CN109118513A - A kind of calculation method and system of two-value sports immunology - Google Patents
A kind of calculation method and system of two-value sports immunology Download PDFInfo
- Publication number
- CN109118513A CN109118513A CN201810907725.7A CN201810907725A CN109118513A CN 109118513 A CN109118513 A CN 109118513A CN 201810907725 A CN201810907725 A CN 201810907725A CN 109118513 A CN109118513 A CN 109118513A
- Authority
- CN
- China
- Prior art keywords
- target video
- video frame
- block
- norm
- value
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/20—Analysis of motion
- G06T7/207—Analysis of motion for motion estimation over a hierarchy of resolutions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/40—Scenes; Scene-specific elements in video content
-
- 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
- G06T2207/10—Image acquisition modality
- G06T2207/10016—Video; Image sequence
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a kind of calculation methods of two-value sports immunology, comprising: obtains and obtains and the corresponding central block of each target video frame and each basic block in the target video frame group according to adjacent blocks division rule；Calculate the norm of each target video frame corresponding central block and each basic block；According to the norm of each target video frame corresponding central block and each basic block, the first two-value sports immunology based on adjacent blocks similar movement corresponding with the target video group is determined respectively and determines the second two-value sports immunology based on adjacent blocks overall movement；Using the first two-value sports immunology and the second two-value sports immunology as the goal description of target video frame group.Above-mentioned calculation method, not only in target video frame central block and basic block divide by preset division rule, but also be included in adjacent blocks overall movement and goal description be described, improved the accuracy of goal description calculating.
Description
Technical field
The present invention relates to the calculation method of technical field of computer vision more particularly to a kind of two-value sports immunology and it is
System.
Background technique
As the important component of action recognition HAR system, describing son can be according to its property (such as floating type or two-value
Type etc.) and size directly influence the overall performance (accuracy rate, rate etc.) of system.In the prior art, since floating-point movement is retouched
State that sub- calculating speed is slow, and memory consumption is big, it is preferred to use two-value sports immunology.The calculating of existing two-value sports immunology
Adjacent blocks overall movement is used in journey.
Inventor to it is existing based on adjacent blocks overall movement calculate two valued description calculating process progress the study found that
Basic block distance center block has a certain distance, leads to the two-value sports immunology inaccuracy being calculated.
Summary of the invention
In view of this, the present invention provides the calculation method and system of a kind of two-value sports immunology, it is existing to solve
In technology using based on adjacent blocks overall movement calculate the two-value sports immunology period of the day from 11 p.m. to 1 a.m, basic block distance center block have it is certain away from
From the problem for causing two-value sports immunology being calculated inaccurate.Concrete scheme is as follows:
A kind of calculation method of two-value sports immunology, comprising:
When detecting the computation requests to two-value sports immunology of target video frame group, obtains and divided according to adjacent blocks
Rule obtains and the corresponding central block of each target video frame and each basic block in the target video frame group；
Calculate the norm of each target video frame corresponding central block and each basic block；
According to the norm of each target video frame corresponding central block and each basic block, determine and the target respectively
Corresponding the first two-value sports immunology based on adjacent blocks similar movement of video group is with determining based on adjacent blocks overall movement
Second two-value sports immunology；
Using the first two-value sports immunology and the second two-value sports immunology as the target video frame group
Goal description.
Above-mentioned method optionally obtains and each mesh in the target video frame group according to adjacent blocks division rule
Mark the corresponding central block of video frame and each basic block, comprising:
For each of target video frame group target video frame, it is divided into the block square that dimension is 5 × 5
Battle array；
In each block matrix while block centered on the third line and tertial piece will be belonged to；
Remaining block except the central block will be removed in each block matrix as basic block.
Above-mentioned method optionally determines first based on adjacent blocks similar movement corresponding with the target video group
Two-value sports immunology, comprising:
It is right in next target video frame adjacent with each target video frame in the target video group to calculate separately
The norm for the central block answered is poor, obtains the first norm difference；
It is right in next target video frame adjacent with each target video frame in the target video group to calculate separately
The norm for each basic block answered is poor, obtains the second norm difference；
It is default to judge whether the absolute value of the difference of each second norm difference and first norm difference is greater than respectively
First threshold；
If it is not, returning to 1.
Above-mentioned method optionally determines the second two-value sports immunology based on adjacent blocks overall movement, comprising:
The norm for calculating separately the central block that each target video frame is adjacent in the target video group is adjacent thereto
The absolute value of the difference of the norm of each basic block of next target video frame, obtains third norm difference；
Judge whether each third norm difference is greater than preset second threshold；
If it is not, returning to 1.
Above-mentioned method, optionally, further includes:
Search goal-based assessment model corresponding with goal description；
Goal description is assessed according to the goal-based assessment model.
A kind of computing system of two-value sports immunology, comprising:
Module is obtained, for obtaining when detecting the computation requests to two-value sports immunology of target video frame group
Central block corresponding with each target video frame in the target video frame group and each is obtained according to adjacent blocks division rule
Basic block；
Computing module, for calculating the norm of each target video frame corresponding central block and each basic block；
First determining module, for the norm according to each target video frame corresponding central block and each basic block,
The first two-value sports immunology based on adjacent blocks similar movement corresponding with the target video group is determined respectively and determines base
In the second two-value sports immunology of adjacent blocks overall movement；
Second determining module is used for the first two-value sports immunology and the sub- conduct of the second two-value sports immunology
Goal description of the target video frame group.
Above-mentioned system, optionally, the acquisition module includes:
Division unit, for being divided into dimension for each of target video frame group target video frame
For 5 × 5 block matrix；
First determination unit, for will belong to centered on the third line and tertial piece simultaneously in each block matrix
Block；
Second determination unit, for remaining block except the central block will to be removed in each block matrix as basic
Block.
Above-mentioned system, optionally, first determining module includes:
First computing unit, it is adjacent with each target video frame next in the target video group for calculating separately
The norm of corresponding central block is poor in a target video frame, obtains the first norm difference；
Second computing unit, it is adjacent with each target video frame next in the target video group for calculating separately
The norm of each corresponding basic block is poor in a target video frame, obtains the second norm difference；
First judging unit, for judging the exhausted of each second norm difference and the difference of first norm difference respectively
Whether preset first threshold is greater than to value；
First return unit is used for if it is not, returning to 1.
Above-mentioned system, optionally, first determining module includes:
Third computing unit, for calculating separately the central block that each target video frame is adjacent in the target video group
Norm next target video frame each basic block adjacent thereto norm difference absolute value, obtain third norm
Difference；
Second judgment unit, for judging whether each third norm difference is greater than preset second threshold；
Second return unit is used for if it is not, returning to 1.
Above-mentioned system, optionally, further includes:
Searching module, for searching goal-based assessment model corresponding with goal description；
Evaluation module, for assessing according to the goal-based assessment model goal description.
Compared with prior art, the present invention includes the following advantages:
The invention discloses a kind of calculation methods of two-value sports immunology, comprising: obtains according to adjacent blocks division rule
It obtains and the corresponding central block of each target video frame and each basic block in the target video frame group；Calculate each mesh
Mark the norm of video frame corresponding central block and each basic block；According to each corresponding central block of target video frame and each
The norm of basic block determines that the first two-value movement based on adjacent blocks similar movement corresponding with the target video group is retouched respectively
It states son and determines the second two-value sports immunology based on adjacent blocks overall movement；By the first two-value sports immunology and institute
State goal description of second two-value sports immunology as the target video frame group.Above-mentioned calculation method, not only to mesh
Central block and basic block in mark video frame are divided by preset division rule, but also have been included in adjacent blocks overall movement
Goal description is described, the accuracy of goal description calculating is improved.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention for those of ordinary skill in the art without creative efforts, can be with
It obtains other drawings based on these drawings.
Fig. 1 is a kind of calculation method flow chart of two-value sports immunology disclosed in the embodiment of the present application；
Fig. 2 is a kind of another method flow diagram of calculation method of two-value sports immunology disclosed in the embodiment of the present application；
Fig. 3 is a kind of block matrix structural schematic diagram disclosed in the embodiment of the present application；
Fig. 4 is a kind of another method flow diagram of calculation method of two-value sports immunology disclosed in the embodiment of the present application；
Fig. 5 is a kind of another method flow diagram of calculation method of two-value sports immunology disclosed in the embodiment of the present application；
Fig. 6 is a kind of computing system structural block diagram of two-value sports immunology disclosed in the embodiment of the present application.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts every other
Embodiment shall fall within the protection scope of the present invention.
The foregoing description of the disclosed embodiments enables those skilled in the art to implement or use the present invention.
Various modifications to these embodiments will be readily apparent to those skilled in the art, as defined herein
General Principle can be realized in other embodiments without departing from the spirit or scope of the present invention.Therefore, of the invention
It is not intended to be limited to the embodiments shown herein, and is to fit to and the principles and novel features disclosed herein phase one
The widest scope of cause.
The present invention discloses a kind of calculation method of two-value sports immunology, and the calculation method, which is applied, was moving identification process
In in the calculating process of two valued description, the executing subject of the method can be processor, controller or calculate assessment
Model etc., the execution process of the calculation method as shown in Figure 1, comprising steps of
S101, when detecting the computation requests to two-value sports immunology of target video frame group, obtain according to closing on
Block division rule obtains and the corresponding central block of each target video frame and each basic block in the target video frame group；
In the embodiment of the present invention, the target video frame group is made of the target video frame of preset quantity, by a large amount of
Test and calculating analysis, it is preferred that when present count described in the target video frame group measures 3, calculating can be combined
Accuracy rate and calculating speed, the present invention implement in, in the target video frame group include 3 target video frames be illustrated,
Wherein, a central block and each basic block that each target video frame is divided according to adjacent blocks division rule are obtained.
S102, the norm for calculating each target video frame corresponding central block and each basic block；
In the embodiment of the present invention, the norm of central block and each basic block is calculated using formula (1), wherein preferred, institute
Stating norm can be that every element is absolute in block for this black Frobenius norm of not Luo Beini, the Frobenius norm
The summation of value square.Calculation method is as follows:
Wherein, A represents central block or the corresponding matrix of basic block,
Aij represents the element that i row j is arranged in matrix
S103, according to the norm of each target video frame corresponding central block and each basic block, determining and institute respectively
It states corresponding the first two-value sports immunology based on adjacent blocks similar movement of target video group and determines overall based on adjacent blocks
Second two-value sports immunology of movement；
In the embodiment of the present invention, according to the norm of each corresponding central block of target video frame and the model of each basic block
Number determines the first two-value sports immunology based on adjacent blocks similar movement corresponding with the target video group respectively and determines
The second two-value sports immunology based on adjacent blocks overall movement, in existing the first two-value fortune based on adjacent blocks similar movement
On the basis of dynamic description, and the second two-value sports immunology based on adjacent blocks overall movement is used, increases the accurate of calculating
Rate.
S104, using the first two-value sports immunology and the second two-value sports immunology as the target video
Goal description of frame group.
In the embodiment of the present invention, when target video frame each in the target video frame group is using described in S101-S103
Method, which calculates, to be completed, using obtained the first two-value sports immunology and the second two-value sports immunology as the mesh
Goal description for marking video frame group, is described the target video frame group according to goal description.
In the embodiment of the present invention, during actual action recognition, obtained video frame is continuously, to need continuous
Repeat S101-S104 calculation method, it is preferred that by obtained goal description corresponding with each target video frame group into
Row saves, convenient for assessing final action recognition process.
Preferably, the present invention establishes assessment models according to machine Learning Theory, by the assessment models Wei thatch it is graceful and
It is tested on human body behavior act identification data set.Entire action recognition process can be divided into training and two stages of test.
In the training stage, initial central block and each basic block are detected in training video frame sequence with block detection first, then
The first two-value sports immunology and the second two-value sports immunology are generated with the central block detected and each basic block.Will from institute
Description for having training video to obtain carries out k-means cluster.All cluster centres constitute a code book.According to description
To the distance of each cluster centre, each description is classified as the classification belonging to the nearest class heart.Then using each
The codebook element of Feature Descriptor come create one description training data in each classification goal-based assessment model.
In test phase, we extract block with mode identical with the training stage, generate description and distribution code book member
Element.On this basis, the goal-based assessment model corresponding with goal description is generated according to training process, utilizes support vector machines
Test video frame sequence is classified as classification the most possible, goal description is carried out according to the goal-based assessment model
Assessment, and calculate final accuracy rate.
The invention discloses a kind of calculation methods of two-value sports immunology, comprising: obtains according to adjacent blocks division rule
It obtains and the corresponding central block of each target video frame and each basic block in the target video frame group；Calculate each mesh
Mark the norm of video frame corresponding central block and each basic block；According to each corresponding central block of target video frame and each
The norm of basic block determines that the first two-value movement based on adjacent blocks similar movement corresponding with the target video group is retouched respectively
It states son and determines the second two-value sports immunology based on adjacent blocks overall movement；By the first two-value sports immunology and institute
State goal description of second two-value sports immunology as the target video frame group.Above-mentioned calculation method, not only to mesh
Central block and basic block in mark video frame are divided by preset division rule, but also have been included in adjacent blocks overall movement
Goal description is described, the accuracy of goal description calculating is improved.
Two-value sports immunology subspace proposed in the embodiment of the present invention is small, and can be calculated using simple mode
It obtains, this to calculate the sub speed of description quickly, can be applied to real-time system.Final goal description is in Weizmann
With can reach 40 frames/second speed on KTH data set.
Show that calculation method proposed by the invention is compared to the prior art in accuracy rate, speed based on a large amount of test
Good balance is reached between occupied space three.
In the embodiment of the present invention, obtain regarding with each target in the target video frame group according to adjacent blocks division rule
The method flow of the corresponding central block of frequency frame and each basic block as shown in Fig. 2, comprising steps of
S201, it is directed to each of target video frame group target video frame, being divided into dimension is 5 × 5
Block matrix；
In the embodiment of the present invention, the schematic diagram of the block matrix of any one target video frame is as shown in figure 3, include 25
Block, wherein include a central block and 24 basic blocks in 25 blocks, the piecemeal divides method to be true by test of many times
Fixed preferred division methods, there is no have the problem of certain distance to the division methods between central block and basic block.
S202, block centered on the third line and tertial piece will be belonged to simultaneously in each block matrix；
In the embodiment of the present invention, block centered on the third line and tertial piece is belonged to while by the block matrix,
The block of the center of the i.e. described block matrix.Wherein, it is 0 that the central block, which is block matrix signal figure label shown in Fig. 3,
Block.
S203, remaining block except the central block will be removed in each block matrix as basic block.
In the embodiment of the present invention, remaining each piece removed except the central block in the block matrix is used as substantially
Block, wherein the quantity of basic block is 24.
In the embodiment of the present invention, the first two-value based on adjacent blocks similar movement corresponding with the target video group is determined
The method flow of sports immunology as shown in figure 4, comprising steps of
S301, next target video frame adjacent with each target video frame in the target video group is calculated separately
In corresponding central block norm it is poor, obtain the first norm difference；
In the embodiment of the present invention, the model in the target video frame group with the central block of each target video frame is obtained
Number, and the norm of the central block in next target video frame adjacent thereto is obtained, the first norm is calculated according to formula (2)
Difference.
Wherein, A- current goal video frame central block
Next target video frame center block B- adjacent with the current goal video frame central block
- the first norm difference
S302, next target video frame adjacent with each target video frame in the target video group is calculated separately
In each corresponding basic block norm it is poor, obtain the second norm difference；
In the embodiment of the present invention, according to method described in formula (2), calculate separately basic with each target video frame
Norm of each corresponding basic block is poor in block next target video frame adjacent with its, traverses 24 basic blocks, obtains
24 second singular differences, wherein in formula (2), the basic block of A- current goal video frame；B- and the current goal video
Corresponding basic block (i.e. the basic block of same position) in the adjacent next target video frame of frame basic block；- the second
Norm difference
S303, judge whether each second norm difference and the absolute value of the difference of first norm difference are greater than respectively
Preset first threshold；
In the embodiment of the present invention, the absolute of each second norm difference and the difference of first norm difference is judged respectively
Value is compared with preset first threshold, judges the absolute of each second norm difference and the difference of first norm difference
Whether value is greater than preset first threshold.Wherein, the preset first threshold be empirical value according to specific identification process into
Row setting.
S304, if it is not, return 1.
In the embodiment of the present invention, when the absolute value of the difference of each second norm difference and first norm difference is greater than
When preset first threshold, 0 is returned to, conversely, returning to 1.
In the embodiment of the present invention, it is preferred that the target video frame group includes 3 target video frames, therefore, obtains one
The binary character string of 24 bits, it is preferred that be divided into 3 groups of every group of 8 bits according to shown in Fig. 3, be together in series generation one 3
Description of byte.If the length is 6 bytes for description of adjacent 3 frame of calculating.
In the embodiment of the present invention, the method flow of the second two-value sports immunology based on adjacent blocks overall movement is determined such as
Shown in Fig. 5, comprising steps of
S401, norm and its phase for calculating separately the central block that each target video frame is adjacent in the target video group
The absolute value of the difference of the norm of each basic block of next target video frame of neighbour, obtains third norm difference；
In the embodiment of the present invention, the calculation method according to formula (2) calculates separately each mesh in the target video group
Mark the difference of the norm of norm next target video frame each basic block adjacent thereto of the adjacent central block of video frame
Absolute value, obtain 24 third norm differences, wherein in formula (2), the central block of A- current goal video frame；B- and institute
State each of the adjacent next target video frame of current goal video frame basic block basic block；Third norm
Difference
S402, judge whether each third norm difference is greater than preset second threshold；
In the embodiment of the present invention, each third norm difference is compared with preset second threshold, is judged each
Whether a third norm difference is greater than preset second threshold, wherein the preset second threshold is empirical value according to specific
Identification process set.
S403, if it is not, return 1.
In the embodiment of the present invention, if third norm difference is greater than the preset second threshold, 0 is returned to, conversely, returning to 1.
In the embodiment of the present invention, it is preferred that the target video frame group includes 3 target video frames, therefore, obtains one
The binary character string of 24 bits, it is preferred that be divided into 3 groups of every group of 8 bits according to shown in Fig. 3, be together in series generation one 3
Description of byte.If the length is 6 bytes for description of adjacent 3 frame of calculating.
In the embodiment of the present invention, based on goal description obtained by the above method, between continuous three target video frames
Only take up 12 bytes.This enables goal description to calculate in many equipment with the limitation of calculating space, such as
Embedded system.
In the embodiment of the present invention, based on the calculation method of above-mentioned two-value sports immunology, the present invention also provides one kind
The computing system of two-value sports immunology, the Structure and Process of the computing system are as shown in Figure 6, comprising:
Obtain module 501, computing module 502, the first determining module 503 and the second determining module 504.
Wherein,
The acquisition module 501, for working as the computation requests detected to two-value sports immunology of target video frame group
When, it obtains and obtains central block corresponding with each target video frame in the target video frame group according to adjacent blocks division rule
With each basic block；
The computing module 502, for calculating the model of each target video frame corresponding central block and each basic block
Number；
First determining module 503, for according to each corresponding central block of target video frame and each basic block
Norm, determine respectively the first two-value sports immunology based on adjacent blocks similar movement corresponding with the target video group and
Determine the second two-value sports immunology based on adjacent blocks overall movement；
Second determining module 504, for retouching the first two-value sports immunology and second two-value movement
State goal description of the son as the target video frame group.
The invention discloses a kind of computing systems of two-value sports immunology, comprising: obtains according to adjacent blocks division rule
It obtains and the corresponding central block of each target video frame and each basic block in the target video frame group；Calculate each mesh
Mark the norm of video frame corresponding central block and each basic block；According to each corresponding central block of target video frame and each
The norm of basic block determines that the first two-value movement based on adjacent blocks similar movement corresponding with the target video group is retouched respectively
It states son and determines the second two-value sports immunology based on adjacent blocks overall movement；By the first two-value sports immunology and institute
State goal description of second two-value sports immunology as the target video frame group.Above-mentioned computing system, not only to mesh
Central block and basic block in mark video frame are divided by preset division rule, but also have been included in adjacent blocks overall movement
Goal description is described, the accuracy of goal description calculating is improved.
In the embodiment of the present invention, the acquisition module 501 includes:
Division unit 505, the first determination unit 506 and the second determination unit 507.
Wherein,
The division unit 505, for being divided for each of target video frame group target video frame
The block matrix for being 5 × 5 at dimension；
First determination unit 506, for the third line and tertial piece of work will to be belonged to simultaneously in each block matrix
Centered on block；
Second determination unit 507 is made for will remove remaining block except the central block in each block matrix
For basic block.
In the embodiment of the present invention, first determining module 503 includes:
First computing unit 508, the second computing unit 509, the first judging unit 510 and the first return unit 511.
Wherein,
First computing unit 508, for calculate separately in the target video group with each target video frame phase
The norm of corresponding central block is poor in adjacent next target video frame, obtains the first norm difference；
Second computing unit 509, for calculate separately in the target video group with each target video frame phase
The norm of each corresponding basic block is poor in adjacent next target video frame, obtains the second norm difference；
First judging unit 510, for judging each second norm difference and first norm difference respectively
Absolute value of the difference whether be greater than preset first threshold；
First return unit 511 is used for if it is not, returning to 1.
In the embodiment of the present invention, first determining module 503 includes:
Third computing unit 512, second judgment unit 513 and the second return unit 514.
Wherein,
The third computing unit 512 is adjacent for calculating separately each target video frame in the target video group
Central block norm next target video frame each basic block adjacent thereto norm difference absolute value, obtain
Third norm difference；
The second judgment unit 513, for judging whether each third norm difference is greater than preset second threshold；
Second return unit 514 is used for if it is not, returning to 1.
In the embodiment of the present invention, the computing system further include:
Searching module 515 and evaluation module 516.
Wherein,
The searching module 515, for searching goal-based assessment model corresponding with goal description；
The evaluation module 516, for assessing according to the goal-based assessment model goal description.
It should be noted that all the embodiments in this specification are described in a progressive manner, each embodiment weight
Point explanation is the difference from other embodiments, and the same or similar parts between the embodiments can be referred to each other.
For device class embodiment, since it is basically similar to the method embodiment, so being described relatively simple, related place ginseng
See the part explanation of embodiment of the method.
Finally, it is to be noted that, herein, relational terms such as first and second and the like be used merely to by
One entity or operation are distinguished with another entity or operation, without necessarily requiring or implying these entities or operation
Between there are any actual relationship or orders.Moreover, the terms "include", "comprise" or its any other variant meaning
Covering non-exclusive inclusion, so that the process, method, article or equipment for including a series of elements not only includes that
A little elements, but also including other elements that are not explicitly listed, or further include for this process, method, article or
The intrinsic element of equipment.In the absence of more restrictions, the element limited by sentence "including a ...", is not arranged
Except there is also other identical elements in the process, method, article or apparatus that includes the element.
For convenience of description, it is divided into various units when description apparatus above with function to describe respectively.Certainly, implementing this
The function of each unit can be realized in the same or multiple software and or hardware when invention.
As seen through the above description of the embodiments, those skilled in the art can be understood that the present invention can
It realizes by means of software and necessary general hardware platform.Based on this understanding, technical solution of the present invention essence
On in other words the part that contributes to existing technology can be embodied in the form of software products, the computer software product
It can store in storage medium, such as ROM/RAM, magnetic disk, CD, including some instructions are used so that a computer equipment
(can be personal computer, server or the network equipment etc.) executes the certain of each embodiment or embodiment of the invention
Method described in part.
The calculation method and system of a kind of two-value sports immunology provided by the present invention are described in detail above,
Used herein a specific example illustrates the principle and implementation of the invention, and the explanation of above embodiments is only used
In facilitating the understanding of the method and its core concept of the invention；At the same time, for those skilled in the art, according to the present invention
Thought, there will be changes in the specific implementation manner and application range, in conclusion the content of the present specification should not be construed as
Limitation of the present invention.
Claims (10)
1. a kind of calculation method of two-value sports immunology characterized by comprising
When detecting the computation requests to two-value sports immunology of target video frame group, obtain according to adjacent blocks division rule
It obtains and the corresponding central block of each target video frame and each basic block in the target video frame group；
Calculate the norm of each target video frame corresponding central block and each basic block；
According to the norm of each target video frame corresponding central block and each basic block, determine and the target video respectively
Corresponding the first two-value sports immunology based on adjacent blocks similar movement of group and determining second based on adjacent blocks overall movement
Two-value sports immunology；
Using the first two-value sports immunology and the second two-value sports immunology as the mesh of the target video frame group
Mark description.
2. the method according to claim 1, wherein being obtained and the target video according to adjacent blocks division rule
Each corresponding central block of target video frame and each basic block in frame group, comprising:
For each of target video frame group target video frame, it is divided into the block matrix that dimension is 5 × 5；
In each block matrix while block centered on the third line and tertial piece will be belonged to；
Remaining block except the central block will be removed in each block matrix as basic block.
3. the method according to claim 1, wherein determining corresponding with the target video group based on adjacent blocks
First two-value sports immunology of similar movement, comprising:
It calculates separately corresponding with the adjacent next target video frame of each target video frame in the target video group
The norm of central block is poor, obtains the first norm difference；
It calculates separately corresponding with the adjacent next target video frame of each target video frame in the target video group
The norm of each basic block is poor, obtains the second norm difference；
Judge whether the absolute value of the difference of each second norm difference and first norm difference is greater than preset the respectively
One threshold value；
If it is not, returning to 1.
4. the method according to claim 1, wherein determining the second two-value movement based on adjacent blocks overall movement
Description, comprising:
Calculate separately adjacent thereto next of norm for the central block that each target video frame is adjacent in the target video group
The absolute value of the difference of the norm of each basic block of a target video frame, obtains third norm difference；
Judge whether each third norm difference is greater than preset second threshold；
If it is not, returning to 1.
5. the method according to claim 1, wherein further include:
Search goal-based assessment model corresponding with goal description；
Goal description is assessed according to the goal-based assessment model.
6. a kind of computing system of two-value sports immunology characterized by comprising
Module is obtained, for obtaining foundation when detecting the computation requests to two-value sports immunology of target video frame group
Adjacent blocks division rule obtains central block corresponding with each target video frame in the target video frame group and each basic
Block；
Computing module, for calculating the norm of each target video frame corresponding central block and each basic block；
First determining module, for the norm according to each target video frame corresponding central block and each basic block, respectively
It determines the first two-value sports immunology based on adjacent blocks similar movement corresponding with the target video group and determines and be based on facing
Second two-value sports immunology of nearly block overall movement；
Second determining module is used for the first two-value sports immunology and the second two-value sports immunology as described in
Goal description of target video frame group.
7. system according to claim 6, which is characterized in that the acquisition module includes:
Division unit for for each of target video frame group target video frame, being divided into dimension is 5 ×
5 block matrix；
First determination unit, for block centered on the third line and tertial piece will to be belonged to simultaneously in each block matrix；
Second determination unit, for remaining block except the central block will to be removed in each block matrix as basic block.
8. system according to claim 6, which is characterized in that first determining module includes:
First computing unit, for calculating separately next mesh adjacent with each target video frame in the target video group
The norm for marking corresponding central block in video frame is poor, obtains the first norm difference；
Second computing unit, for calculating separately next mesh adjacent with each target video frame in the target video group
The norm for marking each corresponding basic block in video frame is poor, obtains the second norm difference；
First judging unit, for judging the absolute value of the difference of each second norm difference Yu first norm difference respectively
Whether preset first threshold is greater than；
First return unit is used for if it is not, returning to 1.
9. system according to claim 6, which is characterized in that first determining module includes:
Third computing unit, for calculating separately the model for the central block that each target video frame is adjacent in the target video group
The absolute value of the difference of the norm of number each basic block of next target video frame adjacent thereto, it is poor to obtain third norm
Value；
Second judgment unit, for judging whether each third norm difference is greater than preset second threshold；
Second return unit is used for if it is not, returning to 1.
10. system according to claim 6, which is characterized in that further include:
Searching module, for searching goal-based assessment model corresponding with goal description；
Evaluation module, for assessing according to the goal-based assessment model goal description.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810907725.7A CN109118513B (en) | 2018-08-10 | 2018-08-10 | Method and system for calculating binary motion descriptor |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810907725.7A CN109118513B (en) | 2018-08-10 | 2018-08-10 | Method and system for calculating binary motion descriptor |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109118513A true CN109118513A (en) | 2019-01-01 |
| CN109118513B CN109118513B (en) | 2022-01-11 |
Family
ID=64852171
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810907725.7A Active CN109118513B (en) | 2018-08-10 | 2018-08-10 | Method and system for calculating binary motion descriptor |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109118513B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102799871A (en) * | 2012-07-13 | 2012-11-28 | Tcl集团股份有限公司 | Method for tracking and recognizing face |
| CN103186775A (en) * | 2013-03-27 | 2013-07-03 | 西安电子科技大学 | Human body motion recognition method based on mixed descriptor |
| US20130259372A1 (en) * | 2012-03-28 | 2013-10-03 | Canon Kabushiki Kaisha | Method and apparatus for object classifier generation, and method and apparatus for detecting object in image |
-
2018
- 2018-08-10 CN CN201810907725.7A patent/CN109118513B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130259372A1 (en) * | 2012-03-28 | 2013-10-03 | Canon Kabushiki Kaisha | Method and apparatus for object classifier generation, and method and apparatus for detecting object in image |
| CN102799871A (en) * | 2012-07-13 | 2012-11-28 | Tcl集团股份有限公司 | Method for tracking and recognizing face |
| CN103186775A (en) * | 2013-03-27 | 2013-07-03 | 西安电子科技大学 | Human body motion recognition method based on mixed descriptor |
Non-Patent Citations (2)
| Title |
|---|
| MICHAEL CALONDER等: "BRIEF:Computing a Local Binary Descriptor Very Fast", 《IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE》 * |
| 张璇 等: "二维运动估值块匹配算法的探讨", 《万方数据库期刊库》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109118513B (en) | 2022-01-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109242041A (en) | A kind of electric energy meter abnormal deviation data examination method, device, equipment and storage medium | |
| CN101673316B (en) | Method for identifying supporting cables needing to adjust cable force based on hybrid monitoring | |
| CN105430099A (en) | Collaborative Web service performance prediction method based on position clustering | |
| WO2006132759A3 (en) | Method and apparatus for candidate evaluation | |
| CN103699795A (en) | Exercise behavior identification method and device and exercise intensity monitoring system | |
| CN112541532B (en) | Target detection method based on dense connection structure | |
| CN110232249A (en) | A kind of rolling bearing method for predicting residual useful life | |
| CN103544325B (en) | For data processing method and device that Webpage clicks on distribution | |
| CN109145446B (en) | Structural damage identification method based on modal strain energy and convolutional neural network | |
| CN111191696B (en) | Deep learning algorithm-based steel bar layering method and system | |
| SE1350894A1 (en) | Method and system for determining performance indicators for periodic movements | |
| CN106017475A (en) | Flight path updating method and flight path updating device | |
| CN107545038A (en) | A kind of file classification method and equipment | |
| CN110457737A (en) | A method of pollution entering the water is quickly positioned based on neural network | |
| CN110263871A (en) | Step counting cheat detection method, device, intelligent terminal and storage medium | |
| CN105334497A (en) | Three-dimensional positioning method and device | |
| CN108876056A (en) | A kind of shared bicycle Demand Forecast method, apparatus, equipment and storage medium | |
| CN108053424A (en) | Method for tracking target, device, electronic equipment and storage medium | |
| CN108415884A (en) | A kind of modal parameters real-time tracing method | |
| CN103927529B (en) | The preparation method and application process, system of a kind of final classification device | |
| CN113742983A (en) | Long-span structural damage identification method based on depth self-encoder neural network | |
| CN109118513A (en) | A kind of calculation method and system of two-value sports immunology | |
| CN109993188A (en) | Data label recognition methods, Activity recognition method and device | |
| CN111310121A (en) | New energy output probability prediction method and system | |
| CN111639194A (en) | Knowledge graph query method and system based on sentence vectors |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |