CN104185041A - Video interaction advertisement automatic generation method and system - Google Patents
Video interaction advertisement automatic generation method and system Download PDFInfo
- Publication number
- CN104185041A CN104185041A CN201410167442.5A CN201410167442A CN104185041A CN 104185041 A CN104185041 A CN 104185041A CN 201410167442 A CN201410167442 A CN 201410167442A CN 104185041 A CN104185041 A CN 104185041A
- Authority
- CN
- China
- Prior art keywords
- advertisement
- user
- matching degree
- type
- video
- 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.)
- Granted
Links
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Two-Way Televisions, Distribution Of Moving Picture Or The Like (AREA)
Abstract
The invention relates to the technical field of video advertisement and discloses a video interaction advertisement automatic generation method and system. The method comprises the following steps: receiving interactive operation of a user on a video, and identifying the user; obtaining an image of a video object subjected to the interactive operation, and identifying the object; calculating matching degree of the video object and each advertisement in an advertisement bank; calculating matching degree of the user and each advertisement in the advertisement bank; carrying out calculating to obtain preference degree of each advertisement in the advertisement bank according to the matching degree of the video object and each advertisement in the advertisement bank, the matching degree of the user and each advertisement and priority of each advertisement; and displaying the advertisement, the preference degree of which is the highest, in the advertisement bank to the user. According to the video interaction advertisement automatic generation method and system, not only characteristics of the object in the video can be fully taken into consideration, but also the preference of the user can be fully taken into consideration, and the most favored advertisement of the video object subjected to the interactive operation can be displayed to the user, so that the effectiveness of the video interaction advertisement is very high.
Description
Technical field
The present invention relates to video ads technical field, relate in particular to a kind of generation method and system of video ads.
Background technology
The existing advertisement technology based on video content, is generally that the advertisement relevant to some object in video inserted to video, and this technology, in the time adding advertisement, although considered the characteristic of video, is not considered user's hobby.And because can not insert too much advertisement in a video, can only insert and the advertisement that in video, part object is relevant, so cannot insert and advertisement that in video, all objects are relevant.
Although existing video interactive advertisement technology has been considered user's hobby, only have and in the time that user clicks an object in video, just eject the advertisement relevant to this object, but existing video interactive Technology Need is manually by associated to the advertisement relevant to object in video and this object video, thereby in video playback, if user and this object video are mutual, could eject respective advertisement.But the mode of artificial associated video object, speed is slow, cost is high, so manual type can only be by part object association interactive advertisement in video, and these associated not necessarily interested advertisements of user of object of interactive advertisement.The object video that causes so possibly advertisement associated is all by user interactions, and user interactions object video there is no associated advertisement, be zero thereby cause the utilance of the interactive advertisement of this video.
Therefore, prior art has yet to be improved and developed.
Summary of the invention
The object of the present invention is to provide a kind of generation method and system of video ads, for the hobby that can not meet user in the existing advertisement technology based on video content, simultaneously insert the high deficiency of interactive advertisement cost for existing video interactive advertisement technology manual type, can be by the object video of Auto-matching and user interactive relevant and with user-dependent preferred advertisement and show user.
A generation method for video ads, comprises the following steps:
Accept the interactive operation of user to video, identify described user;
Obtain the image of the object video of described interactive operation, identify described object;
According to the matching degree of the recognition result of described object being calculated to each advertisement in described object video and advertisement base;
According to the matching degree of described user's recognition result being calculated to each advertisement in described user and advertisement base;
Calculate the preference of described each advertisement according to the priority of the matching degree of each advertisement in the matching degree of each advertisement in described object video and advertisement base, described user and advertisement base, each advertisement;
Give described user by the advertising display of preference maximum in described advertisement base, and user is recorded into advertisement base checking of advertisement.
Preferably, accept the interactive operation of user to video described, before identifying described user's step, described method is further comprising the steps of:
Build advertisement base, each advertisement in described advertisement base comprises field: type, priority, word, picture, prospective users type, user check record, and wherein user checks that record comprises user's ID, the time length of checking;
Each user in described advertisement base is checked to the ratio of checking time span that the time span of advertisement of each type and this user are total, the preference degree of the advertisement as described each user to described each type.
Wherein, preferably, described basis is calculated the step of the matching degree of each advertisement in described object video and advertisement base to the recognition result of described object, specifically comprise:
The picture of each advertisement in the image of described object video and advertisement base is carried out to fuzzy matching, and record the image of described object video and the matching degree of each advertisement;
The word of each advertisement in the type of described object video and advertisement base is carried out to fuzzy matching, and record the type of described object video and the matching degree of each advertisement;
The matching degree of the image of each advertisement and described object video, type is weighted on average, obtains the matching degree of described object video and each advertisement.
Wherein, preferably, described basis is calculated the step of the matching degree of each advertisement in described user and advertisement base to described user's recognition result, specifically comprise:
The prospective users type set of each advertisement in described user's type and described advertisement base is carried out to fuzzy matching, and record described user's type and the matching degree of each advertisement;
Described user is checked to the ratio of checking time span that the time span of advertisement of each type and this user are total, the preference degree of the advertisement as this user to each type;
Obtain the type of each advertisement in advertisement base, the preference degree of the type using described user to each advertisement is as the matching degree of the hobby of described user and each advertisement;
The matching degree of the matching degree of each advertisement and described user's type, hobby is weighted on average, obtains the matching degree of each advertisement and described user in described advertisement base.
