CN104766230A - Advertising effect evaluation method based on human skeletal tracking - Google Patents
Advertising effect evaluation method based on human skeletal tracking Download PDFInfo
- Publication number
- CN104766230A CN104766230A CN201510198748.1A CN201510198748A CN104766230A CN 104766230 A CN104766230 A CN 104766230A CN 201510198748 A CN201510198748 A CN 201510198748A CN 104766230 A CN104766230 A CN 104766230A
- Authority
- CN
- China
- Prior art keywords
- skeleton
- evaluation method
- factor
- trail
- advertising results
- Prior art date
- Legal status (The legal status 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 status listed.)
- Pending
Links
Abstract
The invention relates to an advertising effect evaluation method based on human skeletal tracking. The method comprises the steps that a Kinect controller is used for acquiring images and sending acquired image data into a processing device for identity recognition; the obtained image data are preprocessed and subjected to face detection, and information processing is conducted according to the result of face detection; the audience information is tracked, recorded and stored in a processor database, the acquired audience information is classified, and the advertising effect is evaluated through a factor analysis model according to the audience information. By means of the method, the dynamic advertising condition can be grasped at any time.
Description
Technical field
The present invention relates to intelligent advertisement application and effect assessment technical field thereof, particularly relate to a kind of advertising results evaluation method of following the trail of based on skeleton.
Background technology
No matter 21 century is Internet advertising, or traditional TV, open air, house ads, all experienced by an explosive growth, advertisement has become ubiquitous.In recent years, outdoor advertising is with its high propagation arrival rate, and low propagation cost and the adaptability to modern's lifestyle change, by advertising media's industry widespread use.With regard to China's As-Is analysis, the second largest advertising media being only second to TV has been grown in outdoor advertising in China, accounts for 25% of the total share of Chinese advertisement.The develop rapidly of Chinese Urbanization's and the expansion of Urban Transport Terminals, make outdoor advertising obtain the accreditation of more personages.
Nowadays, advertising sector is just by " assembling a crowd " future development towards " Focus ", advertisement putting is had to the requirement of " essence " and " standard ", but conventional advertisement mediums and most of new media have, and form is outmoded, advertisement putting inefficiency, advertising results cannot monitor this three main drawback.
1) form is outmoded.
Increasing audient, creates aestheticly tired to conventional ads pattern.Tight knot is played more and more faster, house ads under most lines, viewing of cannot attractively having stopped.
2) advertising efficiency is low.
Conventional ads can only rely on one way propagation mode, passes on ad content to audient, rely on repeatedly repeat bomb, impel the memory of user's passive generation brand, advertising efficiency is extremely low.
3) advertising results cannot be monitored.
The great number advertising expenditure of continuous expansion, but cannot monitor the input effect of advertisement, and experience and macro-data model can only be relied on to estimate, cannot adjust advertisement in time and control, this be also annoying one of important worry of numerous advertisers.
From current advertising market form, a kind of novel intelligent advertising message platform of tool advertising results Function of Evaluation, effectively can solve the deficiency that current advertisement exists, and have good using value.At present, this field of Evaluation Model on Effectiveness of intelligent advertisement starts to occur some Preliminary Study Results, but still rare and all slightly not enough.
Summary of the invention
Technical matters to be solved by this invention is to provide a kind of advertising results evaluation method of following the trail of based on skeleton, and it is dynamic to grasp advertising at any time.
The technical solution adopted for the present invention to solve the technical problems is: provide a kind of advertising results evaluation method of following the trail of based on skeleton, comprise the following steps:
(1) adopt Kinect somatosensory controller to carry out the collection of image, and the view data of acquisition feeding treatment facility is carried out identification;
(2) view data obtained is carried out pre-service and carry out Face datection, and carry out information processing according to the result of Face datection;
(3) while, track record audience information, is stored in processor data storehouse, then is classified by the audience information got, and carried out advertising results evaluation according to audience information according to Factor Analysis Model.
The collection adopting Kinect somatosensory controller to carry out image in described step (1) specifically comprises: extract human body and drive skeleton; Extract the attitude skeleton of object; Realize the synchronous of human body attitude and gestures of object; Deformation of body is driven according to the driving skeleton after synchronous with gestures of object.
The mode of active, RFID reader or Quick Response Code scanning is adopted in described step (1) to identify.
In described step (2), Image semantic classification comprises: face righting, the enhancing of facial image and normalized.
In described step (2), Face datection is according to normalized facial image, by the method for knowledge or statistics to face modeling, extract facial characteristics, catch micromechanism, the matching degree of the faceform built in object more to be detected and intelligent database, thus judgement facial expression, return recognition of face information.
Information processing in described step (2) comprises the recognition of face information according to returning, and will integrate the data of Face datection, according to the broadcasting scheme preset, automatic switching advertisement also adds up broadcasting time.
Described Factor Analysis Model is R type Factor analysis, and its matrix form is: x=AF+e, and wherein, F is the common factor of x or the latent factor, and matrix A is Factor load-matrix, and e is the specific factor of x.
Beneficial effect
Owing to have employed above-mentioned technical scheme, the present invention compared with prior art, there is following advantage and good effect: the advertising results evaluation model followed the trail of based on skeleton that the present invention adopts, data message complicated in the various mistake such as audience information, environmental information, people information obtained in Appropriate application advertising platform operational process, in conjunction with the Evaluation Model on Effectiveness in the present invention, can make advertiser need not spended time, manpower, financial resources in addition, just can grasp at any time advertising dynamically, change advertisement putting scheme.
Accompanying drawing explanation
Fig. 1 is human skeleton figure;
Fig. 2 is process flow diagram of the present invention;
Fig. 3 is advertising results feedback data model schematic.
Embodiment
Below in conjunction with specific embodiment, set forth the present invention further.Should be understood that these embodiments are only not used in for illustration of the present invention to limit the scope of the invention.In addition should be understood that those skilled in the art can make various changes or modifications the present invention, and these equivalent form of values fall within the application's appended claims limited range equally after the content of having read the present invention's instruction.
Embodiments of the present invention relate to a kind of advertising results evaluation method of following the trail of based on skeleton, and the collection of image is responsible for by Kinect somatosensory controller, and the data obtained are sent into treatment facility; Treatment facility is computing machine or embedded pc system, the data obtained from collecting device are responsible for carry out image procossing, analytic operation, broadcasting/switching that electric advertisement screen realizes ad content is controlled according to the operation result of advertising results evaluation model, and uploaded onto the server analyzing the information obtained by network, feed back to processing data information center (advertisement manufacturer, survey institute etc.) as required.Its course of work comprises following three steps.
1. identification
Audient enters in identification range, and system utilizes identification equipment initiatively to carry out contactless automatic identification to it, according to different audient's types, initiatively chooses dissimilar advertisement and plays, the attention rate of audient within the scope of raising.
RFID reader: the electronic tag had according to audient, active Intelligent Recognition is carried out to Audience's identity: after label enters magnetic field, receive the radiofrequency signal that plug-in reader sends, the energy obtained by means of induction current sends out storage information in the chips, or initiatively sends the signal of a certain frequency.RFID reader reads information and after decoding, deliver to center of inside infosystem and carry out relevant data process, and treatment facility relevant data being fed back to intelligent advertisement platform carries out identity Intelligent Recognition.
Quick Response Code scans: by the Quick Response Code pattern on paper/electronic curtain, scanned by the RGB camera on Kinect, to be input in treatment facility and to use embedded software to decode.The data message obtained after utilizing decoding carries out Intelligent Recognition.
Human skeleton tracer technique: Kinect has two depth transducers---infrared depth camera.Utilize the information that camera obtains, fusing digital images technology phase, in conjunction with the relevant algorithm for pattern recognition of independent development, follow the trail of human skeleton, obtain the volume coordinate of human body 20 key position skeleton points, extract skeleton information.Obtain the volume coordinate of key position skeleton point again, formed and drive skeleton, extract the key point of object to be synchronized in advertisement, generate the attitude skeleton of object, as accompanying drawing 1.Finally, by affined transformation, driven by human body skeleton to overlap with gestures of object skeleton, make the synchronous of human body attitude and object, realize the function of the driving skeleton driven deformation of body of human body, reach the effect of " servo-actuated ".
Concrete steps are as follows:
1) extract human body and drive skeleton.The driving skeleton of human motion is obtained by sensor.
2) the attitude skeleton of object is extracted.The key point of extract body Model, generates the attitude skeleton of object.
3) human body attitude and gestures of object is synchronous.By affined transformation, driven by human body skeleton to overlap with object skeleton pose, namely realize the synchronous of human body attitude and gestures of object.
4) the driving skeleton after synchronous with gestures of object, drives deformation of body.
After entering interaction mode, in order to make to treat that mutual object and human body are synchronized with the movement under the driving driving skeleton in advertisement, need to allow object lattice along with driving the translation of skeleton, rotation and produce motion.This needs the grid vertex belonging to object to be distributed to the bone affecting its distortion, and arranges weighted value to realize, and then realizes the interlock effect of human body and object.
2. recognition of face
Comprise Image semantic classification, Face datection, information processing three steps.
Image semantic classification: RGB camera carries out facial image acquisition to current audience, at pretreatment stage, is optimized image, removes as far as possible or reduces the interference to pending image such as illumination, imaging system, external environment condition, for subsequent treatment improves quality.To make different facial images complete feature extraction, training and identification as far as possible under identical conditions.Image semantic classification mainly comprises face righting, the enhancing of facial image, and the work such as normalization.Face righting is the facial image rectified to obtain face location, namely makes it forward in the face of camera; Image enhaucament is the quality in order to improve facial image, not only visually picture rich in detail more, and makes image be more conducive to process and the identification of computing machine.The target of normalization work obtains consistent size, standardized face's image that gray scale span is identical.
Face datection: according to normalized facial image, by the method for knowledge or statistics to face modeling, extracts facial characteristics, catch micromechanism, the matching degree of the faceform built in object more to be detected and intelligent database, thus judge facial expression, return recognition of face information;
Information processing: the recognition of face information returned according to Face datection link, the data integrating Face datection said by processor, and according to the broadcasting scheme preset, automatic switching advertisement also adds up broadcasting time.
3. advertising results evaluation
While audient experiences advertising platform, track record audient number, information storage such as experience duration, interaction effect etc. are in processor data storehouse.Again the audience information got is classified, utilizes advertisement machine 3G wireless networking capabilities, by network by process after data upload to remote server.Processing data information center will obtain data message as required by remote server, and model carries out advertising results evaluation (as: advertiser can utilize mobile phone or computer designed according to this invention, the advertising results of each advertisement terminal are monitored by internet, timely adjustment ad content and strategy, realize terminal control).
In advertising platform operational process, understand the flow of the people in constantly macroscopical monitoring unit interval near billboard, pay close attention to the data messages such as the broadcasting time of number and interactive advertisement.And for single advertisement, in advertisement playing process, pay close attention to the data messages such as the sex of audient, age, interaction time, interaction scenario.The data message of this two type can be unified to upload to cloud server by network.Beyond the clouds, carried out the processes such as data analysis, effect assessment, input decision analysis by the algorithm developed, reach mass data precisely analyzed, evaluate, the effect of decision-making.Meanwhile, constantly by upgrade data stored in database, facilitate advertiser's administrative client to observe and change advertisement serving policy (as shown in Figure 2).
Advertising results evaluation index comprises following situation:
1. every thousand people's impression Cost Evaluations system (CPM)
Thousand people's impression costs (Cost Per Mille) of the web advertisement refer to the cost that the web advertisement produces 1000 impression.That is, advertiser throws in a web advertisement, obtain browsing of 1000 people, has impression, has the cognitive cost number that will pay to this advertisement.Thousand people's impression costs are usually using the exposure frequency of the advertisement place page as the index evaluated.
The computing formula of thousand people's impression costs (Cost Per Mille) of the web advertisement is:
CPM=total cost/(broadcasting time/1000)
The CPM numerical value drawn like this is less, show advertiser for the web advertisement to accept by 1,000 audients and the cost paid is lower; The numerical value of CPM is less, and the number of times that also just expression webpage is viewed is more, and the effect of the web advertisement is better.
2. often click Cost Evaluation system (CPC)
The often click cost (Cost Per Click) of the web advertisement refer to utilize the web advertisement clicked and the number of times being linked to address correlation or the detailed content page to weigh the effect of networking advertisement.
The account form often clicking cost (Cost Per Click) is:
CPC=total cost/total touching quantity
CPC assessment to be advertiser be that its advertisement of issuing obtains at every turn after audient clicks, the cost that should pay.The value of CPC is less, shows that the advertisement of advertiser obtainable clicking rate in the budget of advertising expenditure is higher, also just shows that the web advertisement that advertiser issues on the media is more by the welcome of consumer, obtains the concern of more audients.
3. often respond Cost Evaluation system evaluation and analysis (CPR)
What the evaluation index often responding cost (Cost Per Response) utilized is the response of network audience to the web advertisement, and more direct is exactly the interactive number of times with advertiser.
The account form often responding cost (Cost Per Response) utilizes network audience for the response number of times of the interactive link of the web advertisement to weigh advertising results.Pass through computing formula:
CPR=advertising expenditure total cost/audient responds total degree
Show that advertiser often obtains an audient and sends feedback opinion to its web advertisement, need the advertising expenditure paid.And the interaction times of audient and advertiser, online registration, on-line consulting should be comprised, fill in questionnaire etc.These are that client notes, clicks and the action just can taked after having read advertisement.Therefore, audient and the interaction index of advertiser are evaluation indexes more further than the evaluation criterion of CPM and CPC.CPR appraisement system is expected can reflect that consumer is for changing to a greater extent, and advertisement is experienced the most really, weighs the true effect of advertisement with this.
The present invention adopts the Evaluation Model on Effectiveness based on factor-analysis approach.In order to can the effect of reaction advertisement putting of science, all establish relevant data model both at home and abroad and advertisement delivery effect is evaluated.The advertisement of being broadcasted by intelligent advertisement display systems shows there is similar characteristic to Internet advertising, therefore throws in data model from the web advertisement and starts with and to evaluate advertisement delivery effect.
We have drawn the advantage of three kinds of advertising results evaluating data models, create the advertising results evaluating data model of laminating intelligent advertisement data platform.
Intelligent advertisement platform can obtain the record of advertising display, ad click, advertisement viewing time very easily, provide feedback interface easily to dislike the happiness of advertisement in order to evaluate audient simultaneously, therefore we excavate from both direction in length and breadth, establish following advertising results feedback data model (as shown in Figure 3), comprise the data of displaying, click, feedback three dimensions.
Three-dimensional data supports mutually, has objectively responded advertising results.Horizontal excavation is carried out to data simultaneously, make effect plays more bright and clear.
Evaluation Model on Effectiveness mainly adopts factor-analysis approach to set up.Factor analysis is from the relevant dependence in research variable inside, some complicated variablees is summed up as a kind of Multielement statistical analysis method of a few multi-stress (common factor).Its basic thought is: classified by observational variable, and correlativity is higher, namely contacts and divides more closely in same class, and therefore each class variable in fact just represents basic structure, i.e. a common factor.The object of factorial analysis attempts to describe by the linear function of the immesurable so-called common factor of minimum number and specific factor sum each component originally observed.
The basic step of factorial analysis:
1) correlation test
Because one of the main task of factorial analysis is by the information overlap extracting section in original variable and comprehensive composition-factor, and then finally realize the object reducing variable number.Therefore it requires there is stronger correlativity between original variable.Whether whether this step exists correlationship by the original variable of various methods analyst exactly, be applicable to carrying out factorial analysis.
2) factor is extracted
Original aggregation of variable is become the process of a few main gene, if after obtaining main cause subsolution, the Typical Representative variable of each main gene is not bery outstanding, also needs to obtain satisfied main gene by suitable rotation.
3) factors
After aggregation of variable being become a few factor, if the physical meaning of the factor is smudgy, be unfavorable for further analysis.Therefore this step makes the factor extracted have name interpretation by various method.
4) calculated factor score
After Factor Analysis Model is set up, also having an important effect to be that application factor analytical model goes to evaluate the status of each sample in whole model, carrying out comprehensive evaluation by calculating the score of each sample in each factor.
2. Factor Analysis Model describes
1) X=(x
1, x
2..., x
p) ¢ is Observable random vector, mean vector E (x)=0, covariance matrix Cov (X)=Σ, and covariance matrix ∑ equal with correlation matrix R (as long as variable standardization can be realized).
2) F=(F
1, F
2..., F
m) ¢ (m < p) is immesurable vector, its mean vector E (F)=0, covariance matrix Cov (F)=I, namely each component of vector is separate.
3) e=(e
1, e
2..., e
p) ¢ and F is separate, and E (e)=0, the covariance matrix ∑ of e is diagonal matrix, is namely separate between each component e, then model:
x
1=a
11F
1+a
12F
2+…+a
1mF
m+e
1
x
2=a
21F
1+a
22F
2+…+a
2mF
m+e
2
………
x
p=a
p1F
1+a
p2F
2+…+a
pmF
m+e
p
Be called Factor Analysis Model, because this model carries out for variable, each factor is again orthogonal, so also referred to as R type Factor analysis.
Its matrix form is: x=AF+e
Here,
1)m£p;
2) Cov (F, e)=0, namely F and e is incoherent;
3) F
1, F
2..., F
muncorrelated and variance is 1; e
1, e
2..., e
puncorrelated, and variance is different.
We call the common factor of X or the latent factor F, and matrix A is called Factor load-matrix, and e is called the specific factor of X.A=(a
ij), a
ijfor factor loading.Mathematically can prove, factor loading a
ijbe exactly the related coefficient of the i-th variable and the jth factor, reflect the importance of the i-th variable in the jth factor.
Be not difficult to find, the advertising results evaluation model followed the trail of based on skeleton that the present invention adopts, data message complicated in the various mistake such as audience information, environmental information, people information obtained in Appropriate application advertising platform operational process, in conjunction with the Evaluation Model on Effectiveness in the present invention, can make advertiser need not spended time, manpower, financial resources in addition, just can grasp at any time advertising dynamically, change advertisement putting scheme.
Claims (7)
1., based on the advertising results evaluation method that skeleton is followed the trail of, it is characterized in that, comprise the following steps:
(1) adopt Kinect somatosensory controller to carry out the collection of image, and the view data of acquisition feeding treatment facility is carried out identification;
(2) view data obtained is carried out pre-service and carry out Face datection, and carry out information processing according to the result of Face datection;
(3) while, track record audience information, is stored in processor data storehouse, then is classified by the audience information got, and carried out advertising results evaluation according to audience information according to Factor Analysis Model.
2. advertising results evaluation method of following the trail of based on skeleton according to claim 1, is characterized in that, the collection adopting Kinect somatosensory controller to carry out image in described step (1) specifically comprises: extract human body and drive skeleton; Extract the attitude skeleton of object; Realize the synchronous of human body attitude and gestures of object; Deformation of body is driven according to the driving skeleton after synchronous with gestures of object.
3. advertising results evaluation method of following the trail of based on skeleton according to claim 1, is characterized in that, adopts in described step (1) mode of active, RFID reader or Quick Response Code scanning to identify.
4. advertising results evaluation method of following the trail of based on skeleton according to claim 1, is characterized in that, in described step (2), Image semantic classification comprises: face righting, the enhancing of facial image and normalized.
5. advertising results evaluation method of following the trail of based on skeleton according to claim 4, it is characterized in that, in described step (2), Face datection is according to normalized facial image, by the method for knowledge or statistics to face modeling, extract facial characteristics, catch micromechanism, the matching degree of the faceform built in object more to be detected and intelligent database, thus judgement facial expression, return recognition of face information.
6. advertising results evaluation method of following the trail of based on skeleton according to claim 5, it is characterized in that, information processing in described step (2) comprises the recognition of face information according to returning, the data of Face datection will be integrated, according to the broadcasting scheme preset, automatic switching advertisement also adds up broadcasting time.
7. advertising results evaluation method of following the trail of based on skeleton according to claim 1, it is characterized in that, described Factor Analysis Model is R type Factor analysis, its matrix form is: x=AF+e, wherein, F is the common factor of x or the latent factor, and matrix A is Factor load-matrix, and e is the specific factor of x.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510198748.1A CN104766230A (en) | 2015-04-21 | 2015-04-21 | Advertising effect evaluation method based on human skeletal tracking |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510198748.1A CN104766230A (en) | 2015-04-21 | 2015-04-21 | Advertising effect evaluation method based on human skeletal tracking |
Publications (1)
Publication Number | Publication Date |
---|---|
CN104766230A true CN104766230A (en) | 2015-07-08 |
Family
ID=53648043
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510198748.1A Pending CN104766230A (en) | 2015-04-21 | 2015-04-21 | Advertising effect evaluation method based on human skeletal tracking |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN104766230A (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105447466A (en) * | 2015-12-01 | 2016-03-30 | 深圳市图灵机器人有限公司 | Kinect sensor based identity comprehensive identification method |
CN108255173A (en) * | 2017-12-20 | 2018-07-06 | 北京理工大学 | Robot follows barrier-avoiding method and device |
CN108734518A (en) * | 2018-05-22 | 2018-11-02 | 缪涵澄 | A method of counting advertising results using image recognition technology |
WO2019205538A1 (en) * | 2018-04-25 | 2019-10-31 | 华南理工大学 | Large-size screen targeted advertising system and method based on multi-source heterogeneous data analysis |
CN115426199A (en) * | 2022-11-03 | 2022-12-02 | 南京金易众和信息科技有限公司 | Website data security protection method |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030066071A1 (en) * | 2001-10-03 | 2003-04-03 | Koninklijke Philips Electronics N.V. | Program recommendation method and system utilizing a viewing history of commercials |
CN202183087U (en) * | 2011-07-21 | 2012-04-04 | 厦门市鼎朔信息技术有限公司 | Intelligent mirror plane touch display system |
CN102446327A (en) * | 2010-10-15 | 2012-05-09 | 汉王科技股份有限公司 | Advertisement delivery device and method based on face recognition |
US20130046637A1 (en) * | 2011-08-19 | 2013-02-21 | Firethorn Mobile, Inc. | System and method for interactive promotion of products and services |
CN103956128A (en) * | 2014-05-09 | 2014-07-30 | 东华大学 | Intelligent active advertising platform based on somatosensory technology |
-
2015
- 2015-04-21 CN CN201510198748.1A patent/CN104766230A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030066071A1 (en) * | 2001-10-03 | 2003-04-03 | Koninklijke Philips Electronics N.V. | Program recommendation method and system utilizing a viewing history of commercials |
CN102446327A (en) * | 2010-10-15 | 2012-05-09 | 汉王科技股份有限公司 | Advertisement delivery device and method based on face recognition |
CN202183087U (en) * | 2011-07-21 | 2012-04-04 | 厦门市鼎朔信息技术有限公司 | Intelligent mirror plane touch display system |
US20130046637A1 (en) * | 2011-08-19 | 2013-02-21 | Firethorn Mobile, Inc. | System and method for interactive promotion of products and services |
CN103956128A (en) * | 2014-05-09 | 2014-07-30 | 东华大学 | Intelligent active advertising platform based on somatosensory technology |
Non-Patent Citations (1)
Title |
---|
童晓萌: "网络广告媒体选择和效果的评价", 《统计教育》 * |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105447466A (en) * | 2015-12-01 | 2016-03-30 | 深圳市图灵机器人有限公司 | Kinect sensor based identity comprehensive identification method |
CN105447466B (en) * | 2015-12-01 | 2019-07-23 | 深圳市图灵机器人有限公司 | A kind of identity integrated recognition method based on Kinect sensor |
CN108255173A (en) * | 2017-12-20 | 2018-07-06 | 北京理工大学 | Robot follows barrier-avoiding method and device |
WO2019205538A1 (en) * | 2018-04-25 | 2019-10-31 | 华南理工大学 | Large-size screen targeted advertising system and method based on multi-source heterogeneous data analysis |
CN108734518A (en) * | 2018-05-22 | 2018-11-02 | 缪涵澄 | A method of counting advertising results using image recognition technology |
CN115426199A (en) * | 2022-11-03 | 2022-12-02 | 南京金易众和信息科技有限公司 | Website data security protection method |
CN115426199B (en) * | 2022-11-03 | 2023-10-10 | 南京金易众和信息科技有限公司 | Website data safety protection method |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108009871B (en) | Product recommendation based on augmented reality viewpoints | |
WO2019149005A1 (en) | Offline interactive advertisement system | |
CN103956128A (en) | Intelligent active advertising platform based on somatosensory technology | |
WO2018033156A1 (en) | Video image processing method, device, and electronic apparatus | |
JP6267861B2 (en) | Usage measurement techniques and systems for interactive advertising | |
CN104766230A (en) | Advertising effect evaluation method based on human skeletal tracking | |
US20160050465A1 (en) | Dynamically targeted ad augmentation in video | |
US20150220814A1 (en) | Behavioral event measurement system and related method | |
CN105809178A (en) | Population analyzing method based on human face attribute and device | |
CN103098079A (en) | Personalized program selection system and method | |
US20130166394A1 (en) | Saliency-based evaluation of webpage designs and layouts | |
CN105224075A (en) | Sensor-based mobile search, correlation technique and system | |
KR20070006671A (en) | Method and system for managing an interactive video display system | |
Mohammadi et al. | Region based image steganalysis using artificial bee colony | |
CN110415009A (en) | Computerized system and method for being modified in video | |
Tian et al. | Intelligent advertising framework for digital signage | |
US10600060B1 (en) | Predictive analytics from visual data | |
CN111724199A (en) | Intelligent community advertisement accurate delivery method and device based on pedestrian active perception | |
US20150363822A1 (en) | Splitting a purchase panel into sub-groups | |
US20160042362A1 (en) | Recording purchase behavior using device with multiple feedback and input modalities | |
Merino‐Gracia et al. | Real‐time text tracking in natural scenes | |
JP2021508400A (en) | System for providing fast and secure content | |
CN115601081B (en) | Online digital exhibition service management system based on big data analysis | |
US20220309279A1 (en) | Computerized system and method for fine-grained event detection and content hosting therefrom | |
CN110659923A (en) | Information display method and device for user terminal |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
EXSB | Decision made by sipo to initiate substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20150708 |
|
RJ01 | Rejection of invention patent application after publication |