CN106504283A - Information broadcasting method, apparatus and system - Google Patents
Information broadcasting method, apparatus and system Download PDFInfo
- Publication number
- CN106504283A CN106504283A CN201610850683.9A CN201610850683A CN106504283A CN 106504283 A CN106504283 A CN 106504283A CN 201610850683 A CN201610850683 A CN 201610850683A CN 106504283 A CN106504283 A CN 106504283A
- Authority
- CN
- China
- Prior art keywords
- object space
- human body
- information
- body information
- depth image
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/60—Analysis of geometric attributes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30196—Human being; Person
Abstract
The invention discloses information broadcasting method, apparatus and system.Wherein, the method includes：Obtain the depth image of object space；The depth image is analyzed, the human body information in the object space is obtained, wherein, the human body information includes the height of the people in object space, body weight, at least one of sex；Play the information that mates with the human body information.By the way, accurate acquisition human body information can be realized, and adaptive information broadcasting is carried out according to the human body information of collection.
Description
Technical field
The present invention relates to technical field of information processing, more particularly to information broadcasting method, apparatus and system.
Background technology
People are often that information is informed beholder by use information broadcast mode.Common mode is exactly in TV, computer
Or on advertising display platform, play advertisement.However, current information broadcast mode is remained by playback terminal with certain input
Mode solely plays out identical information, and None- identified difference beholder and carry out different information and play.
In addition, at present conventional human body information acquisition mode is view-based access control model image carry out height, surname to human body not and
The acquisition of the information such as body weight.Specifically, mainly use that coloured image does not carry out height, surname not to human body and body weight is estimated roughly
And differentiate.As coloured image is only capable of reflecting the two-dimensional signals such as color and the texture of personage, the impact such as illumination is also suffered from, separately
Outer cannot then carry out height measurement to moving human body, thus the accuracy for obtaining human body information based on Color Image Acquisition is not
High.
Content of the invention
The invention mainly solves the technical problem of providing information broadcasting method, apparatus and system, can realize accurately adopting
Collection human body information, and adaptive information broadcasting is carried out according to the human body information of collection.
For solving above-mentioned technical problem, one aspect of the present invention is：A kind of information broadcasting method is provided, is wrapped
Include：
Obtain the depth image of object space；
The depth image is analyzed, the human body information in the object space is obtained, wherein, the human body information
The height of the people in including object space, body weight, at least one of sex；
Play the information that mates with the human body information.
Wherein, described the depth image is analyzed, obtain human body information in the object space, including：
The view data related to human body information is extracted from the depth image；The mesh is obtained using the view data of the extraction
The human body information in mark space.
Wherein, when the human body information includes height, the extraction from the depth image is related to human body information
View data, including：
The view data related to the height, the related view data bag of the height is extracted from the depth image
Include corresponding first three dimensional space coordinate of head peak pixel in the object space, left thigh peak pixel corresponding
Corresponding 3rd three dimensional space coordinate of second three dimensional space coordinate, right thigh peak pixel, left foot minimum point pixel are corresponding
4th three dimensional space coordinate and corresponding 5th three dimensional space coordinate of right crus of diaphragm minimum point pixel；
The view data using the extraction obtains the human body information in the object space, including：According to described
The three dimensional space coordinate of extraction, is calculated and is referred to by the midpoint of line between the left thigh peak and the right thigh peak
Primary vector to the head peak, the secondary vector for being pointed to the left foot minimum point by the left thigh peak and
The 3rd vector that the right crus of diaphragm minimum point is pointed to by the right thigh peak；Using the primary vector, secondary vector and
Three-dimensional amount is calculated the height in the object space.
Wherein, described it is calculated in the object space using the primary vector, secondary vector and three-dimensional amount
The height of people, including：By the primary vectorSecondary vectorWith the 3rd vectorFollowing formula 1 or formula 2 is substituted into, is obtained
Height Height of the people in the object space,
Wherein, when the human body information includes sex, the extraction from the depth image is related to human body information
View data, including：
The view data related to sex, wherein, the picture number related to sex is extracted from the depth image
According to the contour feature of the face included in the object space, the contour feature of shoulder, the contour feature of chest and buttocks
Contour feature；
The view data using the extraction obtains the human body information in the object space, including：Carry described
The height of the people in the view data related to sex and the object space that take is input into setting grader is classified, and
According to the sex that the classification results for setting grader determine the people in the object space.
Wherein, when the human body information includes body weight, the extraction from the depth image is related to human body information
View data, including：
The view data related to body weight, wherein, the picture number related to body weight is extracted from the depth image
Profile according to the contour feature of face, the contour feature of shoulder, the contour feature of chest, buttocks included in the object space
Feature and the contour feature of abdominal part；
The view data using the extraction obtains the human body information in the object space, including：Carry described
The height and sex of the people in the view data related to body weight for taking and the object space is input into setting regression model pair
Body weight is estimated, and determines the body weight of the people in the object space according to estimation result.
Wherein, the information that the broadcasting is mated with the human body information, including：The human body of the people of the object space is believed
Breath is sorted out, and obtains the crowd's type belonging to the people in the object space；The advertising message with crowd's type matching is searched,
And the advertising message for finding is played out.
For solving above-mentioned technical problem, another technical solution used in the present invention is：A kind of information play apparatus are provided,
Including：
Acquisition module, for obtaining the depth image of object space；
Module is obtained, for being analyzed to the depth image, the human body information in the object space is obtained, its
In, the human body information includes the height of the people in object space, body weight, at least one of sex；
Playing module, for playing the information that mates with the human body information.