Wherein, preferably, the described priority according to the matching degree of each advertisement in the matching degree of each advertisement in described object video and advertisement base, described user and advertisement base, each advertisement calculates the step of the preference of described each advertisement, is specifically included as:
By each advertisement in described object video and described advertisement base and matching degree, described user and the matching degree of each advertisement, the priority of each advertisement be weighted on average, obtain the preference of each advertisement in described advertisement base.
A generation system for video ads, comprising:
Interactive module, for accepting the interactive operation of user to video, identifies described user;
Object type identification module, for obtaining the image of object video of described interactive operation, identifies described object;
Object video matching module, calculates the matching degree of described object video and the each advertisement of advertisement base for basis to the recognition result of described object;
User's matching module, calculates the matching degree of described user and the each advertisement of advertisement base for basis to described user's recognition result;
Preference computing module, for calculating the preference of each advertisement in described advertisement base according to the priority of the matching degree of each advertisement in the matching degree of described object video and the each advertisement of advertisement base, described user and advertisement base, each advertisement;
Display module, for giving described user by the advertising display of described advertisement base preference maximum, and records user into advertisement base to checking of advertisement.
Preferably, described system also comprises:
Advertisement base builds module, be used for building advertisement base, each advertisement in described advertisement base comprises field: type, priority, word, picture, prospective users type, user check record, and wherein user checks that record comprises user's ID, the time length of checking;
User preferences identification module, for the ratio of checking time span of user being checked to the time span of advertisement of each type and this user are total, the preference degree of the advertisement as this user to each type.
Wherein, preferably, described object video matching module comprises:
Object images matching unit, for the picture of the image of described object video and the each advertisement of advertisement base is carried out to fuzzy matching, and records the image of described object video and the matching degree of each advertisement;
Object type matching unit, for the word of the type of described object video and the each advertisement of advertisement base is carried out to fuzzy matching, and records the type of described object video and the matching degree of each advertisement;
Object matching degree computing unit, for the matching degree of the image of each advertisement and described object, type is weighted on average, obtains the matching degree of described object video and each advertisement.
Wherein, preferably, described user's matching module comprises:
User type matching unit, for the prospective users type set of described user's type and the each advertisement of described advertisement base is carried out to fuzzy matching, and records described user's type and the matching degree of each advertisement;
User preferences matching unit, for obtaining the type of each advertisement of advertisement base, the preference degree of the type using described user to each advertisement is as the matching degree of the hobby of described user and each advertisement;
User's matching degree computing unit, for the matching degree of the matching degree of each advertisement and described user's type, hobby is weighted on average, obtains the matching degree of each advertisement and described user in described advertisement base.
Wherein, preferably, described preference computing module, specifically for the priority of the matching degree of each advertisement and described object in described advertisement base, this advertisement and described user's matching degree, this advertisement is weighted on average, obtains the preference of each advertisement in described advertisement base.
The automatic generation method of video interactive advertisement of the present invention and system, Properties of Objects in video can be taken into full account, user's hobby can be taken into full account again, all by the object of user interactions in video, there is a most preferred advertisement can show this user, the not mutual object video of user can pop-up advertisement, makes the validity of video interactive advertisement very high; Simultaneously because add advertisement without manual, so speed is fast, cost is low.
Brief description of the drawings
The flow chart of the automatic generation method of the video interactive advertisement that Fig. 1 provides for the embodiment of the present invention;
Fig. 2 is the flow chart of step 30 in Fig. 1;
Fig. 3 is the flow chart of step 40 in Fig. 1;
The structured flowchart of the automatic creation system of the video interactive advertisement that Fig. 4 provides for the embodiment of the present invention;
Fig. 5 is the structured flowchart of the object video matching module in Fig. 4;
Fig. 6 is the structured flowchart of the user's matching module in Fig. 4.
Embodiment
In order to make object of the present invention, technical scheme and advantage clearer, below in conjunction with drawings and Examples, the present invention is further elaborated.Should be appreciated that specific embodiment described herein, only in order to explain the present invention, is not intended to limit the present invention.
As shown in Figure 1, the automatic generation method of the video interactive advertisement that the embodiment of the present invention provides, comprises the following steps:
Step S30, accepts the interactive operation of user to video, identifies described user.Particularly, user can be both that user utilizes mouse to carry out interactive operation to video to the operation of video, can be also the form of eye control, and for example user utilizes eyeball tracking device and video to carry out alternately.The mode of interactive operation comprises the most frequently used clicking operation.User's type comprises sex, age etc.Wherein, identification user's type is prior art, for example, identify according to user's log-on message, or differentiate user's type according to video frequency pick-up head, and it is prior art, is not described further herein.Wherein, identification user's hobby, in can the historical record that user checks advertisement from advertisement base, analysis mining obtains, for example, the adline that past user sees manyly or the time is longer is the type that user more likes, thereby by counting user to all types of advertisements check number of times or time span, can obtain the fancy grade of user to each series advertisements.
Step S40, obtains the image of the object video of described interactive operation, identifies described object.Particularly, object video refers to the object comprising in video, for example: car, bag, food, clothes etc.The image of object video refers to the image of the object video intercepting from video, for example: image, the image of clothes etc. of the image of car, the image of bag, food.The object video of described interactive operation, for example user has clicked a car in video, and this car is exactly the object video of described interactive operation.The present invention, in the time obtaining the image of object video of described interactive operation, can utilize existing image Segmentation Technology, and the image of the object video in described interactive operation region is split.Object type in described image comprises: " car " class, " bag " class, " food " class, " clothes " class, the present embodiment utilizes conventional images recognition technology can identify the object type in image.
Step S50, according to the matching degree of the recognition result of described object being calculated to each advertisement in described object video and advertisement base.Particularly, advertisement in advertisement base includes but not limited to display advertising, copy, video ads, web advertisement, no matter be the advertisement of what type, in advertisement or include image, include word, not only comprise image but also comprise word, so can utilize existing images match, the technology of characters matching, the image of described object video is mated with advertising image in advertisement base, the word of described object is mated with copy in advertisement base, then images match degree and characters matching degree weighted average are obtained to the matching degree of each advertisement in advertisement base.
Step S60, according to the matching degree of described user's recognition result being calculated to each advertisement in described user and advertisement base.Preferably, by each advertisement and described user's type matching degree, hobby matching degree weighted average, obtain the matching degree of each advertisement and described user in this advertisement base.
Step S70, calculates the preference of described each advertisement according to the priority of the matching degree of each advertisement in the matching degree of each advertisement in described object video and advertisement base, described user and advertisement base, each advertisement.Particularly, the mode of calculating can be the preference that priority that matching degree that the matching degree of described object video and each advertisement is multiplied by described user and each advertisement is multiplied by each advertisement obtains each advertisement in described advertisement base.The mode of calculating can be also the priority of the matching degree of described object video and each advertisement, described user and the matching degree of each advertisement, each advertisement obtains each advertisement in described advertisement base preference by weighted average.The mode of calculating can be also to adopt other function f, make the preference=f of an advertisement in described advertisement base, wherein, matching degree, this advertisement and the described user's of this advertisement and described object matching degree, the priority of this advertisement are three variablees of f, and in described advertisement base, the preference of an advertisement is by function f and the common decision of described three variablees.The principle of structure f be make the matching degree of an advertisement and described object video larger, this advertisement and described user's matching degree priority larger, this advertisement is larger, the preference of this advertisement is larger.
Step S80, gives described user by the advertising display of preference maximum in described advertisement base, and user is recorded into advertisement base checking of advertisement.Particularly, in advertisement base, the advertisement of preference maximum is and user and the highest advertisement of the mutual object video matching degree of user.The mode of display advertisement includes but not limited to pop-up advertisement, is linked to advertisement page, broadcast advertisement video, show advertising pictures.
As shown in Figure 2, in specific implementation process, described step S50 specifically comprises:
Step S510, carries out fuzzy matching by the picture of each advertisement in the image of described object video and advertisement base, and records the image of described object and the matching degree of each advertisement.Particularly, the picture of advertisement refers to the picture comprising in advertisement, is to be generally stored in computer system, for example in the mode of picture file: the picture of bag, the picture of car etc.The image of described object video is the object images splitting from video, for example: the image of bag, the image of car etc.Images match can obtain by prior art the similarity of each advertising pictures in the image of described object video and advertisement base, i.e. matching degree is similar to the diagram technology of searching of searching diagram technology, *** of ***.In advertisement base, there is advertisement a1, a2 .., an, the image of described object and the matching degree of this advertisement are p (ai, v), are p (ai, v)/p after normalization.Matching degree is the degree of coupling, is the number between 0 to 1, in the time being 0, does not mate completely, in the time being 1, mates completely.If do not comprise image in an advertisement, the image of described object and the matching degree of this advertisement are 0.If the image of image and described object video does not have similarity in an advertisement, the image of described object and the matching degree of this advertisement are also 0.Obtain, after the image of described object and the matching degree of each advertisement, being normalized, for example, each matching degree is respectively p (a1, v), p (a2, v) ... p (an, v), p=p (a1, v)+p (a2, v)+... + p (an, v),, after normalization, each corresponding matching degree is p (a1, v)/p, p (a2, v)/p,, p (an, v)/p.Normalized object is the situation of avoiding absolute matching degree too small, thereby replaces absolute matching degree by relative matching degree, because the object of the invention is to optimize more suitable comparatively speaking advertisement from advertisement base.
Step S520, carries out fuzzy matching by the word of each advertisement in the type of described object video and advertisement base, and records the type of described object and the matching degree of each advertisement.Particularly, the type of object, for example: bag type, car type.For example, in the time that the type of described object is bag type, in search advertisements storehouse, whether the word of each advertisement contains " bag ", and the total ratio of the number of times occurring in a copy according to " bag " and this copy is as the type of described object and the matching degree of this advertisement.Also can adopt the technology of existing characters matching to calculate the type of described object and the matching degree of each copy.In advertisement base, there is advertisement a1, a2 .., an, the type of described object and the matching degree of this advertisement are q (ai, v), are q (ai, v)/q after normalization.Matching degree is the degree of coupling, is the number between 0 to 1, in the time being 0, does not mate completely, in the time being 1, mates completely.If do not comprise word in an advertisement, the word of described object and the matching degree of this advertisement are 0.If a word corresponding to type that advertisement Chinese word does not comprise described object, the type of described object and the matching degree of this advertisement are also 0.Obtain, after the type of described object and the matching degree of each advertisement, the matching degree that is not 0 will being normalized, for example, each matching degree is respectively q (a1, v), q (a2, v) ..., q (an, v), q=q (a1, v)+q (a2, v)+... + q (an, v), q=q (a1, v)+q (a2, v)+... + q (an, v), after normalization, each corresponding matching degree is q (a1, v)/q, q (a2, v)/q,, q (an, v)/q.Normalized object is the situation of avoiding absolute matching degree too small, thereby replaces absolute matching degree by relative matching degree, because the object of the invention is to optimize more suitable comparatively speaking advertisement from advertisement base.
Step S530, is weighted the matching degree of the image of each advertisement and described object video, object type on average, obtains the matching degree of described object and each advertisement.Particularly, in advertisement base, there is advertisement a1, a2, .., an, the matching degree of described object v and advertisement ai is x1 (ai, v), the image of described object and the matching degree of this advertisement are p (ai, v), after normalization, be p (ai, v)/p, the type of described object and the matching degree of this advertisement are q (ai, v), after normalization, be q (ai, v)/q.Wherein i span is from 1 to n.X1 (ai, v)=k1 × p (ai, v)/p+k2 × q (ai, v)/q, for example x1 (a1, v)=k1 × p (a1, v)/p+k2 × q (a1, v)/q, x1 (a2, v)=k1 × p (a2, v)/p+k2 × q (a2, v)/q ..., x1 (an, v)=k1 × p (an, v)/p+k2 × q (an, v)/q, wherein k1≤0, k2≤0, k1+k2=1.In the time getting k1=k2=1/2, the images match degree of advertisement and described object and type matching degree are paid attention on an equal basis.In the time of k1>k2, more pay attention to the images match degree of advertisement and described object video.In the time of k2>k1, more pay attention to the object type matching degree of advertisement and described object video.
As shown in Figure 3, in specific implementation process, described step S60 comprises:
Step S610, carries out fuzzy matching by the prospective users type set of each advertisement in described user's type and described advertisement base, and records described user's type and the matching degree of each advertisement.Particularly, user's type includes but not limited to sex types (man, female), age type (old, in, few).For example, in the time that described user's type is maiden's type, in search advertisements storehouse, in the prospective users type of each advertisement, whether contain " maiden ", and the total ratio of the prospective users type of the number of times occurring according to " maiden " and this advertisement is as described user's type and the matching degree of this advertisement in the prospective users type of an advertisement.Also can adopt the technology of existing characters matching to calculate described user's type and the matching degree of each copy.Matching degree is the degree of coupling, is the number between 0 to 1, in the time being 0, does not mate completely, in the time being 1, mates completely.If do not comprise prospective users type in an advertisement, described user's type and the matching degree of this advertisement are 0.If prospective users type does not comprise described user's type in an advertisement, described user's type and the matching degree of this advertisement are also 0.Obtain, after described user's type and the matching degree of each advertisement, the matching degree that is not 0 will being normalized, for example, in advertisement base, there is advertisement a1, a2 .., an, the matching degree of the prospective users type of the type of described user u and advertisement ai is s (ai, u).Wherein i span is from 1 to n.The type of described user u and the matching degree of each advertisement are respectively s (a1, u), s (a2, u) ..., s (an.u), s=s (a1, u)+s (a2, u)+... + s (an, u),, after normalization, each corresponding matching degree is s (a1, u)/s, s (a2, u)/s,, s (an, u)/s.Normalized object is the situation of avoiding absolute matching degree too small, thereby replaces absolute matching degree by relative matching degree, because the object of the invention is to optimize more suitable comparatively speaking advertisement from advertisement base.
S620, obtains the type of each advertisement in advertisement base, and the preference degree of the type using described user to each advertisement is as the hobby matching degree of described user and described each advertisement of advertisement base.Particularly, in advertisement base, there is advertisement a1, a2 .., an, the hobby matching degree of described user u and advertisement ai is t (ai, u).Obtain the type of each advertisement in advertisement base, the for example type of advertisement ai is h (ai), using user u the preference degree g to the advertisement of h (ai) type (h (ai)) as user u the hobby matching degree t (ai to advertisement ai, u), be t (ai, u)=g (h (ai)).Wherein i span is from 1 to n.The hobby matching degree of described user u and advertisement ai is respectively t (a1, u), t (a2, u) ..., t (an.u), t=t (a1, u)+t (a2, u)+... + t (an, u),, after normalization, each corresponding matching degree is t (a1, u)/t, t (a2, u)/t,, t (an, u)/t.Normalized object is the situation of avoiding absolute matching degree too small, thereby replaces absolute matching degree by relative matching degree, because the object of the invention is to optimize more suitable comparatively speaking advertisement from advertisement base.
Step S630, is weighted the matching degree of the matching degree of each advertisement and described user's type, hobby on average, obtains the matching degree of each advertisement and described user in described advertisement base.Particularly, in advertisement base, there is advertisement a1, a2 .., an.Described user is u.The matching degree of advertisement ai and described user u is x2 (ai, u), and the matching degree of the type of described user u and advertisement ai is s (ai, u), after normalization, be s (ai, u)/s, the matching degree of the hobby of described user u and advertisement ai is t (ai, u), is t (ai after normalization, u)/t, x2 (ai, u)=k3 × s (ai, u)/t+k4 × t (ai, u)/t, wherein i span is from 1 to n.For example x2 (a1, u)=k3 × s (a1, u)/s+k4 × t (a1, u)/t, x2 (a2, u)=k3 × s (a2, u)/s+k4 × t (a2, u)/t ... x2 (an, u)=k3 × s (an, u)/s+k4 × t (an, u)/t, wherein k3≤0, k4≤0, k3+k4=1.In the time getting k3=k4=1/2, advertisement and described user's type matching degree and hobby matching degree are paid attention on an equal basis.In the time of k1>k2, more pay attention to advertisement and described user's type matching degree.In the time of k2>k1, more pay attention to advertisement and described user's hobby matching degree.
