CN104036261A - Face recognition method and system - Google Patents
Face recognition method and system Download PDFInfo
- Publication number
- CN104036261A CN104036261A CN201410306005.7A CN201410306005A CN104036261A CN 104036261 A CN104036261 A CN 104036261A CN 201410306005 A CN201410306005 A CN 201410306005A CN 104036261 A CN104036261 A CN 104036261A
- Authority
- CN
- China
- Prior art keywords
- face picture
- classification
- layer
- target face
- collected
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Landscapes
- Image Analysis (AREA)
Abstract
The invention relates to a face recognition method and system, and relates to the technical field of information, wherein the face recognition method and system mainly aim at completing face recognition through the smaller calculation amount compared with the prior art. The method includes the steps of clustering collected face images to obtain multiple first-layer classifications, continuing clustering face images in at least one ith-layer classification through the iterative way to obtain multiple (i+1)th-layer classifications, recognizing the first-layer classification to which a target face image belongs, continuing recognizing the (j+1)th-layer classification to which the target face image belongs in the jth-layer classification to which the target face image belongs through the iterative way, and recognizing a face image similar to the target face image from the jth-layer classification to which the target face image belongs through the iterative way till no (j+1)th-layer classification exists in the jth-layer classification to which the target face image belongs. Compared with an existing technical scheme, the calculation amount is quite small and face recognition efficiency is greatly improved.
Description
Technical field
The present invention relates to areas of information technology, in particular to a kind of face identification method and system.
Background technology
At present, extensive face is retrieved, the work that need to carry out is first face data to be carried out to feature extraction and quantification, and after these work, every face can obtain corresponding multidimensional characteristic data; And judge that by the characteristic that compares two facial images whether two facial images are similar.And in large-scale face data, find similar face need to calculate every image, find nearest face data, need great calculated amount.
Existing scheme is for directly magnanimity face picture in local library being carried out to cluster, and target face picture is compared to find cluster under it with each cluster successively, then in affiliated cluster to the comparison of every face picture to find similar face picture.This scheme has to a certain degree been accelerated face search procedure, but in extensive face search, this scheme still needs to carry out great calculated amount: less if cluster arranges, in each cluster, data scale can be larger, searches for and will expend a large amount of operation time in cluster; If it is too much to divide cluster, under finding, cluster just need to, compared with intensive, all cannot meet the requirement of real-time of search engine.
Summary of the invention
In view of the above problems, the present invention has been proposed to a kind of overcome the problems referred to above or the face identification method addressing the above problem at least in part and system are provided.
According to one aspect of the present invention, a kind of face identification method is provided, it comprises: the face picture of having collected is carried out to cluster and obtain multiple the 1st layer of classification, and continue that by iterative manner the face picture in the classification of at least one i layer is carried out to cluster and obtain multiple i+1 layers classification, i carries out integer value backward from 1; Identify the 1st layer of affiliated classification of target face picture, and continue to identify the j+1 layer classification under described target face picture in the j layer classification under described target face picture by iterative manner, j sequentially carries out integer value backward from 1; By described iterative manner until while there is not the classification of j+1 layer in the classification of j layer under described target face picture, the j layer classification under described target face picture, identify the similar face picture of described target face picture.
Alternatively, aforesaid face identification method, wherein, describedly carries out by the face picture of having collected the step that cluster obtains multiple the 1st layer of classification and comprises: according to the feature of the described face picture of having collected, the proper vector of the face picture of having collected described in generation; Multiple initial center point are set, and according to the distance distance of the proper vector of the described face picture of having collected and each described initial center point, the described face picture of having collected is divided into multiple the 1st layer of classification, and according to the proper vector of the face picture of each the 1st layer of classification, calculate the vectorial central point of described each the 1st layer of classification.
Alternatively, aforesaid face identification method, wherein, describedly carries out by the face picture of having collected the step that cluster obtains multiple the 1st layer of classification and also comprises: calculate the variance between initial center point and the vectorial central point of described each the 1st layer of classification; As described in square extent exceed predetermined threshold value, reset initial center point, and again the described face picture of having collected be divided into multiple the 1st layer of classification, and recalculate the vectorial central point of described each the 1st layer of classification.
Alternatively, aforesaid face identification method, wherein, described in identify the 1st layer of classification under target face picture step comprise: according to the feature of described target face picture, generate the proper vector of described target face picture; Select the 1st layer of classification of distance minimum between vectorial central point and the proper vector of described target face picture, as the 1st layer of classification under described target face picture.
Alternatively, aforesaid face identification method, wherein, the step of the described similar face picture that identifies described target face picture comprises: from the face picture of described j layer classification, select at least one face picture of distance minimum between proper vector and the proper vector of described target face picture, as the described similar face picture of described target face picture.
According to another aspect of the present invention, a kind of face identification system is also provided, it comprises: sort module, for being carried out to cluster, the face picture of having collected obtains multiple the 1st layer of classification, and continue that by iterative manner the face picture in the classification of at least one i layer is carried out to cluster and obtain multiple i+1 layers classification, i sequentially carries out integer value backward from 1; Classification iteration identification module, for identifying the 1st layer of classification under target face picture, and continue to identify the j+1 layer classification under described target face picture in the j layer classification under described target face picture by iterative manner, j carries out integer value backward from 1; Similar face picture recognition module, while there is not the classification of j+1 layer, the j layer classification under described target face picture, identifies the similar face picture of described target face picture for the j layer classification under described target face picture.
Alternatively, aforesaid face identification system, wherein, also comprises: first eigenvector generation module, and for according to the feature of the described face picture of having collected, the proper vector of the face picture of having collected described in generation; Described sort module arranges multiple initial center point, and according to the distance distance of the proper vector of the described face picture of having collected and each described initial center point, the described face picture of having collected is divided into multiple the 1st layer of classification, and according to the proper vector of the face picture of each the 1st layer of classification, calculate the vectorial central point of described each the 1st layer of classification.
Alternatively, aforesaid face identification system, wherein, also comprises: variance computing module, calculates the variance between initial center point and the vectorial central point of described each the 1st layer of classification; As described in square extent exceed predetermined threshold value, described sort module resets initial center point, and again the described face picture of having collected is divided into multiple the 1st layer of classification, and recalculates the vectorial central point of described each the 1st layer of classification.
Alternatively, aforesaid face identification system, wherein, also comprises: Second Characteristic vector generation module, for according to the feature of described target face picture, generates the proper vector of described target face picture; Described sort module is selected the 1st layer of classification of distance minimum between vectorial central point and the proper vector of described target face picture, as the 1st layer of classification under described target face picture.
Alternatively, aforesaid face identification system, wherein, described similar face picture recognition module is from the face picture of described j layer classification, select at least one face picture of distance minimum between proper vector and the proper vector of described target face picture, as the described similar face picture of described target face picture.
According to face identification method of the present invention and system, by cluster last layer cluster result is divided to cluster again, be sandwich construction by iterative manner by the face picture cluster of having collected, and successively find the affiliated classification of target face picture by iterative manner, until finally find the similar face picture of target face picture; With respect to existing technical scheme, the calculated amount in technical scheme of the present invention is very little, has greatly improved recognition of face efficiency.
Above-mentioned explanation is only the general introduction of technical solution of the present invention, in order to better understand technological means of the present invention, and can be implemented according to the content of instructions, and for above and other objects of the present invention, feature and advantage can be become apparent, below especially exemplified by the specific embodiment of the present invention.
Brief description of the drawings
By reading below detailed description of the preferred embodiment, various other advantage and benefits will become cheer and bright for those of ordinary skill in the art.Accompanying drawing is only for the object of preferred implementation is shown, and do not think limitation of the present invention.And in whole accompanying drawing, represent identical parts by identical reference symbol.In the accompanying drawings:
Fig. 1 shows the process flow diagram of face identification method according to an embodiment of the invention;
Fig. 2 shows the work schematic diagram of face identification method according to an embodiment of the invention;
Fig. 3 shows the process flow diagram of face identification method according to an embodiment of the invention;
Fig. 4 shows the process flow diagram of face identification method according to an embodiment of the invention;
Fig. 5 shows the process flow diagram of face identification method according to an embodiment of the invention;
Fig. 6 shows the block diagram of face identification system according to an embodiment of the invention;
Fig. 7 shows the block diagram of face identification method according to an embodiment of the invention;
Fig. 8 shows the block diagram of face identification method according to an embodiment of the invention.
Embodiment
Exemplary embodiment of the present disclosure is described below with reference to accompanying drawings in more detail.Although shown exemplary embodiment of the present disclosure in accompanying drawing, but should be appreciated that and can realize the disclosure and the embodiment that should do not set forth limits here with various forms.On the contrary, it is in order more thoroughly to understand the disclosure that these embodiment are provided, and can be by the those skilled in the art that conveys to complete the scope of the present disclosure.
As shown in the figure, one embodiment of the present of invention provide a kind of face identification method, and it comprises:
Step 110, carries out cluster by the face picture of having collected and obtains multiple the 1st layer of classification, and continues that by iterative manner the face picture in the classification of at least one i layer is carried out to cluster and obtain multiple i+1 layers classification, and i carries out integer value backward from 1.In the present embodiment, as shown in Figure 2, for example, wherein C1 classification comprises C11 to the multistratum classification structure of formation ... multiple classification such as C1m, comprise again the classification such as CN1, CN2 in C11 classification.
Step 120, identifies the 1st layer of affiliated classification of target face picture, and continues to identify the j+1 layer classification under target face picture in the j layer classification under target face picture by iterative manner, and j sequentially carries out integer value backward from 1.
Step 130, by iterative manner until while there is not the classification of j+1 layer in the classification of j layer under target face picture, the j layer classification under target face picture, identify the similar face picture of target face picture.
In the technical scheme of the present embodiment, be sandwich construction by iterative manner by the face picture cluster of having collected by cluster last layer cluster result being divided again to cluster, and successively find the affiliated classification of target face picture by iterative manner, until finally find the similar face picture of target face picture; With respect to existing technical scheme, the calculated amount in technical scheme of the present invention is very little, has greatly improved recognition of face efficiency.
As shown in Figure 3, an alternative embodiment of the invention provides a kind of face identification method, and wherein, step 110 comprises:
Step 111, according to the feature of the face picture of having collected, generates the proper vector of the face picture of having collected.The extraction of the face picture feature of the present embodiment based on to having collected, has collected the feature of face picture and can extract in advance and be stored in feature database.
Step 112, multiple initial center point are set, and according to the distance distance of the proper vector of the face picture of having collected and each initial center point, the face picture of having collected is divided into multiple the 1st layer of classification, and according to the proper vector of the face picture of each the 1st layer of classification, calculate the vectorial central point of each the 1st layer of classification.
, according to the technical scheme of the present embodiment, all, according to the distance between the face picture feature vector of having collected and initial center point, be assigned in the most contiguous classification, afterwards compute vector central point; So repeatedly can form fast multi-level clustering structure.
An alternative embodiment of the invention provides a kind of face identification method, and wherein, step 110 also comprises:
113, calculate the variance between initial center point and the vectorial central point of each the 1st layer of classification.The extraction of the present embodiment based on to target face picture feature, the feature of target face picture can extract real-time.
114, as square extent exceedes predetermined threshold value, reset initial center point, and again the face picture of having collected is divided into multiple the 1st layer of classification, and recalculate the vectorial central point of each the 1st layer of classification.
In the technical scheme of the present embodiment, if variance <0.000001 (example, can get other values), it in presentation class, is all the face picture that feature is close, close otherwise there is feature gap significant discomfort in presentation class the face picture that is placed in same classification, so need to reclassify.Now, the workflow of the face identification method of the present embodiment can as shown in Figure 4, in figure, calculate and specify the step of the number of plies to refer to, is the cluster structures of specifying the number of plies by the face picture cluster of having collected.
As shown in Figure 5, an alternative embodiment of the invention provides a kind of face identification method, and wherein, step 120 comprises:
Step 121, according to the feature of target face picture, generates the proper vector of target face picture.
Step 122, selects the 1st layer of classification of distance minimum between vectorial central point and the proper vector of target face picture, as the 1st layer of classification under target face picture.
, according to the technical scheme of the present embodiment, in each layer of structure, all the vectorial central point of multiple lower floors classification in the proper vector of target face picture and the classification of affiliated upper strata is compared, can find fast the affiliated minimum classification of target face picture.
An alternative embodiment of the invention provides a kind of face identification method, and wherein, step 130 comprises:
From the face picture of j layer classification, select at least one face picture of distance minimum between proper vector and the proper vector of target face picture, as the similar face picture of target face picture.
In conjunction with above embodiment, suppose that collecting face picture is 1,000 ten thousand, while needing therein the similar face picture of searched targets face picture:
If 1 uses direct manner of comparison, the proper vector of target face picture and the proper vector of collecting face picture need to be compared 1,000 ten thousand times.
If 2 use traditional cluster mode, 1,000 ten thousand data are divided into 10000 clusters, need the vectorial central point of target face picture and cluster to compare 10000 times, during each cluster is average, there are 1000 data, each classification inside need to be compared 1000 times, overall relatively 10000+k × 1000 time, as k gets 10 number of comparisons are: 10000+10*1000=20000 time.While finding the similar face picture of target face picture by existing nearest neighbor algorithm in cluster, k represents to choose k neighbour's central point, and 10 is common value.
3, use the technical scheme of the present embodiment, if be divided into 2 layers, 100 clusters of ground floor, the second layer is 200 clusters, in the average each cluster of the second layer, has 500 data, about 100+m × 200+n × 500 time of number of comparisons, as get m=3, and n=10, number of comparisons is: 100+3 × 200+10 × 500=11100 time, with 1,2 comparisons, can significantly reduce number of comparisons.In like manner, m represents that ground floor chooses m neighbour's central point, and n represents that the second layer chooses n neighbour's central point, and 3,10 is common value.
As shown in Figure 6, another embodiment of the present invention also provides a kind of face identification system, and it comprises:
Sort module 610, for being carried out to cluster, the face picture of having collected obtains multiple the 1st layer of classification, and continue that by iterative manner the face picture in the classification of at least one i layer is carried out to cluster and obtain multiple i+1 layers classification, i sequentially carries out integer value backward from 1.In the present embodiment, as shown in Figure 2, for example, wherein C1 classification comprises C11 to the multistratum classification structure of formation ... multiple classification such as C1m, comprise again the classification such as CN1, CN2 in C11 classification.
Classification iteration identification module 620, for identifying the 1st layer of classification under target face picture, and continue to identify the j+1 layer classification under target face picture in the j layer classification under target face picture by iterative manner, j carries out integer value backward from 1.
Similar face picture recognition module 630, while there is not the classification of j+1 layer, the j layer classification under target face picture, identifies the similar face picture of target face picture for the j layer classification under target face picture.
In the technical scheme of the present embodiment, be sandwich construction by iterative manner by the face picture cluster of having collected by cluster last layer cluster result being divided again to cluster, and successively find the affiliated classification of target face picture by iterative manner, until finally find the similar face picture of target face picture; With respect to existing technical scheme, the calculated amount in technical scheme of the present invention is very little, has greatly improved recognition of face efficiency.
As shown in Figure 7, an alternative embodiment of the invention provides a kind of face identification system, wherein, also comprises:
First eigenvector generation module 640, for according to the feature of the face picture of having collected, generates the proper vector of the face picture of having collected.The extraction of the face picture feature of the present embodiment based on to having collected, has collected the feature of face picture and can extract in advance and be stored in feature database.
Sort module 610 arranges multiple initial center point, and according to the distance distance of the proper vector of the face picture of having collected and each initial center point, the face picture of having collected is divided into multiple the 1st layer of classification, and according to the proper vector of the face picture of each the 1st layer of classification, calculate the vectorial central point of each the 1st layer of classification.
, according to the technical scheme of the present embodiment, all, according to the distance between the face picture feature vector of having collected and initial center point, be assigned in the most contiguous classification, afterwards compute vector central point; So repeatedly can form fast multi-level clustering structure.
An alternative embodiment of the invention provides a kind of face identification system, wherein, also comprises:
Variance computing module 650, calculates the variance between initial center point and the vectorial central point of each the 1st layer of classification.The extraction of the present embodiment based on to target face picture feature, the feature of target face picture can extract real-time.
As square extent exceedes predetermined threshold value, sort module 610 resets initial center point, and again the face picture of having collected is divided into multiple the 1st layer of classification, and recalculates the vectorial central point of each the 1st layer of classification.
In the technical scheme of the present embodiment, if variance <0.000001 (example, can get other values), it in presentation class, is all the face picture that feature is close, close otherwise there is feature gap significant discomfort in presentation class the face picture that is placed in same classification, so need to reclassify.Now, the workflow of the face identification method of the present embodiment can as shown in Figure 4, in figure, calculate and specify the step of the number of plies to refer to, is the cluster structures of specifying the number of plies by the face picture cluster of having collected.
As shown in Figure 8, an alternative embodiment of the invention provides a kind of face identification system, wherein, also comprises:
Second Characteristic vector generation module 660, for according to the feature of target face picture, generates the proper vector of target face picture.
Sort module 610 is selected the 1st layer of classification of distance minimum between vectorial central point and the proper vector of target face picture, as the 1st layer of classification under target face picture.
, according to the technical scheme of the present embodiment, in each layer of structure, all the vectorial central point of multiple lower floors classification in the proper vector of target face picture and the classification of affiliated upper strata is compared, can find fast the affiliated minimum classification of target face picture.
An alternative embodiment of the invention provides a kind of face identification system, wherein, similar face picture recognition module 630 is from the face picture of j layer classification, select at least one face picture of distance minimum between proper vector and the proper vector of target face picture, as the similar face picture of target face picture.
In conjunction with above embodiment, suppose that collecting face picture is 1,000 ten thousand, while needing therein the similar face picture of searched targets face picture:
If 1 uses direct manner of comparison, the proper vector of target face picture and the proper vector of collecting face picture need to be compared 1,000 ten thousand times.
If 2 use traditional cluster mode, 1,000 ten thousand data are divided into 10000 clusters, need the vectorial central point of target face picture and cluster to compare 10000 times, during each cluster is average, there are 1000 data, each classification inside need to be compared 1000 times, overall relatively 10000+k × 1000 time, as k gets 10 number of comparisons are: 10000+10*1000=20000 time.While finding the similar face picture of target face picture by existing nearest neighbor algorithm in cluster, k represents to choose k neighbour's central point, and 10 is common value.
3, use the technical scheme of the present embodiment, if be divided into 2 layers, 100 clusters of ground floor, the second layer is 200 clusters, in the average each cluster of the second layer, has 500 data, about 100+m × 200+n × 500 time of number of comparisons, as get m=3, and n=10, number of comparisons is: 100+3 × 200+10 × 500=11100 time, with 1,2 comparisons, can significantly reduce number of comparisons.In like manner, m represents that ground floor chooses m neighbour's central point, and n represents that the second layer chooses n neighbour's central point, and 3,10 is common value.
The algorithm providing at this is intrinsic not relevant to any certain computer, virtual system or miscellaneous equipment with demonstration.Various general-purpose systems also can with based on using together with this teaching.According to description above, it is apparent constructing the desired structure of this type systematic.In addition, the present invention is not also for any certain programmed language.It should be understood that and can utilize various programming languages to realize content of the present invention described here, and the description of above language-specific being done is in order to disclose preferred forms of the present invention.
In the instructions that provided herein, a large amount of details are described.But, can understand, embodiments of the invention can be put into practice in the situation that there is no these details.In some instances, be not shown specifically known method, structure and technology, so that not fuzzy understanding of this description.
Similarly, be to be understood that, in order to simplify the disclosure and to help to understand one or more in each inventive aspect, in the above in the description of exemplary embodiment of the present invention, each feature of the present invention is grouped together into single embodiment, figure or sometimes in its description.But, the method for the disclosure should be construed to the following intention of reflection: the present invention for required protection requires than the more feature of feature of clearly recording in each claim.Or rather, as reflected in claims below, inventive aspect is to be less than all features of disclosed single embodiment above.Therefore, claims of following embodiment are incorporated to this embodiment thus clearly, and wherein each claim itself is as independent embodiment of the present invention.
Those skilled in the art are appreciated that and can the module in the equipment in embodiment are adaptively changed and they are arranged in one or more equipment different from this embodiment.Module in embodiment or unit or assembly can be combined into a module or unit or assembly, and can put them in addition multiple submodules or subelement or sub-component.At least some in such feature and/or process or unit are mutually repelling, and can adopt any combination to combine all processes or the unit of disclosed all features in this instructions (comprising claim, summary and the accompanying drawing followed) and disclosed any method like this or equipment.Unless clearly statement in addition, in this instructions (comprising claim, summary and the accompanying drawing followed) disclosed each feature can be by providing identical, be equal to or the alternative features of similar object replaces.
In addition, those skilled in the art can understand, although embodiment more described herein comprise some feature instead of further feature included in other embodiment, the combination of the feature of different embodiment means within scope of the present invention and forms different embodiment.For example, in the following claims, the one of any of embodiment required for protection can be used with array mode arbitrarily.
All parts embodiment of the present invention can realize with hardware, or realizes with the software module of moving on one or more processor, or realizes with their combination.It will be understood by those of skill in the art that and can use in practice microprocessor or digital signal processor (DSP) to realize the some or all functions according to the some or all parts in the face identification system of the embodiment of the present invention.The present invention can also be embodied as part or all equipment or the device program (for example, computer program and computer program) for carrying out method as described herein.Realizing program of the present invention and can be stored on computer-readable medium like this, or can there is the form of one or more signal.Such signal can be downloaded and obtain from internet website, or provides on carrier signal, or provides with any other form.
It should be noted above-described embodiment the present invention will be described instead of limit the invention, and those skilled in the art can design alternative embodiment in the case of not departing from the scope of claims.In the claims, any reference symbol between bracket should be configured to limitations on claims.Word " comprises " not to be got rid of existence and is not listed as element or step in the claims.Being positioned at word " " before element or " one " does not get rid of and has multiple such elements.The present invention can be by means of including the hardware of some different elements and realizing by means of the computing machine of suitably programming.In the unit claim of having enumerated some devices, several in these devices can be to carry out imbody by same hardware branch.The use of word first, second and C grade does not represent any order.Can be title by these word explanations.
Claims (10)
1. a face identification method, it comprises:
The face picture of having collected is carried out to cluster and obtain multiple the 1st layer of classification, and continue that by iterative manner the face picture in the classification of at least one i layer is carried out to cluster and obtain multiple i+1 layers classification, i carries out integer value backward from 1;
Identify the 1st layer of affiliated classification of target face picture, and continue to identify the j+1 layer classification under described target face picture in the j layer classification under described target face picture by iterative manner, j sequentially carries out integer value backward from 1;
By described iterative manner until while there is not the classification of j+1 layer in the classification of j layer under described target face picture, the j layer classification under described target face picture, identify the similar face picture of described target face picture.
2. face identification method according to claim 1, wherein, describedly carries out by the face picture of having collected the step that cluster obtains multiple the 1st layer of classification and comprises:
According to the feature of the described face picture of having collected, the proper vector of the face picture of having collected described in generation;
Multiple initial center point are set, and according to the distance distance of the proper vector of the described face picture of having collected and each described initial center point, the described face picture of having collected is divided into multiple the 1st layer of classification, and according to the proper vector of the face picture of each the 1st layer of classification, calculate the vectorial central point of described each the 1st layer of classification.
3. according to the face identification method described in claim 1-2 any one, wherein, describedly the face picture of having collected carried out to the step that cluster obtains multiple the 1st layer of classification also comprise:
Calculate the variance between initial center point and the vectorial central point of described each the 1st layer of classification;
As described in square extent exceed predetermined threshold value, reset initial center point, and again the described face picture of having collected be divided into multiple the 1st layer of classification, and recalculate the vectorial central point of described each the 1st layer of classification.
4. according to the face identification method described in claim 1-3 any one, wherein, described in identify the 1st layer of classification under target face picture step comprise:
According to the feature of described target face picture, generate the proper vector of described target face picture;
Select the 1st layer of classification of distance minimum between vectorial central point and the proper vector of described target face picture, as the 1st layer of classification under described target face picture.
5. according to the face identification method described in claim 1-4 any one, wherein, described in identify the similar face picture of described target face picture step comprise:
From the face picture of described j layer classification, select at least one face picture of distance minimum between proper vector and the proper vector of described target face picture, as the described similar face picture of described target face picture.
6. a face identification system, it comprises:
Sort module, for being carried out to cluster, the face picture of having collected obtains multiple the 1st layer of classification, and continue that by iterative manner the face picture in the classification of at least one i layer is carried out to cluster and obtain multiple i+1 layers classification, i sequentially carries out integer value backward from 1;
Classification iteration identification module, for identifying the 1st layer of classification under target face picture, and continue to identify the j+1 layer classification under described target face picture in the j layer classification under described target face picture by iterative manner, j carries out integer value backward from 1;
Similar face picture recognition module, while there is not the classification of j+1 layer, the j layer classification under described target face picture, identifies the similar face picture of described target face picture for the j layer classification under described target face picture.
7. face identification system according to claim 6, wherein, also comprises:
First eigenvector generation module, for according to the feature of the described face picture of having collected, the proper vector of the face picture of having collected described in generation;
Described sort module arranges multiple initial center point, and according to the distance distance of the proper vector of the described face picture of having collected and each described initial center point, the described face picture of having collected is divided into multiple the 1st layer of classification, and according to the proper vector of the face picture of each the 1st layer of classification, calculate the vectorial central point of described each the 1st layer of classification.
8. according to the face identification system described in claim 6-7 any one, wherein, also comprise:
Variance computing module, calculates the variance between initial center point and the vectorial central point of described each the 1st layer of classification;
As described in square extent exceed predetermined threshold value, described sort module resets initial center point, and again the described face picture of having collected is divided into multiple the 1st layer of classification, and recalculates the vectorial central point of described each the 1st layer of classification.
9. according to the face identification system described in claim 6-8 any one, wherein, also comprise:
Second Characteristic vector generation module, for according to the feature of described target face picture, generates the proper vector of described target face picture;
Described sort module is selected the 1st layer of classification of distance minimum between vectorial central point and the proper vector of described target face picture, as the 1st layer of classification under described target face picture.
10. according to the face identification system described in claim 6-9 any one, wherein,
Described similar face picture recognition module is from the face picture of described j layer classification, select at least one face picture of distance minimum between proper vector and the proper vector of described target face picture, as the described similar face picture of described target face picture.
Priority Applications (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410306005.7A CN104036261B (en) | 2014-06-30 | 2014-06-30 | Face identification method and system |
| US15/322,350 US20170132457A1 (en) | 2014-06-27 | 2015-06-26 | Human face similarity recognition method and system |
| PCT/CN2015/082550 WO2015197029A1 (en) | 2014-06-27 | 2015-06-26 | Human face similarity recognition method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410306005.7A CN104036261B (en) | 2014-06-30 | 2014-06-30 | Face identification method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104036261A true CN104036261A (en) | 2014-09-10 |
| CN104036261B CN104036261B (en) | 2017-03-29 |
Family
ID=51467027
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410306005.7A Expired - Fee Related CN104036261B (en) | 2014-06-27 | 2014-06-30 | Face identification method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104036261B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104850600A (en) * | 2015-04-29 | 2015-08-19 | 百度在线网络技术（北京）有限公司 | Method and device for searching images containing faces |
| WO2015197029A1 (en) * | 2014-06-27 | 2015-12-30 | 北京奇虎科技有限公司 | Human face similarity recognition method and system |
| CN108664920A (en) * | 2018-05-10 | 2018-10-16 | 深圳市深网视界科技有限公司 | A kind of cascade face cluster method and apparatus extensive in real time |
| CN110069989A (en) * | 2019-03-15 | 2019-07-30 | 上海拍拍贷金融信息服务有限公司 | Face image processing process and device, computer readable storage medium |
| CN110348274A (en) * | 2018-04-08 | 2019-10-18 | 杭州海康威视数字技术股份有限公司 | A kind of face identification method, device and equipment |
| CN110874419A (en) * | 2019-11-19 | 2020-03-10 | 山东浪潮人工智能研究院有限公司 | Quick retrieval technology for face database |
| CN113553461A (en) * | 2020-04-26 | 2021-10-26 | 北京搜狗科技发展有限公司 | Picture clustering method and related device |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060204058A1 (en) * | 2004-12-07 | 2006-09-14 | Kim Do-Hyung | User recognition system and method thereof |
| CN102402508A (en) * | 2010-09-07 | 2012-04-04 | 华东师范大学 | Similar image search device and search method thereof |
| CN103617217A (en) * | 2013-11-20 | 2014-03-05 | 中国科学院信息工程研究所 | Hierarchical index based image retrieval method and system |
-
2014
- 2014-06-30 CN CN201410306005.7A patent/CN104036261B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060204058A1 (en) * | 2004-12-07 | 2006-09-14 | Kim Do-Hyung | User recognition system and method thereof |
| CN102402508A (en) * | 2010-09-07 | 2012-04-04 | 华东师范大学 | Similar image search device and search method thereof |
| CN103617217A (en) * | 2013-11-20 | 2014-03-05 | 中国科学院信息工程研究所 | Hierarchical index based image retrieval method and system |
Non-Patent Citations (1)
| Title |
|---|
| 刘燕 等: "《基于聚类的大型人脸检索系统》", 《计算机工程》 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2015197029A1 (en) * | 2014-06-27 | 2015-12-30 | 北京奇虎科技有限公司 | Human face similarity recognition method and system |
| CN104850600A (en) * | 2015-04-29 | 2015-08-19 | 百度在线网络技术（北京）有限公司 | Method and device for searching images containing faces |
| CN110348274A (en) * | 2018-04-08 | 2019-10-18 | 杭州海康威视数字技术股份有限公司 | A kind of face identification method, device and equipment |
| CN110348274B (en) * | 2018-04-08 | 2022-03-04 | 杭州海康威视数字技术股份有限公司 | Face recognition method, device and equipment |
| CN108664920A (en) * | 2018-05-10 | 2018-10-16 | 深圳市深网视界科技有限公司 | A kind of cascade face cluster method and apparatus extensive in real time |
| CN108664920B (en) * | 2018-05-10 | 2022-12-20 | 东方网力科技股份有限公司 | Real-time large-scale cascading face clustering method and device |
| CN110069989A (en) * | 2019-03-15 | 2019-07-30 | 上海拍拍贷金融信息服务有限公司 | Face image processing process and device, computer readable storage medium |
| CN110069989B (en) * | 2019-03-15 | 2021-07-30 | 上海拍拍贷金融信息服务有限公司 | Face image processing method and device and computer readable storage medium |
| CN110874419A (en) * | 2019-11-19 | 2020-03-10 | 山东浪潮人工智能研究院有限公司 | Quick retrieval technology for face database |
| CN113553461A (en) * | 2020-04-26 | 2021-10-26 | 北京搜狗科技发展有限公司 | Picture clustering method and related device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104036261B (en) | 2017-03-29 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104036261A (en) | Face recognition method and system | |
| US11481656B2 (en) | Method and apparatus for evaluating a matching degree of multi-domain information based on artificial intelligence, device and medium | |
| CN110660062B (en) | Point cloud instance segmentation method and system based on PointNet | |
| Sattler et al. | Hyperpoints and fine vocabularies for large-scale location recognition | |
| CN106570453B (en) | Method, device and system for pedestrian detection | |
| Dewi et al. | Robust detection method for improving small traffic sign recognition based on spatial pyramid pooling | |
| US20140010407A1 (en) | Image-based localization | |
| US10754887B1 (en) | Systems and methods for multimedia image clustering | |
| Lai et al. | Recognizing complex events in videos by learning key static-dynamic evidences | |
| Lin et al. | Local and global encoder network for semantic segmentation of Airborne laser scanning point clouds | |
| Hu et al. | Massively parallel KD-tree construction and nearest neighbor search algorithms | |
| CN104679818A (en) | Video keyframe extracting method and video keyframe extracting system | |
| Song et al. | Combining Models from Multiple Sources for RGB-D Scene Recognition. | |
| US20150294194A1 (en) | Method of classifying a multimodal object | |
| WO2016095068A1 (en) | Pedestrian detection apparatus and method | |
| CN104036259A (en) | Face similarity recognition method and system | |
| Zhao et al. | Regional gating neural networks for multi-label image classification. | |
| CN105260746A (en) | Expandable multilayer integrated multi-mark learning system | |
| Setyono et al. | Betawi traditional food image detection using ResNet and DenseNet | |
| CN103678513A (en) | Interactive search generation method and system | |
| Polewski et al. | Free Shape Context descriptors optimized with genetic algorithm for the detection of dead tree trunks in ALS point clouds | |
| CN115098717A (en) | Three-dimensional model retrieval method and device, electronic equipment and storage medium | |
| KR102039244B1 (en) | Data clustering method using firefly algorithm and the system thereof | |
| Prasomphan | Toward Fine-grained Image Retrieval with Adaptive Deep Learning for Cultural Heritage Image. | |
| CN111597367A (en) | Three-dimensional model retrieval method based on view and Hash algorithm |
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 | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170329 Termination date: 20210630 |