For solving above-mentioned technical problem, another technical scheme that the present invention is adopted is：A kind of information broadcasting system is provided,
Including the image capture device that is connected with each other and playback equipment, the playback equipment includes processor, memorizer and player；
Described image collecting device is used for the depth image for gathering object space, and exports to the playback equipment；
The memorizer of the playback equipment is used for storing computer instruction and the processor needed for processing procedure
The view data for using；
Computer instruction described in the computing device, is used for：
Obtain the depth image of object space；
The depth image is analyzed, the human body information in the object space is obtained, wherein, the human body information
The height of the people in including object space, body weight, at least one of sex；
To the information that player output is mated with the human body information；
The player is used for playing the information that mates with human body information.
Wherein, described image collecting device is arranged in the playback equipment.
The invention has the beneficial effects as follows：It is analyzed by the depth image from the object space comprising human body and obtains the mesh
The human body information in mark space, and then play the information that mates with the human body information of the object space, it is achieved that according to human body information
Adaptive identifying corresponding informance, is additionally, since depth image comprising three-dimensional information, therefore using the view data in the depth image
The human body information for obtaining is more accurate, and as depth image is for characterizing depth information, and do not affected by ambient light, therefore above-mentioned
The collection of human body information is not limited by ambient light, i.e., no matter which kind of environment does not interfere with the collection of the human body information, improves this
The stability of human body information collection, therefore also improve the accurately fixed and reliability of information self-adapting broadcasting.
Description of the drawings
Fig. 1 is the flow chart of one embodiment of information broadcasting method of the present invention；
Fig. 2 is the structural representation of information broadcasting system in one application scenarios of information broadcasting method of the present invention；
Fig. 3 is the sub-step flow chart that step S12 includes in another embodiment of information broadcasting method of the present invention；
Fig. 4 is the flow chart of information broadcasting method another embodiment of the present invention；
Fig. 5 is the schematic diagram of object space depth image in one application scenarios of information broadcasting method of the present invention.
Fig. 6 is the schematic diagram of face in object space depth image in one application scenarios of information broadcasting method of the present invention；
Fig. 7 is the structural representation of one embodiment of information play apparatus of the present invention；
Fig. 8 is the structural representation of one embodiment of information broadcasting system of the present invention.
Specific embodiment
In order to be better understood from technical scheme, below in conjunction with the accompanying drawings the embodiment of the present invention is retouched in detail
State.
The term for using in embodiments of the present invention is the purpose only merely for description specific embodiment, and is not intended to be limiting
The present invention." one kind ", " described " and " being somebody's turn to do " of singulative used in the embodiment of the present invention and appended claims
It is also intended to include most forms, unless context clearly shows that other implications.It is also understood that term used herein
"and/or" is referred to and comprising one or more associated any or all possible combination for listing project.
Fig. 1 is referred to, Fig. 1 is the flow chart of one embodiment of information broadcasting method of the present invention.In the present embodiment, the method can
Executed by playing device, comprised the following steps：
S11：Obtain the depth image of object space.
Wherein, the depth image of the object space can be obtained using following three kinds of modes：
1) mode based on binocular vision
Specifically, human eye distance two cameras (equivalent to right and left eyes) of setting can be imitated synchronously to be acquired object space,
Left and right two width target image is obtained, the left and right target image is calculated by algorithm, to obtain the depth of the object space
Degree image.Wherein, above-mentioned can be executed by the playing device the step of calculate to left and right target image, or set by other
The standby processing meanss being connected with two cameras are executed.Two cameras can be color camera or be black light
Camera.Wherein, the black light camera can be for example, infrared for the black light camera of the following two kinds mode, black light camera
Camera or ultraviolet-cameras etc..
2) mode based on structure light
Specifically, can be as shown in Fig. 2 encoded by projecting to object space 23 using black light projection module 22
Structured light patterns (such as the speckle pattern of irregular alignment), and adopted using 21 pairs of object spaces 23 of black light camera
Collection, obtains the black light image of the object space with structured light patterns, and black light camera 21 can be passed through according to triangulation
The depth image of the object space is obtained after calculating to the black light image, and is exported to playing device 24.Certainly, exist
In other embodiment, black light camera 21 directly can export the black light image to playing device 24, be filled by the broadcasting
Put 24 depth images that the object space is obtained according to triangulation after calculating to the black light image.
Wherein, above-mentioned black light projection module 22 is typically made up of light source and diffraction optical element, and light source can be side
The black light of transmitting laser can also be vertical cavity surface-emitting laser, and the light source sends and can be recognized by the black light camera 21
Black light.For example, the black light be infrared light, the black light camera 21 be infrared camera, the black light image
For infrared image；Or the black light is ultraviolet light, the black light camera 21 is ultraviolet-cameras, and the black light image is
Ultraviolet image.Diffraction optical element needs to be configured of collimation, beam splitting, diffusion etc. according to different structured light patterns
Function.
Said structure light pattern can be the irregular speckle pattern of distribution.The dense degree of the speckle pattern have impact on depth
Speed and precision that angle value is calculated, speckle particle are more, and calculating speed is slower, but precision is higher.Therefore, the black light is thrown
Shadow module 22 can select suitable speckle particle density according to the approximate depth of the target area of shooting image, ensure to calculate
While speed, still there is higher computational accuracy.Certainly, the speckle particle density also can be by above-mentioned playing device 24 according to certainly
The calculating demand of body and determine, and by the density information of the determination send to black light project module 22.
In the present embodiment, the black light projection module 22 to object space be but do not limit to be projected with certain angle of flare
Speckle particle pattern.
It should be noted that above-mentioned black light projection module 22 is may also be arranged in the black light camera.
3) mode based on jet lag (Time of Flight, TOF)
Specifically, by projecting black light, and the black light returned by black light collected by camera, root to object space
The time that flies according to black light calculating the depth information of object space, to form the depth image of the object space.Wherein,
The black light as being infrared light, the black light of the projection can be sent by the black light camera or by other independently of
Other projective structures of black light camera send.
It is understood that in other embodiments, other modes can be also adopted to obtain the depth image of object space, therefore
Here is not limited.In addition, the S11 steps are alternatively the depth image for directly receiving the object space that other devices send, should
The depth image of object space is to obtain through aforesaid way in advance.
S12：The depth image is analyzed, the human body information in the object space is obtained.
Wherein, the human body information includes the height of the people in object space, body weight, at least one of sex.
Alternatively, after aforesaid way being obtained depth image, human bioequivalence can be carried out to the depth image first, if in depth
Human body is recognized in degree image, is then carried out the S12, otherwise, is not executed S12.Further, when recognizing people in depth image
During body, the human body parts segmentation in depth image can be obtained new depth image first, then the depth picture obtained from segmentation enters
Capable step S12.
Specifically, playing device is calculated the human body of the people in the object space using the information carried in depth image
Information, to realize that human body information is gathered.As shown in figure 3, S12 may include following sub-step：
S121:The view data related to human body information is extracted from the depth image.
In the present embodiment, playing device can according to the human body information of required collection predefine needed for extraction with the human body
The related view data of information, with after the depth image for obtaining the object space, according to the figure of predetermined required extraction
As data carry out image data extraction from the depth image.
For example, the three-dimensional spatial information for setting human body respective pixel in the depth image can extract as height correlation
View data；The contour feature at some positions of human body in the depth image can at least be extracted as related to body weight, sex
View data.The three-dimensional spatial information includes the depth information of the respective pixel of the depth image.The contour feature may include
This pair should human body some adjacent margins multiple pixel coordinates or the further result of calculation to the plurality of pixel coordinate.
S122：The human body information that the object space is obtained using the view data of the extraction.
For example, the target is calculated using the three-dimensional spatial information for setting human body respective pixel in the depth image
The height of interval people, using the people's in the contour feature and/or the object space at some positions of human body in the depth image
Height carries out the sex that Classification and Identification obtains the people of the target interval, and the profile using some positions of human body in the depth image is special
Levy and/or the target interval people height and sex be analyzed the body weight for calculating the people for obtaining the target interval.
S13：Play the information that mates with the human body information.
The playback terminal can targetedly broadcast information according to the human body information of object space.For example, playback terminal is preset
Have the grader to human body information, determine the affiliated type of people in the current object space according to the grader, and from local or
Person is set in server and finds the information that is mated with the affiliated species, and is played out.Wherein, the broadcast mode may include figure
At least one as showing, in sound broadcasting.
It is understood that above- mentioned information player method can be executed, regularly execute or receive user for real-time continuous
Just execute during instruction, for example, the playing device can constantly gather the depth image in current goal region in real time, and by the present invention
Method is obtained in the human body information for not being located at the people of the target area in the same time, and then updates the corresponding information of broadcasting.
In the present embodiment, it is analyzed by the depth image from the object space comprising human body and obtains the object space
Human body information, and then play the information that mates with the human body information of the object space, it is achieved that broadcast according to human body information self adaptation
Corresponding informance is put, and depth image is additionally, since comprising three-dimensional information, therefore the people obtained using the view data in the depth image
Body information is more accurate, and as depth image is for characterizing depth information, and do not affected by ambient light, therefore above-mentioned human body information
Collection do not limited by ambient light, i.e., no matter which kind of environment does not interfere with the collection of the human body information, improves the human body information
The stability of collection, therefore also improve the accurately fixed and reliability of information self-adapting broadcasting.
Fig. 4 is referred to, Fig. 4 is the flow chart of another embodiment of information broadcasting method of the present invention.In the present embodiment, the method
Can be executed by playing device, be comprised the following steps：
S41：Obtain the depth image of object space.
The description of above-mentioned S11 specifically see.
S42：The view data related to height, sex and body weight is extracted from the depth image.
In the present embodiment, being somebody's turn to do the view data related to height may include the head peak of the people in the object space
The corresponding first three dimensional space coordinate P1 of pixel, the corresponding second three dimensional space coordinate P2 of left thigh peak pixel, right thigh
The corresponding 3rd three dimensional space coordinate P3 of peak pixel, the corresponding 4th three dimensional space coordinate P4 of left foot minimum point pixel and the right side
The corresponding 5th three dimensional space coordinate P5 of foot minimum point pixel, as shown in Figure 5.The three dimensional space coordinate of above-mentioned pixel is by the picture
Depth of the human body in the corresponding object space of vegetarian refreshments in the two-dimensional coordinate and the pixel of the object space place plane
Degree information structure.In a concrete application, the human body in the corresponding object space of the pixel is located in the object space
The two-dimensional coordinate of plane can be world coordinates of the pixel in the camera for gathering the depth image, and the world coordinates is appreciated that
It is that the two dimension formed in the plane parallel with the camera lens that the object space is located is sat with the optical center of the camera as origin
Coordinate in mark system.
Being somebody's turn to do the view data related to sex may include：The contour feature of the face in the object space, the wheel of shoulder
The contour feature of wide feature, the contour feature of chest and buttocks.
Being somebody's turn to do the view data related to body weight includes the profile spy of the contour feature of the face in the object space, shoulder
Levy, the contour feature of the contour feature of the contour feature of chest, buttocks and abdominal part.
Wherein, the contour feature of certain human body A in above-mentioned object space is the people in the object space
The one of the human body A of the people in the contour line of body region A, the object space is parallel to object space place plane
The human body A's in ratio, the object space between the length of the contour line in section and width is multiple parallel to the mesh
People in the meansigma methodss of the ratio between the length of the contour line in the section of mark space place plane and width or the object space
In ratio between the length of the contour line in multiple sections parallel to object space place plane of body region A and width
Maximum.The lines that the corresponding pixel in the contour line of position A A edges in position i.e. in the depth image is formed, the target
I.e. parallel with the camera lens plane of space place plane, the length of the contour line in the section are respectively on the section with width
Greatest length under orthogonal two reference directions.As shown in fig. 6, the contour line of the face in the object space is L, should
The length of the contour line at the interface parallel to object space place plane that the nose of face is located is a1, and width is b1, therefore its
Ratio is a1/b1, further, can also obtain the wheel of the left eye pearl place parallel to the section of object space place plane of the face
Length a2 of profile and width b2, obtain its ratio for a2/b2, and obtain the right eye pearl place of the face parallel to target empty
Between place plane section contour line length a3 and width b3, obtain its ratio for a3/b3, take above three section and obtain
Ratio a1/b1, a2/b2, the meansigma methodss of a3/b3 or maximum as its head contour feature.
Further, different according to the feature of different parts, different parts can be directed to and select above different images data to make
For the contour feature at the position, for example, for face, then the multiple sections parallel to object space place plane of face are selected
Contour feature of the meansigma methodss of the ratio between length and width as the face in the object space；For shoulder, chest, abdominal part
And buttocks, then select the position multiple parallel to the section of object space place plane length and width between ratio in
Contour feature of the maximum as the position.
In the present embodiment, the said extracted view data related to sex/body weight is also included in the object space
Angle α between the optical axis of the normal in the front of people and the camera for obtaining the depth image, as shown in Figure 5.Can specifically pass through
Extract the corresponding second three dimensional space coordinate P2 (x of left thigh peak pixel of the people in object space2,y2,z2), right thigh
The corresponding 3rd three dimensional space coordinate P3 (x of peak pixel3,y3,z3), and be calculated using following formula 13,
Certainly, in other embodiments, the view data related to sex/body weight may not include angle α, and here is not limited
Fixed.
The setting of angle can avoid due to human body leaning to one side what caused characteristic information change brought in motor process
Sex, body weight estimation error, such that it is able to realize real-time monitoring.
S43：Height, sex and the body weight that the people in the object space is obtained using the view data of the extraction.
Wherein, the people in the object space is calculated using the view data related to height that extracts in above-mentioned S42
Height.Specifically, the step of this obtains height may include following sub-step：
S431：According to the three dimensional space coordinate of the extraction, it is calculated right big with described by the left thigh peak
Between lower limb peak, the midpoint of line is pointed to the primary vector of the head peak, points to the left side by the left thigh peak
The secondary vector of foot minimum point and the 3rd vector by the right thigh peak sensing right crus of diaphragm minimum point.
For example, as shown in figure 5, the corresponding first three dimensional space coordinate P1 of the head peak pixel that obtains is (x1,y1,z1)、
The corresponding second three dimensional space coordinate P2 of left thigh peak pixel is (x2,y2,z2), right thigh peak pixel corresponding
Three three dimensional space coordinate P3 are (x3,y3,z3), the corresponding 4th three dimensional space coordinate P4 of left foot minimum point pixel be (x4,y4,z4)
The 5th three dimensional space coordinate P5 corresponding with right crus of diaphragm minimum point pixel is (x5,y5,z5), above-mentioned x coordinate, y-coordinate are the collection depth
The world coordinates of the camera of degree image, z coordinate is the depth value of the pixel in depth image, can be calculated by above-mentioned space coordinatess
Obtaining primary vector isSecondary vectorWith the 3rd vector
S432：It is calculated the people's in the object space using the primary vector, secondary vector and three-dimensional amount
Height.
For example, by the primary vectorSecondary vectorWith the 3rd vectorFollowing formula 11 or formula 12 is substituted into, is obtained
Height Height of the people in the object space.
In the present embodiment, using object space in three dimensional space coordinate of the related parts of human body in depth image calculate
Its height is obtained, the height can be caused to be accurately calculated.Human body is not needed to stand with a certain anchor according to this algorithm,
Make one also accurately to measure in motor process, realize monitoring in real time.
Wherein, the people in the object space is calculated using the view data related to sex that extracts in above-mentioned S42
Height.Specifically such as, the height of the people in the view data related to sex and the object space of the extraction is input into
Classified to grader is set, and the property of the people in the object space is determined according to the classification results for setting grader
Not.The setting grader can be nearest neighbor classifier in prior art, Bayes classifier, support vector machine etc..The mesh
The height of the people in mark space can be obtained by the mode of the above-mentioned acquisition height in this S43 or by user input or be adopted
It is calculated in other ways.
Wherein, the people in the object space is calculated using the view data related to body weight that extracts in above-mentioned S42
Height.Concrete such as, by the height and property of the people in the view data related to body weight of the extraction and the object space
Be not input into setting regression model and body weight is estimated, and the body of the people in the object space is determined according to estimation result
Weight.The height of the people in the object space, sex can be obtained by the above-mentioned acquisition height and property in this S43 otherwise or
It is calculated by user input or using other modes.Due to input feature vector polynary, thus can using multiple linear return
Return model as above-mentioned setting regression model, body weight Weight of the people in object space is obtained for example with following formula 14,
Wherein, ωiIt is weight coefficient,For obtain the image data value related to body weight, the n be extraction and body weight
The sum of related view data.
Above-mentioned set that grader is trained to be obtained, this sets what regression model was arrived as study of Confucian classics acquistion.For example, sharp in advance
With depth camera to different heights, surname not and the crowd of body weight measures, and accurately obtain its surname not, height, body weight
Information, these view data is trained to grader and regression model and is learnt.
It is understood that in other embodiment, only can extract in height, sex and body weight wherein one or two
Dependent image data, wherein one or two in the height of the people being calculated in the object space then, sex and body weight.
S44：The human body information of the people of the object space is sorted out, obtains the crowd belonging to the people in the object space
Type.
S45：The advertising message with crowd's type matching is searched, and the advertising message for finding is broadcast
Put.
The present embodiment is used for playing advertising message.Such as, if the people in object space exceedes half or quantity reaches setting
Value belongs to obese type, then throw in fat-reducing series advertisements information；If the people in object space reaches setting value more than half or quantity
Child, then throw in the advertising messages such as toy product；If the people in object space exceedes half or quantity reaches setting value for female
Property, then throw in the advertising messages such as cosmetics.
Further, playing device can also according to the depth image recognize people in the object space whether wearing spectacles,
Walking manner etc., and according to recognition result and the human body information, play the related advertising message of glasses, sports products.
Fig. 7 is referred to, Fig. 7 is the structural representation of one embodiment of information play apparatus of the present invention.In the present embodiment, the letter
Breath playing device 70 includes acquisition module 71, obtains module 72 and playing module 73.Wherein,
Acquisition module 71 is used for the depth image for obtaining object space；
Module 72 is obtained for being analyzed to the depth image, the human body information in the object space is obtained, its
In, the human body information includes the height of the people in object space, body weight, at least one of sex；
Playing module 73 is used for playing the information that mates with the human body information.
Alternatively, this obtains module 72 includes extraction unit 721 and obtains unit 722.
Extraction unit 721 is used for extracting the view data related to human body information from the depth image.Obtain unit
722 are used for obtaining the human body information of the object space using the view data of the extraction.
Alternatively, when the human body information includes height, the extraction unit 721 specifically for：From the depth map
The view data related to the height is extracted as in, and the related view data of the height includes the head in the object space
Corresponding first three dimensional space coordinate of portion's peak pixel, corresponding second three dimensional space coordinate of left thigh peak pixel, the right side
Corresponding 3rd three dimensional space coordinate of thigh peak pixel, corresponding 4th three dimensional space coordinate of left foot minimum point pixel and the right side
Corresponding 5th three dimensional space coordinate of foot minimum point pixel；
Obtain unit 722 specifically for：According to the three dimensional space coordinate of the extraction, it is calculated by the left thigh most
Between high point and the right thigh peak midpoint of line point to the head peak primary vector, by the left thigh most
High point is pointed to the secondary vector of the left foot minimum point and is pointed to the of the right crus of diaphragm minimum point by the right thigh peak
Three vectors；The height that the people in the object space is calculated using the primary vector, secondary vector and three-dimensional amount.
Still optionally further, obtain module 63 to be further used for the primary vectorSecondary vectorWith the 3rd vectorAbove-mentioned formula 11 or formula 12 is substituted into, height Height of the people in the object space is obtained.
Alternatively, when the human body information includes sex, the extraction unit 721 specifically for：From the depth map
The view data related to sex is extracted as in, wherein, the view data related to sex is included in the object space
The contour feature of face, the contour feature of shoulder, the contour feature of the contour feature of chest and buttocks；
Obtain unit 722 specifically for：By in the view data related to sex and the object space of the extraction
People height be input into set grader classified, and according to described set grader classification results determine the target
The sex of the people in space.
Alternatively, when the human body information includes body weight, extraction unit 721 specifically for：From the depth image
The view data related to body weight is extracted, wherein, the view data related to body weight includes the people in the object space
The contour feature of the contour feature of face, the contour feature of shoulder, the contour feature of chest, the contour feature of buttocks and abdominal part；
Obtain unit 722 specifically for：By in the view data related to body weight of the extraction and the object space
People height and sex be input into setting regression model and body weight estimated, and the target empty is determined according to estimation result
Between in people body weight.
Still optionally further, the view data related to sex/body weight of the extraction is also included the object space
In people the normal in front and the optical axis of the camera for obtaining the depth image between angle α.
Wherein, the angle α can be by the three dimensional space coordinate P2 (x2,y2,z2) and the three dimensional space coordinate P3 (x3,
y3,z3) substitute into above-mentioned formula 13 be calculated.
Alternatively, the contour feature of certain human body in the object space is right in the object space
One of corresponding human body in the contour line of human body, the object space is answered parallel to object space place plane
Section contour line length and width between ratio, corresponding human body in the object space multiple parallel to institute
State in meansigma methodss or the object space of ratio between the length of the contour line in the section of object space place plane and width
Corresponding human body multiple sections parallel to object space place plane contour line length and width between
Maximum in ratio.
Alternatively, the playing module 73 is sorted out specifically for the human body information by the people of the object space, obtains described
Crowd's type belonging to people in object space；The advertising message with crowd's type matching is searched, and is found described
Advertising message play out.
Wherein, the above-mentioned module of the information play apparatus is respectively used to execute the corresponding steps in said method embodiment,
The as above embodiment of the method explanation of concrete implementation procedure, therefore not to repeat here.
Fig. 8 is referred to, Fig. 8 is the structural representation of one way of example of information broadcasting system of the present invention.In the present embodiment,
The system 80 includes the image capture device 81 being connected with each other and playback equipment 82.The playback equipment 82 include input interface 821,
Processor 822, memorizer 823 and player 824.
Image capture device 81 is used for the depth image for gathering object space, and exports to the playback equipment 82.Specifically,
The image capture device 81 may include to imitate two cameras arranging in the eyes of human body, with using in said method embodiment based on
Binocular vision mode collects the depth image；The image capture device 81 also can be as shown in Fig. 2 with using said method reality
Apply, in example, the depth image is collected based on structure light mode；Image capture device 81 may also include invisible light source and not
Visible Light Camera, to collect the depth image based on TOF modes in using said method embodiment.Certainly, image is adopted
It can be also other structures to integrate equipment 81, therefore here is not limited.
The input interface 821 of playback equipment 82 is used for obtaining the depth image that image capture device 81 is collected.
Memorizer 823 is used for storing computer instruction, and provides the computer instruction to processor 822, and can store
The data used needed for 822 processing procedure of processor are as being used for being calculated sex, the setting grader of body weight and returning mould
Formula in type and said method embodiment etc., and the depth image that input interface 821 is obtained.
Processor 822 executes the computer instruction, is used for：
Obtain the depth image of object space；
The depth image is analyzed, the human body information in the object space is obtained, wherein, the human body information
The height of the people in including object space, body weight, at least one of sex；
To the information that player output is mated with the human body information；
Player 824 is used for playing the information that mates with human body information.Specifically, the player may include display screen
And/or speaker, the information that mates with human body information is played for display and/or sound.
Alternatively, the image capture device 81 is may be provided in the playback equipment 82, i.e., the image capture device 81 and broadcast
Equipment 82 is put for same equipment.
Alternatively, processor 822 is specifically for extracting the view data related to human body information from the depth image；
The human body information that the object space is obtained using the view data of the extraction.
Specifically, when the human body information includes height, processor 822 can be specifically for：From the depth image
Extract related to the height view data, the view data of the height correlation includes head in the object space most
Corresponding first three dimensional space coordinate of high point pixel, corresponding second three dimensional space coordinate of left thigh peak pixel, right thigh
Corresponding 3rd three dimensional space coordinate of peak pixel, corresponding 4th three dimensional space coordinate of left foot minimum point pixel and right crus of diaphragm are most
Corresponding 5th three dimensional space coordinate of low spot pixel；According to the three dimensional space coordinate of the extraction, it is calculated by described left big
Between lower limb peak and the right thigh peak midpoint of line point to the primary vector of the head peak, by described left big
Lower limb peak points to the secondary vector of the left foot minimum point and points to the right crus of diaphragm minimum point by the right thigh peak
The 3rd vector；The height in the object space is calculated using the primary vector, secondary vector and three-dimensional amount.
Further, processor 822 is used for the primary vectorSecondary vectorWith the 3rd vectorSubstitute into above-mentioned public affairs
Formula 11 or formula 12, obtain height Height of the people in the object space.
Specifically, when the human body information includes sex, processor 822 can be specifically for：From the depth image
The view data related to sex is extracted, wherein, the view data related to sex includes the people in the object space
The contour feature of the contour feature of face, the contour feature of shoulder, the contour feature of chest and buttocks；By the extraction and property
The height of the people in not related view data and the object space is input into setting grader is classified, and according to described
The classification results for setting grader determine the sex of the people in the object space.
Specifically, when the human body information includes body weight, processor 822 can be specifically for carrying from the depth image
The view data related to body weight is taken, wherein, the view data related to body weight includes the face in the object space
Contour feature, the contour feature of shoulder, the contour feature of chest, the contour feature of the contour feature of buttocks and abdominal part；Will
The height and sex of the people in the view data related to body weight of the extraction and the object space is input into setting and returns
Model is estimated to body weight, and determines the body weight of the people in the object space according to estimation result.
Further, the view data related to sex/body weight of the extraction is also included the people in the object space
The normal in front and the optical axis of the camera for obtaining the depth image between angle α.
Wherein, the angle α can be by the three dimensional space coordinate P2 (x2,y2,z2) and the three dimensional space coordinate P3 (x3,
y3,z3) substitute into above-mentioned formula 13 be calculated.
Alternatively, the contour feature of certain human body in the object space is corresponding to for the object space
The contour line of human body, cutting parallel to object space place plane of the corresponding human body of the object space
The corresponding human body of ratio, the object space between the length of the contour line in face and width multiple parallel to the target
The meansigma methodss of the ratio between the length of the contour line in the section of space place plane and width or the corresponding people of the object space
In ratio between the length of the contour line in multiple sections parallel to object space place plane of body region and width
Maximum.
Alternatively, processor 822 executes the information that mates to player output with the human body information, specifically
Including：The human body information of the people of the object space is sorted out, obtains the crowd's type belonging to the people in the object space；Look into
The advertising message with crowd's type matching is looked for, and to the advertising message found described in player output.Described broadcast
Put the advertising message that device is further used for finding described in broadcasting.
The playback equipment 82 can be as above-mentioned playing device, for executing above-described embodiment methods described.For example, above-mentioned
The method that invention embodiment is disclosed is applied in processor 822, or is realized by processor 822.Processor 822 is probably one
Plant IC chip, the disposal ability with signal.During realization, each step of said method can pass through processor
The instruction of the integrated logic circuit of the hardware in 822 or software form is completed.Above-mentioned processor 822 can be general procedure
Device, digital signal processor (DSP), special IC (ASIC), ready-made programmable gate array (FPGA) or other can compile
Journey logical device, discrete gate or transistor logic, discrete hardware components.Can realize or execute the embodiment of the present invention
In disclosed each method, step and logic diagram.General processor can be microprocessor or the processor can also be
Any conventional processor etc..The step of in conjunction with method disclosed in the embodiment of the present invention, can be embodied directly at hardware decoding
Reason device is executed and is completed, or is completed with the hardware in decoding processor and software module combination execution.Software module may be located at
Random access memory, flash memory, read only memory, programmable read only memory or electrically erasable programmable memory, depositor etc.
In the ripe storage medium in this area.The storage medium is located at memorizer 823, and processor 822 reads the letter in respective memory
Breath, the step of complete said method in conjunction with its hardware.
In such scheme, it is analyzed by the depth image from the object space comprising human body and obtains the object space
Human body information, and then play the information that mates with the human body information of the object space, it is achieved that broadcast according to human body information self adaptation
Corresponding informance is put, and depth image is additionally, since comprising three-dimensional information, therefore the people obtained using the view data in the depth image
Body information is more accurate, and as depth image is for characterizing depth information, and do not affected by ambient light, therefore above-mentioned human body information
Collection do not limited by ambient light, i.e., no matter which kind of environment does not interfere with the collection of the human body information, improves the human body information
The stability of collection, therefore also improve the accurately fixed and reliability of information self-adapting broadcasting.
Embodiments of the present invention are the foregoing is only, the scope of the claims of the present invention is not thereby limited, every using this
Equivalent structure or equivalent flow conversion that description of the invention and accompanying drawing content are made, or directly or indirectly it is used in other correlations
Technical field, is included within the scope of the present invention.
Claims (10)
1. a kind of information broadcasting method, it is characterised in that include：
Obtain the depth image of object space；
The depth image is analyzed, the human body information in the object space is obtained, wherein, the human body information includes
The height of the people in object space, body weight, at least one of sex；
Play the information that mates with the human body information.
2. information broadcasting method according to claim 1, it is characterised in that described the depth image is analyzed,
Human body information in the object space is obtained, including：
The view data related to human body information is extracted from the depth image；
The human body information that the object space is obtained using the view data of the extraction.
3. method according to claim 2, it is characterised in that when the human body information includes height, described from described
The view data related to human body information is extracted in depth image, including：
The view data related to the height is extracted from the depth image, and the related view data of the height includes institute
State corresponding first three dimensional space coordinate of head peak pixel in object space, left thigh peak pixel corresponding second
Corresponding 3rd three dimensional space coordinate of three dimensional space coordinate, right thigh peak pixel, left foot minimum point pixel the corresponding 4th
Three dimensional space coordinate and corresponding 5th three dimensional space coordinate of right crus of diaphragm minimum point pixel；
The view data using the extraction obtains the human body information in the object space, including：
According to the three dimensional space coordinate of the extraction, it is calculated by between the left thigh peak and the right thigh peak
Point to the primary vector of the head peak, the left foot minimum point is pointed to by the left thigh peak in the midpoint of line
Secondary vector and the 3rd vector by the right thigh peak sensing right crus of diaphragm minimum point；
The height in the object space is calculated using the primary vector, secondary vector and three-dimensional amount.
4. method according to claim 3, it is characterised in that described using the primary vector, secondary vector and the 3rd
Vector is calculated the height of the people in the object space, including：
By the primary vectorSecondary vectorWith the 3rd vectorFollowing formula 1 or formula 2 is substituted into, the object space is obtained
In people height Height,
5. method according to claim 2, it is characterised in that when the human body information includes sex, described from described
The view data related to human body information is extracted in depth image, including：
The view data related to sex, wherein, the view data bag related to sex is extracted from the depth image
Include the profile of contour feature, the contour feature of shoulder, the contour feature of chest and the buttocks of face in the object space
Feature；
The view data using the extraction obtains the human body information in the object space, including：
The height of the people in the view data related to sex and the object space of the extraction is input into setting classification
Device is classified, and determines the sex of the people in the object space according to the classification results for setting grader.
6. method according to claim 2, it is characterised in that
When the human body information includes body weight, the extraction picture number related to human body information from the depth image
According to, including：
The view data related to body weight, wherein, the view data bag related to body weight is extracted from the depth image
Include the contour feature of face in the object space, the contour feature of shoulder, the contour feature of chest, the contour feature of buttocks
And the contour feature of abdominal part；
The view data using the extraction obtains the human body information in the object space, including：
The height and sex of the people in the view data related to body weight of the extraction and the object space are input into setting
Determine regression model to estimate body weight, and the body weight of the people in the object space is determined according to estimation result.
7. the method according to any one of claim 1 to 6, it is characterised in that the broadcasting is mated with the human body information
Information, including：
The human body information of the people of the object space is sorted out, obtains the crowd's type belonging to the people in the object space；
The advertising message with crowd's type matching is searched, and the advertising message for finding is played out.
8. a kind of information play apparatus, it is characterised in that include：
Acquisition module, for obtaining the depth image of object space；
Module is obtained, for being analyzed to the depth image, the human body information in the object space is obtained, wherein, institute
Stating human body information includes the height of the people in object space, body weight, at least one of sex；
Playing module, for playing the information that mates with the human body information.
9. a kind of information broadcasting system, it is characterised in that including the image capture device that is connected with each other and playback equipment, described broadcasts
The equipment of putting includes processor, memorizer and player；
Described image collecting device is used for the depth image for gathering object space, and exports to the playback equipment；
The memorizer of the playback equipment is used for storing computer instruction and the processor in use needed for processing procedure
View data；
Computer instruction described in the computing device, is used for：
Obtain the depth image of object space；
The depth image is analyzed, the human body information in the object space is obtained, wherein, the human body information includes
The height of the people in object space, body weight, at least one of sex；
To the information that player output is mated with the human body information；
The player is used for playing the information that mates with human body information.
10. system according to claim 9, it is characterised in that described image collecting device is arranged on the playback equipment
In.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610850683.9A CN106504283A (en) | 2016-09-26 | 2016-09-26 | Information broadcasting method, apparatus and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610850683.9A CN106504283A (en) | 2016-09-26 | 2016-09-26 | Information broadcasting method, apparatus and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106504283A true CN106504283A (en) | 2017-03-15 |
Family
ID=58290575
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610850683.9A Pending CN106504283A (en) | 2016-09-26 | 2016-09-26 | Information broadcasting method, apparatus and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106504283A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107330974A (en) * | 2017-07-31 | 2017-11-07 | 广东欧珀移动通信有限公司 | merchandise display method, device and mobile device |
| CN107515844A (en) * | 2017-07-31 | 2017-12-26 | 广东欧珀移动通信有限公司 | Font method to set up, device and mobile device |
| CN108619727A (en) * | 2018-04-20 | 2018-10-09 | 陈军 | Swing based on bodily form detection manages system |
| CN109224460A (en) * | 2018-04-20 | 2019-01-18 | 陈军 | Swing management method based on bodily form detection |
| CN111609908A (en) * | 2019-02-25 | 2020-09-01 | 西门子医疗有限公司 | Patient weight estimation from surface data using a patient model |
| CN114035717A (en) * | 2020-07-20 | 2022-02-11 | 爱思开海力士有限公司 | Motion recognition apparatus using TOF sensor and method of operating the same |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102654913A (en) * | 2011-03-04 | 2012-09-05 | 谢韬 | Method for selectively delivering advertisement |
| CN103474010A (en) * | 2013-09-22 | 2013-12-25 | 广州中国科学院软件应用技术研究所 | Video analysis-based intelligent playing method and device of outdoor advertisement |
| CN103956128A (en) * | 2014-05-09 | 2014-07-30 | 东华大学 | Intelligent active advertising platform based on somatosensory technology |
| CN104881642A (en) * | 2015-05-22 | 2015-09-02 | 海信集团有限公司 | Method and device for content pushing, and equipment |
-
2016
- 2016-09-26 CN CN201610850683.9A patent/CN106504283A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102654913A (en) * | 2011-03-04 | 2012-09-05 | 谢韬 | Method for selectively delivering advertisement |
| CN103474010A (en) * | 2013-09-22 | 2013-12-25 | 广州中国科学院软件应用技术研究所 | Video analysis-based intelligent playing method and device of outdoor advertisement |
| CN103956128A (en) * | 2014-05-09 | 2014-07-30 | 东华大学 | Intelligent active advertising platform based on somatosensory technology |
| CN104881642A (en) * | 2015-05-22 | 2015-09-02 | 海信集团有限公司 | Method and device for content pushing, and equipment |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107330974A (en) * | 2017-07-31 | 2017-11-07 | 广东欧珀移动通信有限公司 | merchandise display method, device and mobile device |
| CN107515844A (en) * | 2017-07-31 | 2017-12-26 | 广东欧珀移动通信有限公司 | Font method to set up, device and mobile device |
| CN107515844B (en) * | 2017-07-31 | 2021-03-16 | Oppo广东移动通信有限公司 | Font setting method and device and mobile device |
| CN108619727A (en) * | 2018-04-20 | 2018-10-09 | 陈军 | Swing based on bodily form detection manages system |
| CN109224460A (en) * | 2018-04-20 | 2019-01-18 | 陈军 | Swing management method based on bodily form detection |
| CN111609908A (en) * | 2019-02-25 | 2020-09-01 | 西门子医疗有限公司 | Patient weight estimation from surface data using a patient model |
| US11703373B2 (en) | 2019-02-25 | 2023-07-18 | Siemens Healthcare Gmbh | Patient weight estimation from surface data using a patient model |
| CN114035717A (en) * | 2020-07-20 | 2022-02-11 | 爱思开海力士有限公司 | Motion recognition apparatus using TOF sensor and method of operating the same |
| US11928892B2 (en) | 2020-07-20 | 2024-03-12 | SK Hynix Inc. | Motion recognition apparatus using ToF sensor and method for operating the same |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106504283A (en) | Information broadcasting method, apparatus and system | |
| CN107748869B (en) | 3D face identity authentication method and device | |
| CN106529399A (en) | Human body information acquisition method, device and system | |
| US7974443B2 (en) | Visual target tracking using model fitting and exemplar | |
| Parameswaran et al. | View invariance for human action recognition | |
| CN102222347B (en) | Creating range image through wave front coding | |
| US8282481B2 (en) | System and method for cyber training of martial art on network | |
| CN101715581B (en) | Volume recognition method and system | |
| CN102262725B (en) | The analysis of three-dimensional scenic | |
| CN106529400A (en) | Mobile terminal and human body monitoring method and device | |
| CN109477966A (en) | The head-mounted display for virtual reality and mixed reality with interior-external position tracking, user's body tracking and environment tracking | |
| CN114399826A (en) | Image processing method and apparatus, image device, and storage medium | |
| CN105740780A (en) | Method and device for human face in-vivo detection | |
| CN106484115A (en) | For strengthening the system and method with virtual reality | |
| US8565477B2 (en) | Visual target tracking | |
| US20100197392A1 (en) | Visual target tracking | |
| CN107481304A (en) | The method and its device of virtual image are built in scene of game | |
| CN110717391A (en) | Height measuring method, system, device and medium based on video image | |
| CN106471523B (en) | Color identification using infrared imaging | |
| CN113239797B (en) | Human body action recognition method, device and system | |
| CN102647606A (en) | Stereoscopic image processor, stereoscopic image interaction system and stereoscopic image display method | |
| CN105912991A (en) | Behavior identification method based on 3D point cloud and key bone nodes | |
| CN105989326A (en) | Method and device for determining three-dimensional position information of human eyes | |
| CN108280421A (en) | Human bodys' response method based on multiple features Depth Motion figure | |
| CN107656611A (en) | Somatic sensation television game implementation method and device, terminal device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170315 |