In specific implementation process, described step S70 comprises:
Step S710, is weighted the priority of the matching degree of the matching degree of each advertisement in described object video and described advertisement base, described user and each advertisement, each advertisement on average, obtains the preference of each advertisement in described advertisement base.Particularly, in advertisement base, there is advertisement a1, a2 .., an.In advertisement base, the priority of each advertisement can read from advertisement base, and the relative importance value of advertisement ai is x3 (ai).X3=x3 (a1)+x3 (a2)+... + x3 (an),, after normalization, each corresponding matching degree is x3 (a1)/x3, x3 (a2)/x3 ..., x3 (an)/x3.Described user is u, and described object is v.The matching degree of advertisement ai and described object v is x1 (ai, t), after normalization, be x1 (ai, v)/x1, the matching degree of advertisement ai and described user u is x2 (ai, u), after normalization, be x2 (ai, u)/x2, the relative importance value of advertisement ai is x3 (ai), is x3 (ai)/x3 after normalization.D1, d2, d3 is weights, and d1+d2+d3=1.The preference of advertisement ai is designated as y (ai)=d1 × x1 (ai, v)/x1+d2 × x2 (ai, u)/x2+d3 × x3 (ai)/x3.Wherein i span is from 1 to n.
The automatic generation method of the video interactive advertisement that another embodiment of the present invention provides also comprised before the step S30 of above-described embodiment:
Step S10, builds advertisement base, and each advertisement in described advertisement base comprises field: type, priority, word, picture, prospective users type, user check record, and wherein user checks that record comprises user's ID, the time length of checking.Particularly, the type of advertisement is for example wrapped series advertisements, car series advertisements etc.The priority of an advertisement is set in advance by advertisement operators, and for example the price of an advertisement is higher, and priority corresponding to this advertisement is just higher.The priority of advertisement generally represents with positive number.The value of the priority of advertisement is larger, and the priority of advertisement is higher.The word of advertisement, picture refer to the word, the picture that in advertisement, comprise.The prospective users type of an advertisement refers to the interested crowd of this advertisement, for example, to be woman to the interested prospective users type of female garment advertisement, is man to the interested prospective users type of in men's style clothes advertisement.The prospective users type of an advertisement can be set in advance in advertisement base.
Step S20, checks each user in described advertisement base the ratio of checking time span that the time span of advertisement of each type and this user are total, the preference degree of the advertisement as described each user to described each type.Particularly, in advertisement base, each advertisement has user to check that record field comprises user's ID, the time length of checking.By this user in all advertisements of advertisement base is sued for peace the time of checking, just can obtain this user total check time span.By the time span of this user in each type advertisement of advertisement base being sued for peace the time of checking, just can obtain this user and checking the advertisement of each type.Suppose the advertisement of total m type of advertisement base, the preference degree of the advertisement of this user to each type is designated as g (1), g (2) ..., g (m), wherein g (1)+g (2)+... + g (m)=1.
The automatic creation system of the corresponding video interactive advertisement providing of the embodiment of the present invention, as shown in Figure 4, it comprises interactive module 30, object type identification module 40, object video matching module 50, user's matching module 60, preference computing module 70 and display module 80.
Wherein, described interactive module 30, for accepting the interactive operation of user to video, is identified described user.Specifically refer to the embodiment of above-mentioned steps S30.
Described object type identification module 40, for obtaining the image of object video of described interactive operation, is identified described object.Specifically refer to the embodiment of above-mentioned steps S40.
Described object video matching module 50 calculates the matching degree of described object video and the each advertisement of advertisement base for basis to the recognition result of described object.Specifically refer to the embodiment of above-mentioned steps S50.
Described user's matching module 60 is for calculating the matching degree of described object video and the each advertisement of advertisement base according to described user's type and hobby.Specifically refer to the embodiment of above-mentioned steps S60.
Described preference computing module 70 is for calculating the preference of each advertisement in described advertisement base according to the priority of the matching degree of each advertisement in the matching degree of described object video and the each advertisement of advertisement base, described user and advertisement base, each advertisement.Specifically refer to the embodiment of above-mentioned steps S70.
Described display module 80, for giving described user by the advertising display of described advertisement base preference maximum, and records user into advertisement base to checking of advertisement.Specifically refer to the embodiment of above-mentioned steps S80.
As shown in Figure 5, in specific implementation process, described object video matching module 50 comprises object images matching unit 510, object type matching unit 520 and object matching degree computing unit 530.
Wherein, described object images matching unit 510 is for the picture of the image of described object video and the each advertisement of advertisement base is carried out to fuzzy matching, and records the image of described object video and the matching degree of each advertisement.Specifically refer to the embodiment of above-mentioned steps S510.
Described object type matching unit 520 is for the word of the type of described object video and the each advertisement of advertisement base is carried out to fuzzy matching, and records the type of described object video and the matching degree of each advertisement.Specifically refer to the embodiment of above-mentioned steps S520.
Described object matching degree computing unit 530, for the matching degree of the image of each advertisement and described object video, object type is weighted on average, obtains the matching degree of described object and each advertisement.Specifically refer to the embodiment of above-mentioned steps S530.
As shown in Figure 6, in specific implementation process, described user's matching module 60 comprises user type matching unit 610, user preferences matching unit 620, user's matching degree computing unit 630.
Wherein, described user type matching unit 610 is for the prospective users type set of described user's type and the each advertisement of described advertisement base is carried out to fuzzy matching, and records described user's type and the matching degree of each advertisement.Specifically refer to the embodiment of above-mentioned steps S610.
Described user preferences matching unit 620, for obtaining the type of each advertisement of advertisement base, the preference degree of the type using described user to each advertisement is as the matching degree of the hobby of described user and each advertisement.Specifically refer to the embodiment of above-mentioned steps S620.
Described user's matching degree computing unit 630 is for being weighted the matching degree of the matching degree of each advertisement and described user's type, hobby on average, obtains the matching degree of each advertisement and described user in described advertisement base.Specifically refer to the embodiment of above-mentioned steps S630.
In specific implementation process, described preference computing module 70 comprises preference weighted average unit 710.
Wherein, described preference weighted average unit 710, for the priority of the matching degree of each advertisement of described advertisement base and described object, this advertisement and described user's matching degree, this advertisement is weighted on average, obtains the preference of each advertisement in described advertisement base.Specifically refer to the embodiment of above-mentioned steps S710.
The automatic creation system of the corresponding video interactive advertisement providing of another embodiment of the present invention, also comprises that in the system-based of above-described embodiment advertisement base builds module 10, user preferences identification module 20.
Advertisement base builds module 10, be used for building advertisement base, each advertisement in described advertisement base comprises field: type, priority, word, picture, prospective users type, user check record, and wherein user checks that record comprises user's ID, the time length of checking.Specifically refer to the embodiment of above-mentioned steps S10.
User preferences identification module 20, for the ratio of checking time span of each user of described advertisement base being checked to the time span of advertisement of each type and this user are total, the preference degree of the advertisement as described each user to described each type.Specifically refer to the embodiment of above-mentioned steps S20.
In sum, the automatic generation method of video interactive advertisement of the present invention and system, Properties of Objects in video can be taken into full account, user's hobby can be taken into full account again, all by the object of user interactions in video, there is a most preferred advertisement can show this user, the not mutual object video of user can pop-up advertisement, makes the validity of video interactive advertisement very high; Simultaneously because add advertisement without manual, so speed is fast, cost is low.
The foregoing is only preferred embodiment of the present invention, not in order to limit the present invention, all any amendments of doing within the spirit and principles in the present invention, be equal to and replace and improvement etc., within all should being included in protection scope of the present invention.
Claims (10)
1. a generation method for video ads, is characterized in that, said method comprising the steps of:
Accept the interactive operation of user to video, identify described user;
Obtain the image of the object video of described interactive operation, identify described object;
According to the matching degree of the recognition result of described object being calculated to each advertisement in described object video and advertisement base;
According to the matching degree of described user's recognition result being calculated to each advertisement in described user and advertisement base;
Calculate the preference of described each advertisement according to the priority of the matching degree of each advertisement in the matching degree of each advertisement in described object video and advertisement base, described user and advertisement base, each advertisement;
Give described user by the advertising display of preference maximum in described advertisement base, and user is recorded into advertisement base checking of advertisement.
2. the generation method of video ads according to claim 1, is characterized in that, accepts the interactive operation of user to video described, and before identifying described user's step, described method is further comprising the steps of:
Build advertisement base, each advertisement in described advertisement base comprises field: type, priority, word, picture, prospective users type, user check record, and wherein user checks that record comprises user's ID, the time length of checking;
Each user in described advertisement base is checked to the ratio of checking time span that the time span of advertisement of each type and this user are total, the preference degree of the advertisement as described each user to described each type.
3. the generation method of video ads according to claim 1, is characterized in that, described basis is calculated the step of the matching degree of each advertisement in described object video and advertisement base to the recognition result of described object, specifically comprise:
The picture of each advertisement in the image of described object video and advertisement base is carried out to fuzzy matching, and record the image of described object video and the matching degree of each advertisement;
The word of each advertisement in the type of described object video and advertisement base is carried out to fuzzy matching, and record the type of described object video and the matching degree of each advertisement;
The matching degree of the image of each advertisement and described object video, object type is weighted on average, obtains the matching degree of described object video and each advertisement.
4. the generation method of video ads according to claim 1, is characterized in that, described basis is calculated the step of the matching degree of each advertisement in described user and advertisement base to described user's recognition result, specifically comprise:
The prospective users type set of each advertisement in described user's type and described advertisement base is carried out to fuzzy matching, and record described user's type and the matching degree of each advertisement;
Obtain the type of each advertisement in advertisement base, the preference degree of the type using described user to each advertisement is as the matching degree of the hobby of described user and each adline;
The matching degree of the matching degree of each advertisement and described user's type, each adline hobby is weighted on average, obtains the matching degree of each advertisement and described user in described advertisement base.
5. the generation method of video ads according to claim 1, it is characterized in that, the described priority according to the matching degree of each advertisement in the matching degree of each advertisement in described object video and advertisement base, described user and advertisement base, each advertisement calculates the step of the preference of described each advertisement, specifically comprises:
By each advertisement in described object video and described advertisement base and matching degree, described user and the matching degree of each advertisement, the priority of each advertisement be weighted on average, obtain the preference of each advertisement in described advertisement base.
6. a generation system for video ads, is characterized in that, comprising:
Interactive module, for accepting the interactive operation of user to video, identifies described user;
Object type identification module, for obtaining the image of object video of described interactive operation, identifies described object;
Object video matching module, calculates the matching degree of described object video and the each advertisement of advertisement base for basis to the recognition result of described object;
User's matching module, calculates the matching degree of described user and the each advertisement of advertisement base for basis to described user's recognition result;
Preference computing module, for calculating the preference of each advertisement in described advertisement base according to the priority of the matching degree of each advertisement in the matching degree of described object video and the each advertisement of advertisement base, described user and advertisement base, each advertisement;
Display module, for giving described user by the advertising display of described advertisement base preference maximum, and records user into advertisement base to checking of advertisement.
7. the generation system of video ads according to claim 6, is characterized in that, described system also comprises:
Advertisement base builds module, be used for building advertisement base, each advertisement in described advertisement base comprises field: type, priority, word, picture, prospective users type, user check record, and wherein user checks that record comprises user's ID, the time length of checking;
User preferences identification module, for the ratio of checking time span of user being checked to the time span of advertisement of each type and this user are total, the preference degree of the advertisement as this user to each type.
8. the generation system of video ads according to claim 6, is characterized in that, described object video matching module comprises:
Object images matching unit, for the picture of the image of described object video and the each advertisement of advertisement base is carried out to fuzzy matching, and records the image of described object video and the matching degree of each advertisement;
Object type matching unit, for the word of the type of described object video and the each advertisement of advertisement base is carried out to fuzzy matching, and records the type of described object video and the matching degree of each advertisement;
Object matching degree computing unit, for the matching degree of the image of each advertisement and described object, type is weighted on average, obtains the matching degree of described object video and each advertisement.
9. the generation system of video ads according to claim 6, is characterized in that, described user's matching module comprises:
User type matching unit, for the prospective users type set of described user's type and the each advertisement of described advertisement base is carried out to fuzzy matching, and records described user's type and the matching degree of each advertisement;
User preferences matching unit, for obtaining the type of each advertisement of advertisement base, the preference degree of the type using described user to each advertisement is as the matching degree of the hobby of described user and each advertisement;
User's matching degree computing unit, for the matching degree of the matching degree of each advertisement and described user's type, hobby is weighted on average, obtains the matching degree of each advertisement and described user in described advertisement base.
10. the generation system of video ads according to claim 6, it is characterized in that, described preference computing module, specifically for the priority of the matching degree of each advertisement and described object in described advertisement base, this advertisement and described user's matching degree, this advertisement is weighted on average, obtains the preference of each advertisement in described advertisement base.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410167442.5A CN104185041B (en) | 2014-04-24 | 2014-04-24 | The automatic generation method and system of video interactive advertisement |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410167442.5A CN104185041B (en) | 2014-04-24 | 2014-04-24 | The automatic generation method and system of video interactive advertisement |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104185041A true CN104185041A (en) | 2014-12-03 |
| CN104185041B CN104185041B (en) | 2018-05-11 |
Family
ID=51965733
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410167442.5A Active CN104185041B (en) | 2014-04-24 | 2014-04-24 | The automatic generation method and system of video interactive advertisement |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104185041B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105992029A (en) * | 2015-02-12 | 2016-10-05 | 广东欧珀移动通信有限公司 | Wallpaper recommendation method and system, server, and mobile terminal |
| CN106331742A (en) * | 2016-08-25 | 2017-01-11 | 武克易 | Video advertisement playing system |
| CN108052671A (en) * | 2017-12-29 | 2018-05-18 | 北京奇虎科技有限公司 | A kind of method and system that content item push is carried out based on Dynamic Matching |
| CN108124167A (en) * | 2016-11-30 | 2018-06-05 | 阿里巴巴集团控股有限公司 | A kind of play handling method, device and equipment |
| CN108769808A (en) * | 2018-05-24 | 2018-11-06 | 安徽质在智能科技有限公司 | Interactive video playback method and system |
| CN109102335A (en) * | 2018-08-09 | 2018-12-28 | 佛山市苔藓云链科技有限公司 | A method of executing media interaction in the server |
| CN110582015A (en) * | 2019-07-26 | 2019-12-17 | 马力 | extension method, system and terminal for controlling video picture information |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101141607A (en) * | 2006-09-08 | 2008-03-12 | 百视通网络电视技术发展有限责任公司 | Mutual correlation method for IPTV and implementing system thereof |
| CN102118401A (en) * | 2009-12-31 | 2011-07-06 | 华为技术有限公司 | Media playing method and system, content transmission network equipment and advertisement server (AS) |
| CN102254265A (en) * | 2010-05-18 | 2011-11-23 | 北京首家通信技术有限公司 | Rich media internet advertisement content matching and effect evaluation method |
| US20120230654A1 (en) * | 2007-09-30 | 2012-09-13 | International Businesss Machines Corporation | Labeling a video, for modifying a video, and for video processing |
| CN102843589A (en) * | 2012-09-29 | 2012-12-26 | 乐视网信息技术（北京）股份有限公司 | Intelligent advertisement insertion system and intelligent advertisement insertion method |
| CN103368898A (en) * | 2012-03-26 | 2013-10-23 | 中兴通讯股份有限公司 | Method and system for accomplishing information push |
| CN103384311A (en) * | 2013-07-18 | 2013-11-06 | 博大龙 | Method for generating interactive videos in batch mode automatically |
-
2014
- 2014-04-24 CN CN201410167442.5A patent/CN104185041B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101141607A (en) * | 2006-09-08 | 2008-03-12 | 百视通网络电视技术发展有限责任公司 | Mutual correlation method for IPTV and implementing system thereof |
| US20120230654A1 (en) * | 2007-09-30 | 2012-09-13 | International Businesss Machines Corporation | Labeling a video, for modifying a video, and for video processing |
| CN102118401A (en) * | 2009-12-31 | 2011-07-06 | 华为技术有限公司 | Media playing method and system, content transmission network equipment and advertisement server (AS) |
| CN102254265A (en) * | 2010-05-18 | 2011-11-23 | 北京首家通信技术有限公司 | Rich media internet advertisement content matching and effect evaluation method |
| CN103368898A (en) * | 2012-03-26 | 2013-10-23 | 中兴通讯股份有限公司 | Method and system for accomplishing information push |
| CN102843589A (en) * | 2012-09-29 | 2012-12-26 | 乐视网信息技术（北京）股份有限公司 | Intelligent advertisement insertion system and intelligent advertisement insertion method |
| CN103384311A (en) * | 2013-07-18 | 2013-11-06 | 博大龙 | Method for generating interactive videos in batch mode automatically |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105992029A (en) * | 2015-02-12 | 2016-10-05 | 广东欧珀移动通信有限公司 | Wallpaper recommendation method and system, server, and mobile terminal |
| CN106331742A (en) * | 2016-08-25 | 2017-01-11 | 武克易 | Video advertisement playing system |
| CN108124167A (en) * | 2016-11-30 | 2018-06-05 | 阿里巴巴集团控股有限公司 | A kind of play handling method, device and equipment |
| CN108052671A (en) * | 2017-12-29 | 2018-05-18 | 北京奇虎科技有限公司 | A kind of method and system that content item push is carried out based on Dynamic Matching |
| CN108052671B (en) * | 2017-12-29 | 2021-01-08 | 北京奇元科技有限公司 | Method and system for pushing content items based on dynamic matching |
| CN108769808A (en) * | 2018-05-24 | 2018-11-06 | 安徽质在智能科技有限公司 | Interactive video playback method and system |
| CN108769808B (en) * | 2018-05-24 | 2021-01-29 | 浙报融媒体科技(浙江)有限责任公司 | Interactive video playing method and system |
| CN109102335A (en) * | 2018-08-09 | 2018-12-28 | 佛山市苔藓云链科技有限公司 | A method of executing media interaction in the server |
| CN110582015A (en) * | 2019-07-26 | 2019-12-17 | 马力 | extension method, system and terminal for controlling video picture information |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104185041B (en) | 2018-05-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20200175550A1 (en) | Method for identifying advertisements for placement in multimedia content elements | |
| CN104185041A (en) | Video interaction advertisement automatic generation method and system | |
| US11620678B2 (en) | Advertising method, device and system, and computer-readable storage medium | |
| CN105745643B (en) | System and method for creating image-based content based on text-based content | |
| US9043860B2 (en) | Method and apparatus for extracting advertisement keywords in association with situations of video scenes | |
| WO2015149307A1 (en) | Systems and methods for optimizing content layout using behavior metrics | |
| US8346604B2 (en) | Facilitating bidding on images | |
| US20170213248A1 (en) | Placing sponsored-content associated with an image | |
| US8688514B1 (en) | Ad selection using image data | |
| US10798425B1 (en) | Personalized key object identification in a live video stream | |
| US11609943B2 (en) | Contextual content distribution | |
| EP2874102A2 (en) | Generating models for identifying thumbnail images | |
| US10380623B2 (en) | System and method for generating an advertisement effectiveness performance score | |
| CN112818224B (en) | Information recommendation method and device, electronic equipment and readable storage medium | |
| US11334924B2 (en) | Automated image ads | |
| CN112989213A (en) | Content recommendation method, device and system, electronic equipment and storage medium | |
| US11762900B2 (en) | Customized selection of video thumbnails to present on social media webpages | |
| US20160171548A1 (en) | Method for identifying advertisements for placement in multimedia content elements | |
| US20120095825A1 (en) | Incentive Selection of Region-of-Interest and Advertisements for Image Advertising | |
| CN114881685A (en) | Advertisement delivery method, device, electronic device and storage medium | |
| US11367103B2 (en) | Method for modeling digital advertisement consumption | |
| US20150227970A1 (en) | System and method for providing movie file embedded with advertisement movie | |
| US20150181288A1 (en) | Video sales and marketing system | |
| US20220164090A1 (en) | Abstract generation method and apparatus | |
| CN113901244A (en) | Label construction method and device for multimedia resource, electronic equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| TA01 | Transfer of patent application right | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20180321 Address after: 523000, Guangdong province Dongguan Songshan Lake hi tech Industrial Development Zone Creative Life City shopping mall B two floor shopping mall 2 part of the site (No. 201) Applicant after: Great power innovative Intelligent Technology (Dongguan) Co., Ltd. Address before: 510630 Guangzhou University City, Guangdong Applicant before: Zhu Dingju |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |