CN105868816A - Label image generation method and apparatus - Google Patents
Label image generation method and apparatus Download PDFInfo
- Publication number
- CN105868816A CN105868816A CN201610218723.8A CN201610218723A CN105868816A CN 105868816 A CN105868816 A CN 105868816A CN 201610218723 A CN201610218723 A CN 201610218723A CN 105868816 A CN105868816 A CN 105868816A
- Authority
- CN
- China
- Prior art keywords
- label image
- effective
- distance
- label
- images
- 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
- G06K—GRAPHICAL DATA READING; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
- G06K19/00—Record carriers for use with machines and with at least a part designed to carry digital markings
- G06K19/06—Record carriers for use with machines and with at least a part designed to carry digital markings characterised by the kind of the digital marking, e.g. shape, nature, code
- G06K19/06009—Record carriers for use with machines and with at least a part designed to carry digital markings characterised by the kind of the digital marking, e.g. shape, nature, code with optically detectable marking
- G06K19/06037—Record carriers for use with machines and with at least a part designed to carry digital markings characterised by the kind of the digital marking, e.g. shape, nature, code with optically detectable marking multi-dimensional coding
Abstract
The invention provides a label image generation method and apparatus. The method comprises the following steps: if the quantity of effective labels in an effective label image set is smaller than a first preset quantity, randomly generating a first label image; calculating a distance between the first label image and the effective label image set; and if the distance between the first label image and the effective label image set is greater than a preset distance threshold, determining that the first label image is an effective label image, and adding the first label image to the effective label image set. Therefore, there are certain distances between label images in the effective label image set, and there are quite large differences between the label images.
Description
Technical field
The invention belongs to technical field of image processing, especially relate to a kind of label image and generate method and dress
Put.
Background technology
In recent years, various types of label images such as such as two-dimension code label image, extensively should
It is used in the life of people, is such as used for identifying object, payment etc..These label images are by multiple volumes
Code unit generates according to certain coding rule, and typically carries data letter in these label images
Breath, this data message refers to the relevant information of the object of the used mark of label image, such as object
Title, product batch number etc..
But, in actual application, the use demand of a kind of label image is used for positioning.Citing
For, the such as such as four sides wall of the ad-hoc location in a room is respectively arranged with different labels
Image, when robot walks about in a room, can know based on the detection to label image
Go at any sidewalls, need to which directional steering to walk.Now, can be not in label image
Carrying data message, simply needs to generate different label images.
In the actual application of above-mentioned align_type, how to generate the label image with robustness, with
Ensure, between a large amount of label images generated, there is larger difference, in order to label figure in actual application
Accurately identifying of picture, is problem demanding prompt solution.
Summary of the invention
For the problem of above-mentioned existence, the present invention provides a kind of label image to generate method and apparatus, uses
To generate multiple label images with bigger diversity.
The invention provides a kind of label image and generate method, including:
If effectively the effective number of labels in label image set is less than the first predetermined number, then stochastic generation
First label image;
Calculate the distance of described first label image and described effective label image set;
If described first label image is more than predeterminable range threshold with the distance of described effective label image set
Value, it is determined that described first label image is effective label image, is added to by described first label image
In described effective label image set.
The invention provides a kind of label image generating means, including:
Processor, memorizer, described processor, described memorizer are connected by bus;
Wherein, storing batch processing code in described memorizer, described processor calls in described memorizer
The program code of storage is to perform following processing procedure:
If effectively the effective number of labels in label image set is less than the first predetermined number, then stochastic generation
First label image;
Calculate the distance of described first label image and described effective label image set;
If described first label image is more than predeterminable range threshold with the distance of described effective label image set
Value, it is determined that described first label image is effective label image, is added to by described first label image
In described effective label image set.
The label image that the present invention provides generates method and apparatus, if in the currently active label image set
Effective number of labels less than the first predetermined number of required generation, then one label image of stochastic generation,
And calculate the distance of this label image and effective label image set, when this label image and effective label figure
When the distance that image set closes is more than certain distance threshold value, this label image of generation and effective label image are described
Between other the effective label images existed in set, there is notable difference, now determine the mark of generation
Signing image is effective label image, is added in effective label image set, thus, effective label
All there is between each label image in image collection a certain distance, between each label image, there is bigger difference
The opposite sex.
Accompanying drawing explanation
Fig. 1 is the flow chart that label image of the present invention generates embodiment of the method one;
Fig. 2 is the flow chart that label image of the present invention generates embodiment of the method two;
Fig. 3 is the structural representation of label image generating means embodiment one of the present invention.
Detailed description of the invention
Fig. 1 is the flow chart that label image of the present invention generates embodiment of the method one, as it is shown in figure 1, the party
Method comprises the steps:
If the effective number of labels in step 101 effectively label image set is less than the first predetermined number,
Then stochastic generation the first label image.
When user needs the label image generating the first predetermined number, can first define an effective label
Image collection D, for depositing effective label image of generation.Time initial, this effective label image set D
For sky, when first label image of stochastic generation, directly this first label image generated is added
In effective label image set.Above-mentioned first label image is not to limit first label figure generated
Picture.
In the present embodiment, illustrating as a example by generating any one first label image, generating at needs should
During the first label image, first determine whether in effective label image set D the quantity of effective label image whether
Reach the first predetermined number, if not having, then one the first label image of stochastic generation.
Before the stochastic generation mode specifically introducing label image, the first definition to label image is said
Bright:
Assume with mgRepresent any one the first label image, then mg=(w0,w1,...,wn), wiIt it is the first label
Image mgIn the i-th row coding vector, i takes from 0 to n.Wherein, wi=(b0,b1,, bn|bj∈ 0,1}), bj
It it is the pixel value of jth coding unit in the i-th row coding vector.
It is to say, the image array that the coding unit that any one label image is arranged by n row n forms,
Each coding unit is as an element in this image array.This image array is by n column vector i.e.
N row coding vector forms, and each column coding vector is made up of n coding unit, and each coding unit is corresponding
Pixel value be 0 or 1, i.e. black or white binarized pixel.Therefore, each coding unit is the equal of
One coding unit lattice, fills with black or white.Therefore, generate label image to be equivalent to generate exactly
This image array, each coding unit in the image array that will generate afterwards is filled according to its pixel value with right
The color answered, thus define label image.
Therefore, when generating label image, the present embodiment is with each row in stochastic generation label image
The mode of coding vector generates label image, it is, of course, understood that if it is desired to label image
Being expressed as being made up of n row coding vector, often row coding vector is made up of n coding unit, stochastic generation
Often the principle of row coding vector similarly, does not repeats.
Specifically, stochastic generation the first label image, including:
The first label image m is determined according to formula (1)gIn the generating probability of each row coding vector
P (w=wi), with according to generating probability P (w=wi) generate the first label image mg:
Wherein, wγFor either rank coding vector, Z represents 2nPlant coded system.What deserves to be explained is, wγ
For either rank coding vector, it is not to limit wγIt is the first label image mgIn either rank coding vector,
And refer to: if either rank coding vector is made up of n coding unit, this n coding unit is altogether
The coded system existed is Z kind, then to there being Z kind coding vector, and namely middle wγFor Z kind coding to
Any one in amount.Above-mentioned generating probability refers to coding vector wiIt is the first label image mgMiddle string encodes
The probability of vector.
Wherein, T (wi) it is wiConversion ratio, determine according to formula (2);O(wi, D) and it is wiThere iing criterion
Sign the probability occurred in image collection D, determine according to formula (3):
Wherein,For wiIn q-th coding unit, whenTime, δ (wγ,wi) equal to 0, whenTime, δ (wγ,wi) equal to 1, | D | is label image quantity effective in D.
Below with the implication of the conversion ratio described in an above-mentioned formula (2) the brightest, it is assumed that wiBe by
The coding vector of 5 coding unit compositions, the coded system of these 6 coding unit compositions is 010110,
Then T (010110)=4/5, wherein, this conversion ratio embodies this wiIn the most adjacent two coding units
Diversity, in this citing, first coding unit is 0, and second coding unit is 1, from first
Coding unit to second coding unit changes, and varied number is designated as 1;Accordingly, from second volume
Code unit to the 3rd coding unit becomes 0 from 1, changes, and varied number adds 1 and becomes 2;From
Three coding unit to the 4th coding units become 1 from 0, change, and varied number adds 1 and becomes 3;
Becoming 1 from the 4th coding unit to the 5th coding unit from 1, do not change, varied number keeps 3
Constant;Become 0 from the 5th coding unit to the 6th coding unit from 1, change, change number
Amount adds 1 and becomes 4.And in theory, the varied number that these 6 coding units at most can have is 5, because of
This, T (010110)=4/5.For either rank coding vector, its conversion ratio is the highest, for label
For the identification of image, its can identification the highest, more can guarantee that the accuracy of identification.
Based on above-mentioned formula (2) and the implication of formula (3), formula (1) is interpreted as: if
wiConversion ratio the highest, it is at label image mgThe probability of middle generation should be the highest, otherwise, generation general
Rate should be the lowest;If wiOther label images generated in effective label image set occur
Probability is the highest, then it is at label image mgThe probability of middle generation should be the lowest, otherwise, the probability of generation
Should be the highest.
To sum up, the generating probability expressed by formula (1), is so that label image mgAnd have criterion
Sign other label images generated in image collection and there is more preferable diversity as far as possible.
Step 102, calculate the distance of the first label image and effective label image set.
In the present embodiment, the first label image mgWith the distance of effective label image set D, refer to first
Minimum range between other label images each in label image and effectively label image set D is the most permissible
The first label image m is calculated by equation below (4)gDistance with effective label image set D
L(mg, D):
Wherein, mhFor the arbitrary effective label image in D, L (mg,mh) it is mgWith mhBetween distance, root
Determine according to formula (5):
Wherein, H () represents Hamming distance, Rk(mh) represent mhRotate k × 90 degree.
If step 103 first label image is more than predeterminable range threshold with the distance of effective label image set
Value, it is determined that the first label image is effective label image, adds the first label image to effective label
In image collection.
In the present embodiment, if the first label image mgWith the distance of effective label image set D more than pre-
If distance threshold τ0, it is determined that the first label image mgFor effective label image, by the first label image mg
Add in effective label image set D.
It is understood that as a label image mgAfter adding in effective label image set D, can
To continue executing with step 101, i.e. judge whether the effective number of labels in effective label image set D reaches
First predetermined number, if reached, then end-tag image generation process, without reaching,
Then continue to generate new label image.
In the present embodiment, the minimum difference distance of the effective label image set D generated by the way
τ (D) can be expressed as:
Wherein,
Wherein, S (mh) it is arbitrary effective label image m in DhSpinning distance, mlFor in D any one
With mhDifferent effective label images.
The implication of this minimum difference distance τ (D) refers to, for the effective label image set D generated,
Its error-detecting and error correcting capability can reach floor [(τ (D)-1)/2], say, that effective when generate
When label image exists the coding unit of floor [(τ (D)-1)/2] individual mistake, it is possible to correct it as correspondence
Correct label image.Wherein, floor () is downward bracket function.And, the label image of generation because of
For carrying rotatory, therefore, it is possible to the direction that label image is put detected, have the highest in actual applications
Robustness and accuracy.
Alternatively, after step 103, it is also possible to comprise the steps 104:
Step 104, generating the framed picture corresponding with the first label image, the first label image is filled in
In the housing that framed picture characterizes.
In actual application, the label image of generation it is generally required to be attached on object, and, generation
Label image is bianry image, has black-and-white two color, if the background patterns of the object being attached at should
Label image is not easily distinguishable, for identifying that the identification device arranged on device, such as robot is accurately positioned
Existence to label image is likely to result in inconvenience.
Therefore, in the present embodiment, alternatively, it is also possible to peripheral at the label image generated, housing is generated
Image, this framed picture is such as made up of the black coding unit occupying certain coding unit quantity, i.e. phase
When in generating a circle dark border image in label image surrounding, in order to be accurately positioned label image.
In the present embodiment, if the effective number of labels in the currently active label image set is less than required life
The first predetermined number become, then one label image of stochastic generation, and calculate this label image and have criterion
Sign the distance of image collection, when the distance of this label image with effective label image set is more than certain distance
During threshold value, other illustrating to have existed in this label image and the effective label image set generated are effective
There is between label image notable difference, now determine that the label image of generation is effective label image, will
It adds in effective label image set, thus, each label image in effective label image set it
Between all there is a certain distance, there is larger difference between each label image.
Fig. 2 is the flow chart that label image of the present invention generates embodiment of the method two, as in figure 2 it is shown, at figure
On the basis of 1 illustrated embodiment, can also comprise the steps: after above-mentioned steps 102
If step 201 first label image is not more than predeterminable range with the distance of effective label image set
Threshold value, then abandon the first label image, adds one by invalid tag amount of images.
On the basis of embodiment illustrated in fig. 1, if the first label image calculated by step 102
mgIt is not more than predeterminable range threshold tau with the distance of effective label image set D0, the first label image is described
mgBig not with the diversity of other the effective label images in effective label image set D, the most permissible
Abandon this first label image m of generationg, invalid tag amount of images ρ is added one.
Step 202, judge invalid tag amount of images whether less than the second predetermined number, if being less than, then
Perform step 203-206, otherwise perform step 207-212.
Step 203, stochastic generation the second label image.
Step 204, judge that whether the distance of the second label image and effective label image set is more than presetting
Distance threshold, if more than predeterminable range threshold value, then performs step 205, otherwise, performs step 206.
Step 205, determine that the second label image is effective label image, add effective label image collection to
In conjunction, and invalid tag amount of images is set to 0.
Step 206, abandon described second label image, invalid tag amount of images is added one.
After step 206, the judgement performing step 202 processes.
In the present embodiment, as the first label image m generatedgFor invalid tag image, by invalid tag figure
After adding one as quantity ρ, it is judged that invalid tag amount of images has reached certain threshold value, i.e. second
Predetermined number.If invalid tag amount of images is less than the second predetermined number, currently employed presetting is described
Distance threshold τ0Can also continue to use, i.e. with this distance threshold τ0The most effective as label image
According to still setting up.Wherein, this establishment refers to use this distance threshold τ0, it would still be possible to meet user and generate
One predetermined number, distance difference is bigger each other effective label.
Now, according to formula (1) stochastic generation next one label image, the i.e. second label image, and then
With reference to embodiment illustrated in fig. 1, calculate the distance of the second label image and effective label image set D, as
Really the second label image and the distance of effective label image set D are more than above-mentioned predeterminable range threshold tau0, then
Illustrate that the second label image is effective label image, be added in effective label image set D, and
Invalid tag amount of images is set to 0, and then, carry out and the effective label in effective label image set D
Whether quantity reaches the judgement of the first predetermined number, without reaching, continues to generate new label image.
It is contrary, if the distance of the second label image and effective label image set D is not more than above-mentioned pre-
If distance threshold τ0, illustrate that the second label image is invalid tag image, then by invalid tag amount of images
ρ adds one.If invalid tag amount of images ρ now is still not reaching to the second predetermined number, then continue
Continuous execution step 203-step 206.
What deserves to be explained is, in the present embodiment, invalid tag amount of images ρ actually statistics is the most raw
How many invalid tag images are become, as long as being not reaching to the second present count in invalid tag amount of images ρ
Before amount, generate an effective label image, then invalid tag amount of images ρ is set to 0, now illustrates
Predeterminable range threshold tau0Or effective, i.e. still can be as the most effective judgment basis of label image
's.
Relative, if invalid tag amount of images ρ has reached the second predetermined number, then perform such as
Under step 207 and step process afterwards:
Step 207, updating described predeterminable range threshold value, the distance threshold after renewal is less than predeterminable range threshold
Value.
Step 208, judge the distance threshold after updating whether more than the cut-off distance threshold preset, if greatly
In, then perform step 209, otherwise terminate.
In the present embodiment, if invalid tag amount of images ρ has reached the second predetermined number, explanation
Predeterminable range threshold tau0The most invalid, using secondary as threshold value, although to ensure that between the label image of generation
Diversity bigger, but the label image of the first predetermined number needed for user cannot be generated.Therefore,
Need to update this threshold value, be such as updated to τ1=τ0-1, with the distance threshold τ after updating1As being subsequently generated
Label image whether effective foundation.
But, what deserves to be explained is, be not unrestrictedly to update distance threshold, because if unrestrictedly
Ground updates distance threshold, although may meet user and generate the demand of substantial amounts of label image, but, mark
The diversity signed between image can not preferably be ensured, therefore, in the present embodiment, updates distance threshold
There is certain restriction, if the distance threshold τ after current renewal1Reach the cut-off distance threshold limited,
Then end-tag image generation process.
Contrary, without reaching the cut-off distance threshold that limits, then can continue to generate new label
Image, the i.e. the 3rd label image, and use the distance threshold τ after renewal1As the 3rd label image whether
Effective judgment basis.
Step 209, stochastic generation the 3rd label image.
Step 210, judge that whether the distance of the 3rd label image and effective label image set is more than updating
After distance threshold, if more than update after distance threshold, then perform step 211, otherwise, perform step
Rapid 212.
Step 211, determine that described 3rd label image is effective label image, add to described effectively
In label image set, and invalid tag amount of images is set to 0.
Step 212, abandon described 3rd label image, invalid tag amount of images is added one.
After step 212, the judgement continuing executing with step 202 processes.
The process logic of above-mentioned steps 209 to step 212 may refer to described above, repeats no more.
In the present embodiment, by the quantity of the invalid tag that statistics continuously generates, and by distance threshold
Value update mechanism, is ensureing to have between the effective label image generated bigger diversity, and is ensureing
Reasonable tradeoff has been carried out between a number of label image needed for generating abundant user.
Fig. 3 is the structural representation of label image generating means embodiment one of the present invention, as it is shown on figure 3,
This label image generating means includes:
Processor 11, memorizer 12, described processor 11, described memorizer 12 are connected by bus 13.
Wherein, storing batch processing code in described memorizer 12, described processor 11 calls described memorizer
In 12, the program code of storage is to perform following processing procedure:
If effectively the effective number of labels in label image set is less than the first predetermined number, then stochastic generation
First label image;
Calculate the distance of described first label image and described effective label image set;
If described first label image is more than predeterminable range threshold with the distance of described effective label image set
Value, it is determined that described first label image is effective label image, is added to by described first label image
In described effective label image set.
Alternatively, described processor 11 is additionally operable to:
If the distance of described first label image and described effective label image set be not more than described preset away from
From threshold value, then abandon described first label image, invalid tag amount of images is added one;
If invalid tag amount of images is less than the second predetermined number, then stochastic generation the second label image;
If described second label image is more than described predeterminable range with the distance of described effective label image set
Threshold value, it is determined that described second label image is effective label image, adds described effective label image to
In set, and invalid tag amount of images is set to 0;
If the distance of described second label image and described effective label image set be not more than described preset away from
From threshold value, then abandon described second label image, invalid tag amount of images is added one.
Alternatively, described processor 11 is additionally operable to:
If the distance of described first label image and described effective label image set be not more than described preset away from
From threshold value, then abandon described first label image, invalid tag amount of images is added one;
If invalid tag amount of images is more than the second predetermined number, then update described predeterminable range threshold value, more
Distance threshold after Xin is less than predeterminable range threshold value;
If the distance threshold after Geng Xining is more than the cut-off distance threshold preset, then stochastic generation the 3rd label figure
Picture;
If described 3rd label image is more than the distance after updating with the distance of described effective label image set
Threshold value, it is determined that described 3rd label image is effective label image, adds described effective label image to
In set, and invalid tag amount of images is set to 0.
The label image generating means of the present embodiment may be used for performing embodiment of the method shown in Fig. 1, Fig. 2
Technical scheme, it is similar with technique effect that it realizes principle, and here is omitted.
In the embodiment of above-mentioned label recognizer, it should be appreciated that this processor can be that central authorities process list
Unit's (Central Processing Unit is called for short CPU), it is also possible to be other general processors, figure
As processor, digital signal processor (Digital Signal Processor, be called for short DSP), special
Integrated circuit (Application Specific Integrated Circuit is called for short ASIC) etc..
The processor etc. that general processor can be microprocessor or this processor can also be any routine.Knot
The step of conjunction method disclosed in the embodiment of the present invention can be embodied directly in hardware processor and perform,
Or combine execution by the hardware in processor and software module to complete.
One of ordinary skill in the art will appreciate that: realize all or part of step of said method embodiment
Can be completed by the hardware that programmed instruction is relevant, aforesaid program can be stored in a computer-readable
Taking in storage medium, this program upon execution, performs to include the step of said method embodiment;And it is aforementioned
Storage medium include: various Jie that can store program code such as ROM, RAM, magnetic disc or CD
Matter.
Last it is noted that various embodiments above is only in order to illustrate technical scheme, rather than right
It limits;Although the present invention being described in detail with reference to foregoing embodiments, this area common
Skilled artisans appreciate that the technical scheme described in foregoing embodiments still can be modified by it,
Or the most some or all of technical characteristic is carried out equivalent;And these amendments or replacement, and
The essence not making appropriate technical solution departs from the scope of various embodiments of the present invention technical scheme.
Claims (10)
1. a label image generates method, it is characterised in that including:
If effectively the effective number of labels in label image set is less than the first predetermined number, then stochastic generation
First label image;
Calculate the distance of described first label image and described effective label image set;
If described first label image is more than predeterminable range threshold with the distance of described effective label image set
Value, it is determined that described first label image is effective label image, is added to by described first label image
In described effective label image set.
Method the most according to claim 1, it is characterised in that described method also includes:
If the distance of described first label image and described effective label image set be not more than described preset away from
From threshold value, then abandon described first label image, invalid tag amount of images is added one.
Method the most according to claim 2, it is characterised in that described in abandon described first label figure
As afterwards, described method also includes:
If invalid tag amount of images is less than the second predetermined number, then stochastic generation the second label image;
If described second label image is more than described predeterminable range with the distance of described effective label image set
Threshold value, it is determined that described second label image is effective label image, adds described effective label image to
In set, and invalid tag amount of images is set to 0;
If the distance of described second label image and described effective label image set be not more than described preset away from
From threshold value, then abandon described second label image, invalid tag amount of images is added one.
The most according to the method in claim 2 or 3, it is characterised in that described method also includes:
If invalid tag amount of images is more than the second predetermined number, then update described predeterminable range threshold value, more
Distance threshold after Xin is less than predeterminable range threshold value;
If the distance threshold after Geng Xining is more than the cut-off distance threshold preset, then stochastic generation the 3rd label figure
Picture;
If described 3rd label image is more than the distance after updating with the distance of described effective label image set
Threshold value, it is determined that described 3rd label image is effective label image, adds described effective label image to
In set, and invalid tag amount of images is set to 0.
The most according to the method in any one of claims 1 to 3, it is characterised in that described random life
Become the first label image, including:
The generating probability of each row coding vector in described first label image is determined according to formula (1)
P (w=wi), with according to described generating probability P (w=wi) generate described first label image mg:
Wherein, mg=(w0,w1,...,wn), wiIt is the first label image mgIn the i-th row coding vector,
wi=(b0,b1,,,bn|bj∈ 0,1}), bjPixel for the jth coding unit in described i-th row coding vector
Value;wγFor either rank coding vector, Z represents 2nPlant coded system;
Wherein, T (wi) it is wiConversion ratio, determine according to formula (2);O(wi, D) and it is wiThere iing criterion
Sign the probability occurred in image collection D, determine according to formula (3):
Wherein,For wiIn q-th coding unit, whenTime, δ (wγ,wi) equal to 0, whenTime, δ (wγ,wi) equal to 1, | D | is label image quantity effective in D.
The most according to the method in any one of claims 1 to 3, it is characterised in that described calculating institute
State the distance of the first label image and described effective label image set, including:
Described first label image m is calculated according to formula (4)gAnd between described effective label image set D
Distance L (mg, D):
Wherein, mhFor the arbitrary effective label image in D, L (mg,mh) it is mgWith mhBetween distance, root
Determine according to formula (5):
Wherein, H () represents Hamming distance, Rk(mh) represent mhRotate k × 90 degree.
The most according to the method in any one of claims 1 to 3, it is characterised in that described method is also
Including:
Generating the framed picture corresponding with described first label image, described first label image is filled in institute
State in the housing that framed picture characterizes.
8. a label image generating means, it is characterised in that including:
Processor, memorizer, described processor, described memorizer are connected by bus;
Wherein, storing batch processing code in described memorizer, described processor calls in described memorizer
The program code of storage is to perform following processing procedure:
If effectively the effective number of labels in label image set is less than the first predetermined number, then stochastic generation
First label image;
Calculate the distance of described first label image and described effective label image set;
If described first label image is more than predeterminable range threshold with the distance of described effective label image set
Value, it is determined that described first label image is effective label image, is added to by described first label image
In described effective label image set.
Label generating means the most according to claim 8, it is characterised in that described processor is also used
In:
If the distance of described first label image and described effective label image set be not more than described preset away from
From threshold value, then abandon described first label image, invalid tag amount of images is added one;
If invalid tag amount of images is less than the second predetermined number, then stochastic generation the second label image;
If described second label image is more than described predeterminable range with the distance of described effective label image set
Threshold value, it is determined that described second label image is effective label image, adds described effective label image to
In set, and invalid tag amount of images is set to 0;
If the distance of described second label image and described effective label image set be not more than described preset away from
From threshold value, then abandon described second label image, invalid tag amount of images is added one.
Label generating means the most according to claim 8 or claim 9, it is characterised in that described process
Device is additionally operable to:
If the distance of described first label image and described effective label image set be not more than described preset away from
From threshold value, then abandon described first label image, invalid tag amount of images is added one;
If invalid tag amount of images is more than the second predetermined number, then update described predeterminable range threshold value, more
Distance threshold after Xin is less than predeterminable range threshold value;
If the distance threshold after Geng Xining is more than the cut-off distance threshold preset, then stochastic generation the 3rd label figure
Picture;
If described 3rd label image is more than the distance after updating with the distance of described effective label image set
Threshold value, it is determined that described 3rd label image is effective label image, adds described effective label image to
In set, and invalid tag amount of images is set to 0.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610218723.8A CN105868816B (en) | 2016-04-08 | 2016-04-08 | Label image generation method and device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610218723.8A CN105868816B (en) | 2016-04-08 | 2016-04-08 | Label image generation method and device |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105868816A true CN105868816A (en) | 2016-08-17 |
CN105868816B CN105868816B (en) | 2019-02-01 |
Family
ID=56637320
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610218723.8A Active CN105868816B (en) | 2016-04-08 | 2016-04-08 | Label image generation method and device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105868816B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109389148A (en) * | 2018-08-28 | 2019-02-26 | 昆明理工大学 | A kind of similar determination method of image based on improvement DHash algorithm |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20080061146A1 (en) * | 2006-09-13 | 2008-03-13 | Konica Minolta Business Technologies, Inc. | Barcode image generating device, barcode image reading device and barcode image generating and reading system |
CN102638761A (en) * | 2012-04-24 | 2012-08-15 | 北京信息科技大学 | WIFI (Wireless Fidelity) positioning method and positioning system thereof |
CN103994762A (en) * | 2014-04-21 | 2014-08-20 | 刘冰冰 | Mobile robot localization method based on data matrix code |
CN104361379A (en) * | 2014-11-28 | 2015-02-18 | 天津联云网络技术有限公司 | Two-dimensional code label data interaction system with positioning function |
-
2016
- 2016-04-08 CN CN201610218723.8A patent/CN105868816B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20080061146A1 (en) * | 2006-09-13 | 2008-03-13 | Konica Minolta Business Technologies, Inc. | Barcode image generating device, barcode image reading device and barcode image generating and reading system |
CN102638761A (en) * | 2012-04-24 | 2012-08-15 | 北京信息科技大学 | WIFI (Wireless Fidelity) positioning method and positioning system thereof |
CN103994762A (en) * | 2014-04-21 | 2014-08-20 | 刘冰冰 | Mobile robot localization method based on data matrix code |
CN104361379A (en) * | 2014-11-28 | 2015-02-18 | 天津联云网络技术有限公司 | Two-dimensional code label data interaction system with positioning function |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109389148A (en) * | 2018-08-28 | 2019-02-26 | 昆明理工大学 | A kind of similar determination method of image based on improvement DHash algorithm |
CN109389148B (en) * | 2018-08-28 | 2021-11-23 | 昆明理工大学 | Image similarity judgment method based on improved DHash algorithm |
Also Published As
Publication number | Publication date |
---|---|
CN105868816B (en) | 2019-02-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US9743084B2 (en) | Image encoding and decoding method and apparatus | |
CN102147873A (en) | Method and system for generating color code as well as method and system for analyzing color code | |
US20100271371A1 (en) | Data processing device and method for converting coordinates of a curved surface | |
RU2017143389A (en) | TRACKING OBJECTS BY LEARNING WITHOUT CONTROL | |
CN104754323B (en) | The scaling method of camera optical axis testing equipment | |
CN102661811B (en) | Remote sensing earth surface temperature up-scaling method and system | |
CN107122816A (en) | A kind of self-defined view finding figure Quick Response Code and generation method | |
CN108885787A (en) | Method for training image restoration model, image restoration method, device, medium, and apparatus | |
CN104615262A (en) | Input method and input system used for virtual keyboard | |
CN112001383A (en) | Water meter code intelligent identification method based on convolutional neural network technology | |
CN105139007A (en) | Positioning method and apparatus of face feature point | |
CN105868816A (en) | Label image generation method and apparatus | |
US20160148040A1 (en) | Apparatus and method for object detection based on dominant pixel information | |
JP6636963B2 (en) | Image processing apparatus and image processing method | |
CN112465886A (en) | Model generation method, device, equipment and readable storage medium | |
CN103218820A (en) | Camera calibration error compensation method based on multi-dimensional characteristics | |
CN115082788B (en) | Air pressure center identification method and device, electronic equipment and storage medium | |
CN102376093A (en) | Calibration method of camera | |
CN103871068A (en) | High-precision calibration method based on genetic algorithm | |
CN103164690A (en) | Method and device for utilizing motion tendency to track augmented reality three-dimensional multi-mark | |
CN102903097B (en) | Method and device for image perspective correction | |
CN104809459B (en) | A kind of character array structure light code word bearing calibration | |
CN106248070A (en) | A kind of navigator started based on iris identification | |
CN112465932A (en) | Image filling method, device, equipment and storage medium | |
CN110619354A (en) | Image recognition system and method for unmanned sales counter |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |