image when the difference value of the behavior image and the reference image is lower than a preset threshold;
  • the comparison sub-unit 34 is configured to compare the collected facial expression image with the system preset expression. When the user's expression matches the preset expression, the user feature corresponding to the preset time period is marked as an attractive feature.
  • the collecting sub-unit 33 is configured to collect a facial expression image corresponding to each preset time period behavior image when the difference value between the behavior image and the reference image is lower than a preset threshold;
  • the sub-unit 34 is configured to compare the collected facial expression image with the system preset expression.
  • the user feature corresponding to the preset time period is marked as an attractive feature. Since the user characteristics are judged on the basis of obtaining the expression, the user's satisfaction with the advertisement can be more accurately known, which is beneficial to the merchant to further improve the advertisement and improve the advertisement efficiency.
  • the advertisement quality evaluation apparatus further includes:
  • the detecting module 50 is configured to: when the interrupted advertisement instruction sent by the preset input device is detected, mark the user feature corresponding to the preset time period according to the interrupted advertisement event;
  • the second mapping module 60 is configured to map the advertisement quality of the preset time period according to the user feature for evaluation by an evaluator.
  • the user When the interrupted advertisement instruction sent by the preset input device is detected, the user is generally dissatisfied or uninterested in the corresponding time period advertisement, and the user feature corresponding to the preset time period is marked according to the interrupted advertisement event, according to the user.
  • the feature maps the quality of the advertisement for the preset time period for evaluation by the evaluator.
  • the detecting module 50 is configured to: when the interrupted advertisement instruction sent by the preset input device is detected, mark the user feature corresponding to the preset time period according to the interrupted advertisement event; and the second mapping module 60 uses The advertisement quality of the preset time period is mapped according to the user feature for evaluation by an evaluator. Ad quality is assessed more fully by considering the disruption of advertising sent by the input device.
  • the first marking module includes:
  • An acquiring unit configured to acquire position information of a reference point and a center point of the user's head in each of the collected behavior images, and obtain, between the comparison image and the reference image, according to position information of the center point of the user head Difference value.
  • the difference value between the comparison image and the reference image is obtained according to the position information of the center point of the user's head.
  • the acquiring unit is configured to acquire position information of a reference point and a center point of the user's head in each of the collected behavior images, and obtain the comparison according to position information of the center point of the user head.
  • the difference value between the image and the reference image Since the difference value between the comparison image and the reference image is obtained according to the location information of the center point of the user's head, the user feature corresponding to the preset time period can be accurately obtained, thereby providing a basis for accurately obtaining the user's satisfaction with the advertisement. .
  • the advertisement quality evaluation apparatus further includes:
  • the third mapping module is configured to acquire a user feature of the preset number of users in each preset time period of the advertisement, and map the advertisement quality of the preset time period according to the user feature of the preset number of users, for the evaluator to evaluate.
  • the third mapping module is configured to acquire a user feature of a preset number of users in each preset time period of the advertisement, and map the preset time segments according to the user features of the preset number of users.
  • the quality of the advertisement for evaluation by the evaluator Since the satisfaction of the preset number of users for each preset time period is obtained, the accidental factors are eliminated, the accuracy is improved, and the benefit of the advertisement of the merchant is further improved.

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Accounting & Taxation (AREA)
  • Development Economics (AREA)
  • Strategic Management (AREA)
  • Finance (AREA)
  • Game Theory and Decision Science (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

一种广告质量评估方法,所述广告质量评估方法包括:拍摄广告播放初始时刻用户的行为图像,并将该行为图像作为参考图像;在广告播放过程中,每间隔预设时间段采集用户的行为图像,并将所采集的各幅行为图像与参考图像进行比对;获取所述行为图像与参考图像的差异值,根据行为图像与参考图像的差异值标记对应预设时间段的用户特征;将各个预设时间段的用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。本方案解决了现有技术中无法简单快速得知用户对广告的满意度,从而不能及时提高广告效率的技术问题。

Description

广告质量评估方法及装置
技术领域
本发明涉及网络及移动通讯技术领域,尤其涉及一种广告质量评估方法及装置。
背景技术
目前,可访问互联网的智能设备已被广泛使用,广告商通常会通过智能设备向用户推送广告,然而,现有技术中广告商在广告播出以后却无法简单快速监控用户对广告的满意度,即无法简单快速得知此次推送的广告,哪一部分能吸引用户,哪一部分用户并不感兴趣,也就无法对广告中的不足进行改进或者提高广告质量,从而不能及时提升相应广告的效率。
发明内容
本发明的主要目的在于提供一种广告质量评估方法及装置,旨在解决现有技术中无法无法简单快速得知用户对广告的满意度,从而不能及时提高广告效率的技术问题
为实现上述目的,本发明提供一种广告质量评估方法,所述广告质量评估方法方法包括:
拍摄广告播放初始时刻用户的头部行为图像,并将该头部行为图像作为参考图像;
在广告播放过程中,每间隔预设时间段采集用户的头部行为图像,并将所采集的各幅头部行为图像与参考图像进行比对;
获取所述头部行为图像与参考图像的差异值,根据头部行为图像与参考图像的差异值标记对应预设时间段的用户特征;
将各个预设时间段的用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。
所述广告质量评估方法方法还包括:
拍摄广告播放初始时刻用户的行为图像,并将该行为图像作为参考图像;
在广告播放过程中,每间隔预设时间段采集用户的行为图像,并将所采集的各幅行为图像与参考图像进行比对;
获取所述行为图像与参考图像的差异值,根据行为图像与参考图像的差异值标记对应预设时间段的用户特征;
将各个预设时间段的用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。
优选地,所述获取所述行为图像与参考图像的差异值,根据行为图像与参考图像的差异值标记对应预设时间段的用户特征步骤包括:
获取所述行为图像与参考图像的差异值,当行为图像与参考图像的差异值超过第一预设阀值时,标记对应预设时间段的用户特征为排斥类特征;
当行为图像与参考图像差异值低于第一预设阀值时,标记对应预设时间段的用户特征为吸引类特征。
优选地,所述当行为图像与参考图像差异值低于第一预设阀值时,标记对应预设时间段的用户特征为吸引类特征步骤包括:
当行为图像与参考图像的差异值低于预设阀值时,采集对应各个预设时间段行为图像所对应的人脸表情图像;
将采集的人脸表情图像与系统预设表情进行比对,当用户的表情与预设表情匹配时,标记对应预设时间段的用户特征为吸引类特征。
优选地,所述广告质量评估方法还包括:
当检测到预设输入设备发送的中断广告指令时,根据所述中断广告事件标记对应预设时间段的用户特征;
根据所述用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。
优选地,所述获取所述比对图像与参考图像之间的差异值步骤包括:
获取参考图像与采集的各幅行为图像中用户头部中心点所处的位置信息,根据所述用户头部中心点的位置信息获取所述比对图像与参考图像之间的差异值。
优选地,所述广告质量评估方法还包括:
在广告各个预设时间段内获取预设数量用户的用户特征,根据所述预设数量用户的用户特征映射所述各个预设时间段的广告质量,以供评估人员进行评估。
为实现上述目的,本发明还提供一种广告质量评估装置,所述广告质量评估装置包括:
拍摄模块,用于拍摄广告播放初始时刻用户的行为图像,并将该行为图像作为参考图像;
采集模块,用于在广告播放过程中,每间隔预设时间段采集用户的行为图像,并将所采集的各幅行为图像与参考图像进行比对;
第一标记模块,用于获取所述行为图像与参考图像的差异值,根据行为图像与参考图像的差异值标记对应预设时间段的用户特征;
第一映射模块,用于将各个预设时间段的用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。
可选地,所述第一标记模块包括:
第一标记单元,用于获取所述行为图像与参考图像的差异值,当行为图像与参考图像的差异值超过第一预设阀值时,标记对应预设时间段的用户特征为排斥类特征;
第二标记单元,用于当行为图像与参考图像差异值低于第一预设阀值时,标记对应预设时间段的用户特征为吸引类特征。
可选地,所述广告质量评估装置还包括:
检测模块,用于当检测到预设输入设备发送的中断广告指令时,根据所述中断广告事件标记对应预设时间段的用户特征;
第二映射模块,用于根据所述用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。
可选地,所述第一标记模块包括:
获取单元,用于获取获取参考图像与采集的各幅行为图像中用户头部中心点所处的位置信息,根据所述用户头部中心点的位置信息获取所述比对图像与参考图像之间的差异值。
可选地,所述广告质量评估装置还包括:
第三映射模块,用于在广告各个预设时间段内获取预设数量用户的用户特征,根据所述预设数量用户的用户特征映射所述各个预设时间段的广告质量,以供评估人员进行评估。
本发明通过一种基于用户行为的广告质量评估方法,该方法通过拍摄广告播放初始时刻用户的行为图像,并将该行为图像作为参考图像;在广告播放过程中,每间隔预设时间段采集用户的行为图像,并将所采集的各幅行为图像与参考图像进行比对;获取所述行为图像与参考图像的差异值,根据行为图像与参考图像的差异值标记对应预设时间段的用户特征;将各个预设时间段的用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。由于每间隔预设时间段采集用户的行为图像,根据行为图像与参考图像的差异值与差异值对应的用户特征映射所述预设时间段的广告质量,因而能够快速得知用户对各个预设时间段广告的满意度,并反馈给广告商,广告商根据所述反馈改进相应广告,进而提升广告效率。
附图说明
图1为本发明广告质量评估方法第一实施例的流程示意图;
图2为本发明广告质量评估方法第二实施例中获取所述行为图像与参考图像的差异值,根据行为图像与参考图像的差异值标记对应预设时间段的用户特征步骤的细化流程示意图;
图3为本发明广告质量评估方法第三实施例中当行为图像与参考图像差异值低于第一预设阀值时,标记对应预设时间段的用户特征为吸引类特征步骤的细化流程示意图;
图4为本发明广告质量评估方法第四实施例的流程示意图;
图5为本发明基于用户行为的广告质量评估装置第一实施例的细化功能模块示意图;
图6为本发明基于用户行为的广告质量评估装置第二实施例中第一标记模块的细化功能模块示意图;
图7为本发明基于用户行为的广告质量评估装置第三实施例中第二标记单元的细化功能模块示意图;
图8为本发明基于用户行为的广告质量评估装置第四实施例的细化功能模块示意图。
本发明目的的实现、功能特点及优点将结合实施例,参照附图做进一步说明。
具体实施方式
应当理解,此处所描述的具体实施例仅仅用以解释本发明,并不用于限定本发明。
为更好理解本发明,在此提供广告质量评估方法,在广告质量评估方法第一实施例中,参照图1,该方法包括:
步骤S10,拍摄广告播放初始时刻用户的行为图像,并将该行为图像作为参考图像;
当开始放广告时,记录此时时间为T1, 初始化并打开智能设备摄像头,初始化过程降底图像的分辨率并增大图片的尺寸,这样利于检测人的位移以及人肢体语言的变化,以便根据人的位移以及人肢体语言的变化对图片进行分析,由于在广告播放前,用户一般在观看自身感兴趣的各项目标内容或者打开智能终端相应应用以观看目标内容,因而在广告播放初始时刻用户做出对广告内容感兴趣的行为,如靠近智能终端,拍摄广告播放初始时刻用户的行为图像,因而拍摄出来的图片可是用户头部或者肢体处在图片中心位置,将该图像作为判断用户行为的参考模板P,即为参考图像,然后开始捕捉用户观看动画时的图像。
步骤S20,在广告播放过程中,每间隔预设时间段采集用户的行为图像,并将所采集的各幅行为图像与参考图像进行比对;
在广告播放过程中,即是当广告持续进行时,每间隔预设时间段采集用户的行为图像,记录采集图像的时刻T1,T2,T3…Tn,采集的行为图像标记为P1,P2,P3 … Pn,每张行为图像记录该预设时间段用户的行为特征,因而把整个广告播放时间切分为n个小时间片,所述预设时间段标记对应的广告时段为A1,A2,A3…An,将所采集的各幅行为图像与参考图像进行比对,所述比对除比对图像中人的位移以及人肢体语言的变化,还包括对用户摇头,离开等动作的判断,当用户出现大的位移变化即远离相应广告播放终端时,用户在采集图像上的位置信息会发生变化,如用户不在采集图像上或者用户在所采集图像的边缘位置,而当用户出现偏头等动作时,其头部在图像上所处的位置信息也会发生变化。
步骤S30,获取所述行为图像与参考图像的差异值,根据行为图像与参考图像的差异值标记对应预设时间段的用户特征;
获取所述行为图像与参考图像的差异值,如获取参考图像中用户头部所处的位置信息,所述位置信息包括用户头部中心点在图像中的横坐标与纵坐标信息,参考图像中用户头部中心点所处的位置可在图像的正中央,获取行为图像中用户头部所处的位置信息,即获取行为图像中用户头部中心点在图片中的横坐标与纵坐标信息,获取所述行为图像与参考图像的差异值包括获取用户头部中心点在参考图像与行为图像中位置信息的差异值或者用户肢体中心点在在参考图像与行为图像中位置信息的差异值,根据行为图像与参考图像的差异值标记对应预设时间段的用户特征,所述用户特征包括吸引类特征与排斥类特征,当用户特征为吸引类特征时,表示用户对广告关注度很高,例如用户没有发生动作,始终注视着广告,当用户特征为排斥类特征时,表示广告对用户缺乏吸引力,例如摇头,用户离开,用户使用遥控器终止广告播放等,当差异值超过预设阀值时,则判断用户行为发生了变化。
步骤S40,将各个预设时间段的用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。
将各个预设时间段的用户特征映射所述预设时间段的广告质量,当预设时间段内用户特征为吸引类特征时,说明该预设时间段内广告质量高,用户对所投放广告的满意度高,广告商可以继续沿用该段时间段的广告,当预设时间段内用户特征为排斥类特征时,说明用户对所投放广告的满意度低,即是各个预设时间段的用户特征映射所述预设时间段的广告质量,广告商据此对广告进行改进或裁剪该预设时间段用户不感兴趣或者满意度低的广告时段。
在本实施例中,通过拍摄广告播放初始时刻用户的行为图像,并将该行为图像作为参考图像;在广告播放过程中,每间隔预设时间段采集用户的行为图像,并将所采集的各幅行为图像与参考图像进行比对;获取所述行为图像与参考图像的差异值,根据行为图像与参考图像的差异值标记对应预设时间段的用户特征;将各个预设时间段的用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。由于每间隔预设时间段采集用户的行为图像,根据行为图像与参考图像的差异值与差异值对应的用户特征映射所述预设时间段的广告质量,因而能够快速得知用户对各个预设时间段广告的满意度,将用户对各个预设时间段的满意度反馈给广告商,广告商根据所述反馈改进相应广告,进而提升广告效率。
进一步地,在本发明广告质量评估方法第一实施例的基础上,提出广告质量评估方法第二实施例,参照图2,在第二实施例中,步骤S30包括:
步骤S31,获取所述行为图像与参考图像的差异值,当行为图像与参考图像的差异值超过第一预设阀值时,标记对应预设时间段的用户特征为排斥类特征;
步骤S32,当行为图像与参考图像差异值低于第一预设阀值时,标记对应预设时间段的用户特征为吸引类特征。
获取所述行为图像与参考图像的差异值,行为图像与参考图像的差异值超过第一预设阀值时,标记对应预设时间段的用户特征为排斥类特征,如图片P和Pt差异较大,差异值大于40%,则认为用户离开了,利用播广告的时间休息一下,此时标记对应预设时间段的用户特征为排斥类特征,而用户没有发生动作,则差异小于阀值%3,则图片P和Pt差异较小,差异大于阀值%3小于%40,也认为用户没有发生动作,设置阀值是为了减少误判,比如用户眨了眨眼睛等不会认为用户特征为排斥类特征。
在本实施例中,通过获取所述行为图像与参考图像的差异值,当行为图像与参考图像的差异值超过第一预设阀值时,标记对应预设时间段的用户特征为排斥类特征;当行为图像与参考图像差异值低于第一预设阀值时,标记对应预设时间段的用户特征为吸引类特征。由于当行为图像与参考图像的差异值超过第一预设阀值时,标记对应预设时间段的用户特征为排斥类特征;当行为图像与参考图像差异值低于第一预设阀值时,标记对应预设时间段的用户特征为吸引类特征,即是对用户特征进行准确判断,而对用户特征进行准确判断是准确获取用户对各个预设时间段广告满意度的基础,从而有利于商家进一步改进广告,提升广告效益。
进一步地,在本发明广告质量评估方法第二实施例的基础上,提出广告质量评估方法第三实施例,参照图3,在第三实施例中,步骤S32包括:
步骤S33,当行为图像与参考图像的差异值低于预设阀值时,采集对应各个预设时间段行为图像所对应的人脸表情图像;
步骤S34,将采集的人脸表情图像与系统预设表情进行比对,当用户的表情与预设表情匹配时,标记对应预设时间段的用户特征为吸引类特征。
当行为图像与参考图像的差异值低于预设阀值时,采集对应各个预设时间段行为图像所对应的人脸表情图像;将采集的人脸表情图像与系统预设表情进行比对,当用户的表情与预设表情匹配时,标记对应预设时间段的用户特征为吸引类特征。由于当行为图像与参考图像的差异值低于预设阀值时,用户对广告的满意度不一定高,如用户可能在闭眼休息,或者脸上出现对广告的嫌弃等各种表情,为提高准确度,采集对应各个预设时间段行为图像所对应的人脸表情图像,并与系统预设表情进行比对,系统预设表情为用户对广告感兴趣或者被广告吸引的表情,因而当用户的表情与预设表情匹配时,即两者表情都是高兴,或者睁大眼睛盯着屏幕,标记对应预设时间段的用户特征为吸引类特征。
在本实施例中,通过当行为图像与参考图像的差异值低于预设阀值时,采集对应各个预设时间段行为图像所对应的人脸表情图像;将采集的人脸表情图像与系统预设表情进行比对,当用户的表情与预设表情匹配时,标记对应预设时间段的用户特征为吸引类特征。由于在获取表情的基础上对用户特征进行判断,因而能够更准确得知用户对广告的满意度,有利于商家进一步改进广告,提升广告效益。
进一步地,在本发明广告质量评估方法第一实施例的基础上,提出广告质量评估方法第四实施例,参照图4,在第四实施例中,所述广告质量评估方法还包括:
步骤S50,当检测到预设输入设备发送的中断广告指令时,根据所述中断广告事件标记对应预设时间段的用户特征;
步骤S60,根据所述用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。
当检测到预设输入设备发送的中断广告指令时,此时用户一般对该对应时间段广告不满意或者不感兴趣,根据所述中断广告事件标记对应预设时间段的用户特征,根据所述用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。
在本实施例中,通过当检测到预设输入设备发送的中断广告指令时,根据所述中断广告事件标记对应预设时间段的用户特征;根据所述用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。由于考虑输入设备发送的中断广告情况,因而更能全面对广告质量进行评估。
进一步地,在本发明广告质量评估方法第一实施例的基础上,提出广告质量评估方法第五实施例,在第五实施例中,所述获取所述比对图像与参考图像之间的差异值步骤包括:
获取参考图像与采集的各幅行为图像中用户头部中心点所处的位置信息,根据所述用户头部中心点的位置信息获取所述比对图像与参考图像之间的差异值。
由于当用户头部中心点所处的位置信息变化时,用户的行为一般会发生变化,因而根据所述用户头部中心点的位置信息获取所述比对图像与参考图像之间的差异值。
在本实施例中,通过获取参考图像与采集的各幅行为图像中用户头部中心点所处的位置信息,根据所述用户头部中心点的位置信息获取所述比对图像与参考图像之间的差异值。由于根据所述用户头部中心点的位置信息获取所述比对图像与参考图像之间的差异值,因而能够准确获取对应预设时间段的用户特征,为准确获取用户对广告满意度提供基础。
进一步地,在本发明广告质量评估方法第一实施例的基础上,提出广告质量评估方法第六实施例,在第六实施例中,所述广告质量评估方法还包括:
在广告各个预设时间段内获取预设数量用户的用户特征,根据所述预设数量用户的用户特征映射所述各个预设时间段的广告质量,以供评估人员进行评估。
增大用户统计样本,提高广告特征时间片的准确性,例如:针对一个广告,统计成千上万用户的反馈。统计大多数用户共同标注的吸引类特征时间片,这样得到的广告效益更高,也可对广告进行自适应性改进,如先在一个地方区域内投放广告,针对广告的排斥类时间片改进后,再次进行投放,继续收集改进后的广告特征,再对排斥类时间片进行改进,然后再次投放。经过多次提炼后,得到一个内容引人入胜且时间长度恰当的广告, 最终投放到全国区域内,这样能保证广告在第一次大规模投放时给用户一个良好的体验。
在本实施例中,通过在广告各个预设时间段内获取预设数量用户的用户特征,根据所述预设数量用户的用户特征映射所述各个预设时间段的广告质量,以供评估人员进行评估。由于获取预设数量的用户对各个预设时间段的满意度,因而排除了偶然因素,提升了准确度,有益于进一步提高商家广告的效益。
本发明还提供广告质量评估方法装置,在基于用户行为的广告质量评估装置第一实施例中,参照图5,该装置包括:
拍摄模块10,用于拍摄广告播放初始时刻用户的行为图像,并将该行为图像作为参考图像;
当开始放广告时,记录此时时间为T1, 初始化并打开智能设备摄像头,初始化过程降底图像的分辨率并增大图片的尺寸,这样利于检测人的位移以及人肢体语言的变化,以便根据人的位移以及人肢体语言的变化对图片进行分析,由于在广告播放前,用户一般在观看自身感兴趣的各项目标内容或者打开智能终端相应应用以观看目标内容,因而在广告播放初始时刻用户做出对广告内容感兴趣的行为,如靠近智能终端,拍摄广告播放初始时刻用户的行为图像,因而拍摄出来的图片可是用户头部或者肢体处在图片中心位置,将该图像作为判断用户行为的参考模板P,即为参考图像,然后开始捕捉用户观看动画时的图像。
采集模块20,用于在广告播放过程中,每间隔预设时间段采集用户的行为图像,并将所采集的各幅行为图像与参考图像进行比对;
在广告播放过程中,即是当广告持续进行时,每间隔预设时间段采集用户的行为图像,记录采集图像的时刻T1,T2,T3…Tn,采集的行为图像标记为P1,P2,P3 … Pn,每张行为图像记录该预设时间段用户的行为特征,因而把整个广告播放时间切分为n个小时间片,所述预设时间段标记对应的广告时段为A1,A2,A3…An,将所采集的各幅行为图像与参考图像进行比对,所述比对除比对图像中人的位移以及人肢体语言的变化,还包括对用户摇头,离开等动作的判断,当用户出现大的位移变化即远离相应广告播放终端时,用户在采集图像上的位置信息会发生变化,如用户不在采集图像上或者用户在所采集图像的边缘位置,而当用户出现偏头等动作时,其头部在图像上所处的位置信息也会发生变化。
第一标记模块30,用于获取所述行为图像与参考图像的差异值,根据行为图像与参考图像的差异值标记对应预设时间段的用户特征;
获取所述行为图像与参考图像的差异值,如获取参考图像中用户头部所处的位置信息,所述位置信息包括用户头部中心点在图像中的横坐标与纵坐标信息,参考图像中用户头部中心点所处的位置可在图像的正中央,获取行为图像中用户头部所处的位置信息,即获取行为图像中用户头部中心点在图片中的横坐标与纵坐标信息,获取所述行为图像与参考图像的差异值包括获取用户头部中心点在参考图像与行为图像中位置信息的差异值或者用户肢体中心点在在参考图像与行为图像中位置信息的差异值,根据行为图像与参考图像的差异值标记对应预设时间段的用户特征,所述用户特征包括吸引类特征与排斥类特征,当用户特征为吸引类特征时,表示用户对广告关注度很高,例如用户没有发生动作,始终注视着广告,当用户特征为排斥类特征时,表示广告对用户缺乏吸引力,例如摇头,用户离开,用户使用遥控器终止广告播放等,当差异值超过预设阀值时,则判断用户行为发生了变化。
第一映射模块40,用于将各个预设时间段的用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。
将各个预设时间段的用户特征映射所述预设时间段的广告质量,当预设时间段内用户特征为吸引类特征时,说明该预设时间段内广告质量高,用户对所投放广告的满意度高,广告商可以继续沿用该段时间段的广告,当预设时间段内用户特征为排斥类特征时,说明用户对所投放广告的满意度低,即是各个预设时间段的用户特征映射所述预设时间段的广告质量,广告商据此对广告进行改进或裁剪该预设时间段用户不感兴趣或者满意度低的广告时段。
在本实施例中,通过拍摄模块10,用于拍摄广告播放初始时刻用户的行为图像,并将该行为图像作为参考图像;采集模块20,用于在广告播放过程中,每间隔预设时间段采集用户的行为图像,并将所采集的各幅行为图像与参考图像进行比对;第一标记模块30,用于获取所述行为图像与参考图像的差异值,根据行为图像与参考图像的差异值标记对应预设时间段的用户特征;第一映射模块40,用于将各个预设时间段的用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。由于每间隔预设时间段采集用户的行为图像,根据行为图像与参考图像的差异值与差异值对应的用户特征映射所述预设时间段的广告质量,因而能够快速得知用户对各个预设时间段广告的满意度,将用户对各个预设时间段的满意度反馈给广告商,广告商根据所述反馈改进相应广告,进而提升广告效率。
进一步地,在本发明广告质量评估装置第一实施例的基础上,提出广告质量评估装置第二实施例,参照图6,在第二实施例中,第一标记模块包括:
第一标记单元31,用于获取所述行为图像与参考图像的差异值,当行为图像与参考图像的差异值超过第一预设阀值时,标记对应预设时间段的用户特征为排斥类特征;
第二标记单元32,用于当行为图像与参考图像差异值低于第一预设阀值时,标记对应预设时间段的用户特征为吸引类特征。
获取所述行为图像与参考图像的差异值,行为图像与参考图像的差异值超过第一预设阀值时,标记对应预设时间段的用户特征为排斥类特征,如图片P和Pt差异较大,差异值大于40%,则认为用户离开了,利用播广告的时间休息一下,此时标记对应预设时间段的用户特征为排斥类特征,而用户没有发生动作,则差异小于阀值%3,则图片P和Pt差异较小,差异大于阀值%3小于%40,也认为用户没有发生动作,设置阀值是为了减少误判,比如用户眨了眨眼睛等不会认为用户特征为排斥类特征。
在本实施例中,通过第一标记单元31,用于获取所述行为图像与参考图像的差异值,当行为图像与参考图像的差异值超过第一预设阀值时,标记对应预设时间段的用户特征为排斥类特征;第二标记单元32,用于当行为图像与参考图像差异值低于第一预设阀值时,标记对应预设时间段的用户特征为吸引类特征。由于当行为图像与参考图像的差异值超过第一预设阀值时,标记对应预设时间段的用户特征为排斥类特征;当行为图像与参考图像差异值低于第一预设阀值时,标记对应预设时间段的用户特征为吸引类特征,即是对用户特征进行准确判断,而对用户特征进行准确判断是准确获取用户对各个预设时间段广告满意度的基础,从而有利于商家进一步改进广告,提升广告效益。
进一步地,在本发明广告质量评估装置第二实施例的基础上,提出广告质量评估装置第三实施例,参照图7,在第三实施例中,所述第二标记单元包括:
采集子单元33,用于当行为图像与参考图像的差异值低于预设阀值时,采集对应各个预设时间段行为图像所对应的人脸表情图像;
比对子单元34,用于将采集的人脸表情图像与系统预设表情进行比对,当用户的表情与预设表情匹配时,标记对应预设时间段的用户特征为吸引类特征。
在本实施例中,通过采集子单元33,用于当行为图像与参考图像的差异值低于预设阀值时,采集对应各个预设时间段行为图像所对应的人脸表情图像;比对子单元34,用于将采集的人脸表情图像与系统预设表情进行比对,当用户的表情与预设表情匹配时,标记对应预设时间段的用户特征为吸引类特征。由于在获取表情的基础上对用户特征进行判断,因而能够更准确得知用户对广告的满意度,有利于商家进一步改进广告,提升广告效益。
进一步地,在本发明广告质量评估装置第一实施例的基础上,提出广告质量评估装置第四实施例,参照图8,在第四实施例中,所述广告质量评估装置还包括:
检测模块50,用于当检测到预设输入设备发送的中断广告指令时,根据所述中断广告事件标记对应预设时间段的用户特征;
第二映射模块60,用于根据所述用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。
当检测到预设输入设备发送的中断广告指令时,此时用户一般对该对应时间段广告不满意或者不感兴趣,根据所述中断广告事件标记对应预设时间段的用户特征,根据所述用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。
在本实施例中,通过检测模块50,用于当检测到预设输入设备发送的中断广告指令时,根据所述中断广告事件标记对应预设时间段的用户特征;第二映射模块60,用于根据所述用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。由于考虑输入设备发送的中断广告情况,因而更能全面对广告质量进行评估。
进一步地,在本发明广告质量评估装置第一实施例的基础上,提出广告质量评估装置第五实施例,在第五实施例中,所述第一标记模块包括:
获取单元,用于获取参考图像与采集的各幅行为图像中用户头部中心点所处的位置信息,根据所述用户头部中心点的位置信息获取所述比对图像与参考图像之间的差异值。
由于当用户头部中心点所处的位置信息变化时,用户的行为一般会发生变化,因而根据所述用户头部中心点的位置信息获取所述比对图像与参考图像之间的差异值。
在本实施例中,通过获取单元,用于获取参考图像与采集的各幅行为图像中用户头部中心点所处的位置信息,根据所述用户头部中心点的位置信息获取所述比对图像与参考图像之间的差异值。由于根据所述用户头部中心点的位置信息获取所述比对图像与参考图像之间的差异值,因而能够准确获取对应预设时间段的用户特征,为准确获取用户对广告满意度提供基础。
进一步地,在本发明广告质量评估装置第一实施例的基础上,提出广告质量评估装置第六实施例,在第六实施例中,所述广告质量评估装置还包括:
第三映射模块,用于在广告各个预设时间段内获取预设数量用户的用户特征,根据所述预设数量用户的用户特征映射所述各个预设时间段的广告质量,以供评估人员进行评估。
在本实施例中,通过第三映射模块,用于在广告各个预设时间段内获取预设数量用户的用户特征,根据所述预设数量用户的用户特征映射所述各个预设时间段的广告质量,以供评估人员进行评估。由于获取预设数量的用户对各个预设时间段的满意度,因而排除了偶然因素,提升了准确度,有益于进一步提高商家广告的效益。
以上仅为本发明的优选实施例,并非因此限制本发明的专利范围,凡是利用本发明说明书及附图内容所作的等效结构或等效流程变换,或直接或间接运用在其他相关的技术领域,均同理包括在本发明的专利保护范围内。

Claims (19)

  1. 一种广告质量评估方法,其特征在于,所述广告质量评估方法包括:
    拍摄广告播放初始时刻用户的头部行为图像,并将该头部行为图像作为参考图像;
    在广告播放过程中,每间隔预设时间段采集用户的头部行为图像,并将所采集的各幅头部行为图像与参考图像进行比对;
    获取所述头部行为图像与参考图像的差异值,根据头部行为图像与参考图像的差异值标记对应预设时间段的用户特征;
    将各个预设时间段的用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。
  2. 如权利要求1所述的广告质量评估方法,其特征在于,所述获取所述头部行为图像与参考图像的差异值,根据头部行为图像与参考图像的差异值标记对应预设时间段的用户特征步骤包括:
    获取所述头部行为图像与参考图像的差异值,当头部行为图像与参考图像的差异值超过第一预设阀值时,标记对应预设时间段的用户特征为排斥类特征;
    当头部行为图像与参考图像差异值低于第一预设阀值时,标记对应预设时间段的用户特征为吸引类特征。
  3. 如权利要求2所述的广告质量评估方法,其特征在于,所述当头部行为图像与参考图像差异值低于第一预设阀值时,标记对应预设时间段的用户特征为吸引类特征步骤包括:
    当头部行为图像与参考图像的差异值低于预设阀值时,采集对应各个预设时间段头部行为图像所对应的人脸表情图像;
    将采集的人脸表情图像与系统预设表情进行比对,当用户的表情与预设表情匹配时,标记对应预设时间段的用户特征为吸引类特征。
  4. 一种广告质量评估方法,其特征在于,所述广告质量评估方法包括:
    拍摄广告播放初始时刻用户的行为图像,并将该行为图像作为参考图像;
    在广告播放过程中,每间隔预设时间段采集用户的行为图像,并将所采集的各幅行为图像与参考图像进行比对;
    获取所述行为图像与参考图像的差异值,根据行为图像与参考图像的差异值标记对应预设时间段的用户特征;
    将各个预设时间段的用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。
  5. 如权利要求4所述的广告质量评估方法,其特征在于,所述获取所述行为图像与参考图像的差异值,根据行为图像与参考图像的差异值标记对应预设时间段的用户特征步骤包括:
    获取所述行为图像与参考图像的差异值,当行为图像与参考图像的差异值超过第一预设阀值时,标记对应预设时间段的用户特征为排斥类特征;
    当行为图像与参考图像差异值低于第一预设阀值时,标记对应预设时间段的用户特征为吸引类特征。
  6. 如权利要求5所述的广告质量评估方法,其特征在于,所述当行为图像与参考图像差异值低于第一预设阀值时,标记对应预设时间段的用户特征为吸引类特征步骤包括:
    当行为图像与参考图像的差异值低于预设阀值时,采集对应各个预设时间段行为图像所对应的人脸表情图像;
    将采集的人脸表情图像与系统预设表情进行比对,当用户的表情与预设表情匹配时,标记对应预设时间段的用户特征为吸引类特征。
  7. 如权利要求4所述的广告质量评估方法,所述广告质量评估方法还包括:
    当检测到预设输入设备发送的中断广告指令时,根据所述中断广告事件标记对应预设时间段的用户特征;
    根据所述用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。
  8. 如权利要求5所述的广告质量评估方法,所述广告质量评估方法还包括:
    当检测到预设输入设备发送的中断广告指令时,根据所述中断广告事件标记对应预设时间段的用户特征;
    根据所述用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。
  9. 如权利要求6所述的广告质量评估方法,所述广告质量评估方法还包括:
    当检测到预设输入设备发送的中断广告指令时,根据所述中断广告事件标记对应预设时间段的用户特征;
    根据所述用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。
  10. 如权利要求4所述的广告质量评估方法,其特征在于,所述获取所述比对图像与参考图像之间的差异值步骤包括:
    获取参考图像与采集的各幅行为图像中用户头部中心点所处的位置信息,根据所述用户头部中心点的位置信息获取所述比对图像与参考图像之间的差异值。
  11. 如权利要求4所述的广告质量评估方法,其特征在于,所述广告质量评估方法还包括:
    在广告各个预设时间段内获取预设数量用户的用户特征,根据所述预设数量用户的用户特征映射所述各个预设时间段的广告质量,以供评估人员进行评估。
  12. 一种广告质量评估装置,其特征在于,所述广告质量评估装置包括:
    拍摄模块,用于拍摄广告播放初始时刻用户的行为图像,并将该行为图像作为参考图像;
    采集模块,用于在广告播放过程中,每间隔预设时间段采集用户的行为图像,并将所采集的各幅行为图像与参考图像进行比对;
    第一标记模块,用于获取所述行为图像与参考图像的差异值,根据行为图像与参考图像的差异值标记对应预设时间段的用户特征;
    第一映射模块,用于将各个预设时间段的用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。
  13. 如权利要求12所述的广告质量评估装置,其特征在于,所述第一标记模块包括:
    第一标记单元,用于获取所述行为图像与参考图像的差异值,当行为图像与参考图像的差异值超过第一预设阀值时,标记对应预设时间段的用户特征为排斥类特征;
    第二标记单元,用于当行为图像与参考图像差异值低于第一预设阀值时,标记对应预设时间段的用户特征为吸引类特征。
  14. 如权利要求12所述的广告质量评估装置,其特征在于,所述第二标记单元包括:
    采集子单元,用于当行为图像与参考图像的差异值低于预设阀值时,采集对应各个预设时间段行为图像所对应的人脸表情图像;
    比对子单元,用于将采集的人脸表情图像与系统预设表情进行比对,当用户的表情与预设表情匹配时,标记对应预设时间段的用户特征为吸引类特征。
  15. 如权利要求12所述的广告质量评估装置,所述广告质量评估装置还包括:
    检测模块,用于当检测到预设输入设备发送的中断广告指令时,根据所述中断广告事件标记对应预设时间段的用户特征;
    第二映射模块,用于根据所述用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。
  16. 如权利要求13所述的广告质量评估装置,所述广告质量评估装置还包括:
    检测模块,用于当检测到预设输入设备发送的中断广告指令时,根据所述中断广告事件标记对应预设时间段的用户特征;
    第二映射模块,用于根据所述用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。
  17. 如权利要求14所述的广告质量评估装置,所述广告质量评估装置还包括:
    检测模块,用于当检测到预设输入设备发送的中断广告指令时,根据所述中断广告事件标记对应预设时间段的用户特征;
    第二映射模块,用于根据所述用户特征映射所述预设时间段的广告质量,以供评估人员进行评估。
  18. 如权利要求12所述的广告质量评估装置,其特征在于,所述第一标记模块包括:
    获取单元,用于获取获取参考图像与采集的各幅行为图像中用户头部中心点所处的位置信息,根据所述用户头部中心点的位置信息获取所述比对图像与参考图像之间的差异值。
  19. 如权利要求12所述的广告质量评估装置,其特征在于,所述广告质量评估装置还包括:
    第三映射模块,用于在广告各个预设时间段内获取预设数量用户的用户特征,根据所述预设数量用户的用户特征映射所述各个预设时间段的广告质量,以供评估人员进行评估。
PCT/CN2016/112595 2016-11-16 2016-12-28 广告质量评估方法及装置 WO2018090447A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201611027268.X 2016-11-16
CN201611027268.XA CN106779802A (zh) 2016-11-16 2016-11-16 广告质量评估方法及装置

Publications (1)

Publication Number Publication Date
WO2018090447A1 true WO2018090447A1 (zh) 2018-05-24

Family

ID=58971206

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/112595 WO2018090447A1 (zh) 2016-11-16 2016-12-28 广告质量评估方法及装置

Country Status (2)

Country Link
CN (1) CN106779802A (zh)
WO (1) WO2018090447A1 (zh)

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111177556A (zh) * 2019-12-30 2020-05-19 绍兴市上虞区理工高等研究院 一种科技成果转化系统及方法
CN111210276A (zh) * 2020-01-08 2020-05-29 秒针信息技术有限公司 一种广告播放策略调整方法、装置、电子设备及介质
CN117499477A (zh) * 2023-11-16 2024-02-02 北京易华录信息技术股份有限公司 一种基于大模型训练的信息推送方法以及系统
CN117499477B (zh) * 2023-11-16 2024-06-07 北京易华录信息技术股份有限公司 一种基于大模型训练的信息推送方法以及系统

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN109003141A (zh) * 2018-07-30 2018-12-14 杨超 一种户外媒体投放价值评估方法
CN109345953A (zh) * 2018-08-28 2019-02-15 信利光电股份有限公司 基于广告机的广告推送方法及系统、存储介质和移动终端
CN109685611A (zh) * 2018-12-15 2019-04-26 深圳壹账通智能科技有限公司 一种产品推荐方法、装置、计算机设备及存储介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090025026A1 (en) * 2007-07-19 2009-01-22 Cisco Technology, Inc. Conditional response signaling and behavior for ad decision systems
CN103154953A (zh) * 2010-09-30 2013-06-12 阿弗科迪瓦公司 测量具有上网功能的应用程序的情感数据
CN103377293A (zh) * 2013-07-05 2013-10-30 河海大学常州校区 多源输入、信息智能优化处理的全息触摸交互展示系统
CN104298682A (zh) * 2013-07-18 2015-01-21 广州华久信息科技有限公司 一种基于人脸表情图像的信息推荐效果的评价方法及手机

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1372214A (zh) * 2001-02-19 2002-10-02 意蓝科技股份有限公司 使用者行为导向的网络行销系统及其处理方法
CN102881239A (zh) * 2011-07-15 2013-01-16 鼎亿数码科技(上海)有限公司 基于图像识别的广告投播系统及方法
US20140188528A1 (en) * 2012-12-31 2014-07-03 Stubhub, Inc. Customized Advertisement for Venue Seat Map
CN103324283A (zh) * 2013-05-23 2013-09-25 广东欧珀移动通信有限公司 基于人脸识别控制视频播放的方法及其终端
CN104252679A (zh) * 2013-06-30 2014-12-31 北京百度网讯科技有限公司 一种品牌广告评估体系的构建方法及系统

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090025026A1 (en) * 2007-07-19 2009-01-22 Cisco Technology, Inc. Conditional response signaling and behavior for ad decision systems
CN103154953A (zh) * 2010-09-30 2013-06-12 阿弗科迪瓦公司 测量具有上网功能的应用程序的情感数据
CN103377293A (zh) * 2013-07-05 2013-10-30 河海大学常州校区 多源输入、信息智能优化处理的全息触摸交互展示系统
CN104298682A (zh) * 2013-07-18 2015-01-21 广州华久信息科技有限公司 一种基于人脸表情图像的信息推荐效果的评价方法及手机

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111177556A (zh) * 2019-12-30 2020-05-19 绍兴市上虞区理工高等研究院 一种科技成果转化系统及方法
CN111210276A (zh) * 2020-01-08 2020-05-29 秒针信息技术有限公司 一种广告播放策略调整方法、装置、电子设备及介质
CN111210276B (zh) * 2020-01-08 2023-12-08 秒针信息技术有限公司 一种广告播放策略调整方法、装置、电子设备及介质
CN117499477A (zh) * 2023-11-16 2024-02-02 北京易华录信息技术股份有限公司 一种基于大模型训练的信息推送方法以及系统
CN117499477B (zh) * 2023-11-16 2024-06-07 北京易华录信息技术股份有限公司 一种基于大模型训练的信息推送方法以及系统

Also Published As

Publication number Publication date
CN106779802A (zh) 2017-05-31

Similar Documents

Publication Publication Date Title
WO2018090447A1 (zh) 广告质量评估方法及装置
WO2016017987A1 (en) Method and device for providing image
WO2013139239A1 (en) Method for recommending users in social network and the system thereof
WO2019080406A1 (zh) 电视机语音交互方法、语音交互控制装置及存储介质
WO2016091011A1 (zh) 字幕切换方法及装置
WO2016201745A1 (zh) 基于用户终端的就诊提示方法、用户终端和网络医院平台
WO2018166224A1 (zh) 全景视频的目标追踪显示方法、装置及存储介质
WO2018018681A1 (zh) 视频节目预览方法及装置
WO2017177524A1 (zh) 音视频同步播放的方法及装置
WO2014187158A1 (zh) 终端数据云分享的控制方法、服务器及终端
WO2019076087A1 (zh) 电视机及其显示图效控制方法、计算机可读存储介质
WO2017164640A1 (en) System and method for editing video contents automatically technical field
WO2018045682A1 (zh) 音画同步测试方法及装置
WO2018036057A1 (zh) 软件后台自适应升级方法及装置
WO2017121066A1 (zh) 应用程序显示方法和系统
WO2021261830A1 (en) Video quality assessment method and apparatus
WO2017090833A1 (en) Photographing device and method of controlling the same
WO2019010926A1 (zh) 广告的推送方法、装置及计算机可读存储介质
WO2022025423A1 (en) Video quality assessing method and apparatus
WO2017148035A1 (zh) 图像处理方法及装置
WO2019085543A1 (zh) 电视机系统及电视机控制方法
WO2018023925A1 (zh) 拍摄方法及系统
WO2017190452A1 (zh) 背光自适应调节方法及装置
WO2019137016A1 (zh) 电视节目推荐方法、设备及计算机可读存储介质
WO2017084297A1 (zh) 多媒体文件的播放方法及系统

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16921801

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 18/10/2019)

122 Ep: pct application non-entry in european phase

Ref document number: 16921801

Country of ref document: EP

Kind code of ref document: A1