CN106295475A - A kind of based on sum of ranks than the registered fingerprint replacement method of method - Google Patents

A kind of based on sum of ranks than the registered fingerprint replacement method of method Download PDF

Info

Publication number
CN106295475A
CN106295475A CN201510289626.3A CN201510289626A CN106295475A CN 106295475 A CN106295475 A CN 106295475A CN 201510289626 A CN201510289626 A CN 201510289626A CN 106295475 A CN106295475 A CN 106295475A
Authority
CN
China
Prior art keywords
fingerprint
row
ranks
sum
registered fingerprint
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN201510289626.3A
Other languages
Chinese (zh)
Inventor
张威
吴敏
阿勇
赵彤
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
BEIJING EASTERN GOLDEN FINGER TECHNOLOGY Co Ltd
Original Assignee
BEIJING EASTERN GOLDEN FINGER TECHNOLOGY Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by BEIJING EASTERN GOLDEN FINGER TECHNOLOGY Co Ltd filed Critical BEIJING EASTERN GOLDEN FINGER TECHNOLOGY Co Ltd
Priority to CN201510289626.3A priority Critical patent/CN106295475A/en
Publication of CN106295475A publication Critical patent/CN106295475A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/12—Fingerprints or palmprints
    • G06V40/1347—Preprocessing; Feature extraction
    • G06V40/1353—Extracting features related to minutiae or pores
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
    • G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
    • G06F16/583—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content
    • G06F16/5838—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using metadata automatically derived from the content using colour
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • G06V40/12—Fingerprints or palmprints
    • G06V40/1365—Matching; Classification
    • G06V40/1371—Matching features related to minutiae or pores

Abstract

A kind of based on sum of ranks than the registered fingerprint replacement method of method, 1) same fingerprint judges, compares newly gathering fingerprint with original registered fingerprint, if same piece of fingerprint, execution following steps.Replace if it is not, then do not do fingerprint；2) build raw data list, by be evaluated fingerprint (1,2 ..., n) and evaluation index (1,2 ..., m) be arranged in n row m row initial data sum of ranks compare matrix；The described fingerprint being evaluated is for newly gathering fingerprint, original registered fingerprint；3) compile order, compose power respectively to newly gathering different evaluation index row's orders of fingerprint, original registered fingerprint, compose power；4) calculate respectively newly gather fingerprint, the sum of ranks of original registered fingerprint compares RSR；5) if newly gathering the comprehensive evaluation result RSR of fingerprint1It is better than the RSR of original registered fingerprint2Comprehensive evaluation result, then replace the image of original registered fingerprint as existing registered fingerprint with the new image gathering fingerprint.The present invention is with a kind of fingerprint Comprehensive Assessment based on sum of ranks ratio method, after registered fingerprint is changed, it is possible to significantly improve fingerprint comparison precision.

Description

A kind of based on sum of ranks than the registered fingerprint replacement method of method
Technical field
The invention belongs to living things feature recognition field, be a kind of based on sum of ranks than the registered fingerprint replacement method of method.
Background technology
Fingerprint identification technology is divided into two processes: (1) fingerprint register process.By reading fingerprint image, find in fingerprint image Finger-print region, the feature that fingerprint mutually can be made a distinction in the region that takes the fingerprint.These characteristics are saved in data In storehouse, and as the representative of this piece of fingerprint, thus complete fingerprint register process；(2) feature comparison process.Newly adopt having extracted After the feature of the fingerprint to be compared of collection, registered fingerprints whole in the appointment registered fingerprint in data base or data base are compared, Confirm that the new fingerprint gathered process whether in data base is characterized comparison process.Fingerprint identification technology relates generally to fingerprint image Collection, fingerprint image process, feature extraction, preservation data, eigenvalue comparison with the process such as mate.Along with computer skill Art and information processing are continuous progressive with identification technology, and fingerprint identification technology has obtained swift and violent development, become the most ripe, The most acceptable a kind of biometrics identification technology, be can prove public security, network management, bank, social security, employee, sea Close the technology that many fields such as identity authentication, electronic access are applied extensively and profoundly, there is important theoretical significance and market should By value.The most how the design registered fingerprint replacement method of science has and important practical significance.
Traditional fingerprint evaluation methodology based on minutia, i.e. extracts the minutiae feature (destination node of crestal line or friendship in fingerprint Crunode) characterize fingerprint image as feature, and identify the registered fingerprint image of replacement by extracting these features.As clearly The CN1588425A of Hua Da, it utilizes the same finger fingerprint image of multi collect to carry out the fusion of minutia.The defect of this technology It is, if the minutia of the fingerprint image gathered for each time is not accurately aimed at, fingerprint comparison precision degradation to be caused.
Sum of ranks is a kind of statistical analysis technique integrating classic parameter estimation and nonparametric statistics in modern age each advantage than method, it It is applicable to the overall merit of row × list data.Wherein, sum of ranks ratio (Rank-sum ratio, RSR) refers to row in table (or row) The meansigma methods of rank, is a nonparametric metering, has the feature of 0～1 interval continuous variable.Its basic thought is at a n In row (n evaluation object) m row (m evaluation index) matrix, changed by order, it is thus achieved that nondimensional statistic RSR, with RSR The quality of evaluation object is ranked up or grading sorting by value.The overall merit contrasted before and after sum of ranks is used for fingerprint image than method, Multiple evaluation objectives can be carried out choosing comprehensively, easy and simple to handle.
Summary of the invention
The problem existed for above-mentioned prior art, the purpose of the present invention is to propose to a kind of registered fingerprint based on sum of ranks ratio method and replaces Change method, after the advantage of the method is to utilize this registered fingerprint Exchange rings to replace registered fingerprint, it is possible to significantly improve fingerprint ratio To precision.
The present invention be achieved in that a kind of based on sum of ranks than the registered fingerprint replacement method of method, it is characterised in that include following step Rapid:
1) same fingerprint judges
Compare newly gathering fingerprint with original registered fingerprint, if after same piece of fingerprint, perform following steps and determine and be No needs carries out fingerprint replacing；Replace if it is not, then do not do fingerprint；
2) raw data list is built
Consider the indices that fingerprint is evaluated, select corresponding fingerprint evaluation index so that it is can synthetically reflect fingerprint Quality, by be evaluated fingerprint (1,2 ..., n) and evaluation index (1,2 ..., m) be arranged in n row m row initial data sum of ranks ratio Matrix, n and m is the integer more than 0；The described fingerprint being evaluated is for newly gathering fingerprint, original registered fingerprint, i.e. n=2；
Described initial data sum of ranks has 2 pieces of fingerprints than the row representative in matrix and participates in overall merit, wherein sets the 1st row and represents new Gathering fingerprint, set the 2nd row and represent original registered fingerprint, the row in matrix represent m evaluation index, arbitrary unit in matrix Element is RijRepresent row's order result of the jth index of i-th piece of fingerprint；
3) compile order, compose power
Respectively to newly gather fingerprint, original registered fingerprint different evaluation index row order, different evaluation indexes is carried out compose power: Estimator can improve the weight coefficient of corresponding index for the attention degree of a certain index along with the change of desired value；
4) calculate respectively newly gather fingerprint, the sum of ranks of original registered fingerprint compares RSR
RSR i = Σ j = 1 m W j R ij n × m
In formula, n is the fingerprint being evaluated, and m is evaluation index, WjFor weight coefficient；
RijIt it is the rank of the i-th row jth column element；
5) if newly gathering the comprehensive evaluation result RSR of fingerprint1It is better than the RSR of original registered fingerprint2Comprehensive evaluation result, then use The new image gathering fingerprint replaces the image of original registered fingerprint as existing registered fingerprint.
Further, select picture quality and fingerprint characteristic number as evaluation index；Described evaluation index m=2；
Described step 2) structure raw data list be
A, calculates respectively and newly gathers fingerprint, the mass fraction of original registered fingerprint
Quality evaluation for fingerprint image method is utilized to calculate the mass fraction of the fingerprint image newly gathering fingerprint, original registered fingerprint respectively, Record the mass fraction of each finger fingerprint image；
B, calculates respectively and newly gathers fingerprint, the Characteristic Number of original registered fingerprint
Utilize Finger print characteristic abstract algorithm to extract respectively and newly gather fingerprint, the fingerprint characteristic of original registered fingerprint, record each finger fingerprint The number of feature；
Building initial data sum of ranks than matrix is R 2 × 2 = R 11 R 12 R 21 R 22 ; Row in matrix represents and newly gathers fingerprint, original registration refers to Stricture of vagina has 2 pieces of fingerprints and participates in overall merit, wherein sets the 1st row representative and newly gathers fingerprint, and the 2nd row represents original registered fingerprint, Row in matrix represent 2 evaluation indexes, and the 1st is classified as and singly refers to fingerprint quality, and the 2nd is classified as fingerprint characteristic number；In matrix Either element is Rij, represent row's order result of the jth index of i-th piece of fingerprint.
Further, weight W of described picture quality1=1, weight W of described fingerprint characteristic number2=1.
Further, weight W of described picture quality1=0.8, weight W of described fingerprint characteristic number2=1.
Further, the comprehensive evaluation result RSR of described new collection fingerprint1It is better than the RSR of original registered fingerprint2Comprehensive evaluation result, For the new comprehensive evaluation result RSR gathering fingerprint1Value more than the RSR of original registered fingerprint2Value.
The present invention has the most useful technique effect,
1) present invention propose a kind of based on sum of ranks than the integrated evaluating method of the fingerprint of method, change weight factor is tied mutually with sum of ranks ratio method Close, and use it for fingerprint superior and inferior evaluating；
2) picture quality and fingerprint characteristic number are combined by the present invention, overcome the tradition deficiency based on single evaluation target；
3) present invention is by newly gathering fingerprint and the integrated evaluating method of original registered fingerprint utilization RSRw method, calculates respective RSR value, evaluates new collection fingerprint and the quality of original registered fingerprint by this comprehensive evaluation index RSR value, for decision is No former registered fingerprint is changed, fingerprint comparison precision can be improved significantly, definitely.
Accompanying drawing explanation
Fig. 1 is fingerprint image comprehensive evaluation index schematic diagram of the present invention.
The flow chart of Fig. 2 implementation of the present invention.
Detailed description of the invention
Embodiment 1
As shown in Figure 1-2, the method comprises the following steps:
Step S01: compare newly gathering fingerprint with original registered fingerprint, it is determined whether be same piece of fingerprint.If it is same After piece fingerprint, perform following steps and determine the need for carrying out fingerprint replacing.Replace if it is not, then fingerprint cannot be done.
Assuming that have registered one piece of fingerprint in fingerprint database, again gather this piece of fingerprint, and carry out fingerprint comparison, confirm that this refers to That piece of fingerprint registered in stricture of vagina and fingerprint database is as same piece of fingerprint.Now need by performing each step following, it is judged that Whether to replace existing registered fingerprint.
Step S02: utilize Quality evaluation for fingerprint image method to calculate the fingerprint image newly gathering fingerprint, original registered fingerprint respectively Mass fraction, records the mass fraction of each finger fingerprint image.
As it is shown in figure 1, the quality discrimination of the image of fingerprint can be from finger-print region size, the dry and wet situation of finger-print region, streakline Orientation consistency angularly differentiates, can provide a synthesis result according to prior art, obtain a mass fraction.
Certain by prior art singly refers to Quality evaluation for fingerprint image method, and giving makes new advances gathers fingerprint, original registered fingerprint image Mass fraction: 86,80.
Step S03: utilize Finger print characteristic abstract algorithm to extract respectively and newly gather fingerprint, the fingerprint characteristic of original registered fingerprint, record Respectively refer to the number of fingerprint characteristic.
By existing single piece of fingerprint image characteristics extracting method, obtain the Characteristic Number of fingerprint, it is however generally that, it is referred to China The clear and definite fingerprint minutiae feature of industry standards of public safety GA426-2008 that the Ministry of Public Security issues and number expression method.
For given, by certain single piece of fingerprint image characteristics extracting method of prior art, reference standard GA426-2008 marks Quasi-acquirement newly gathers fingerprint, the feature point number of original registered fingerprint is: 100,75.
Step S04: compile order, compose power
Evaluation index is divided into high excellent index and low excellent index, and high excellent index compiles order from big to small, and maximum is compiled with 1；Low excellent index Compiling order from small to large, minima is compiled with 1；Identical person averages, and order is the least, and index is the most excellent；Fingerprint quality mark and fingerprint Characteristic Number is high excellent order, is the forward index of fingerprint evaluation, and index is the biggest, and rank is the highest.Respectively to newly gathering fingerprint, former The fingerprint quality of registered fingerprint, fingerprint characteristic number two class index is had to compile order (R).
Carry out different evaluation indexes composing power: estimator changes along with the change of desired value for the attention degree of a certain index, Desired value is the most bad, and the overall merit of fingerprint image is affected the biggest by this index, therefore improves the corresponding weight coefficient of this index. Embodiment 1 print quality, the weight coefficient of fingerprint characteristic are respectively 1.
The data utilizing step S01-S03 can form following data form, and calculates the order (for order in parantheses) of indices:
Table 1 fingerprint compiles order
Fingerprint Fingerprint image quality Fingerprint characteristic number
Newly gather fingerprint 86(2) 100(2)
Original registered fingerprint 80(1) 75(1)
Definition sum of ranks than matrix is R 2 × 2 = R 11 R 12 R 21 R 22 , Row in matrix represents and newly gathers fingerprint, original registered fingerprint has 2 Piece fingerprint participates in overall merit (can set the 1st row to represent and newly gather fingerprint, the 2nd row represents original registered fingerprint), in matrix Row represent and have 2 evaluation indexes, the 1st is classified as and singly refers to fingerprint quality, and the 2nd is classified as fingerprint characteristic number.Arbitrary unit in matrix Element is Rij, represent row's order result of the jth index of i-th piece of fingerprint.
Being drawn by table 1, sum of ranks compares matrix R 2 × 2 = 2 2 1 1
Step S05: calculate respectively newly gather fingerprint, original registered fingerprint sum of ranks than RSR, using this result as newly gather fingerprint, Original registered fingerprint comprehensive evaluation result.
According to formula(wherein i, j=1,2, n, m=2, W1=1, W2=1) weighting of fingerprint image is calculated Sum of ranks compares: RSR 1 = 1 4 ( 2 + 2 ) = 1 , RSR 2 = 1 4 ( 1 + 1 ) = 0.5 .
Step S06: if the comprehensive evaluation result newly gathering fingerprint is better than the comprehensive evaluation result of original registered fingerprint, then carry out Replace the image of original registered fingerprint as existing registered fingerprint with the new image gathering fingerprint.
Because RSR1＞ RSR2, so replacing original registered fingerprint with the new fingerprint gathered.
The present invention proposes a kind of registered fingerprint replacement method based on sum of ranks ratio method, select by based on sum of ranks than method to finger Stricture of vagina carries out Comprehensive Assessment, can significantly improve fingerprint comparison precision after registered fingerprint is changed.

Claims (5)

1. one kind based on sum of ranks than the registered fingerprint replacement method of method, it is characterised in that comprise the following steps:
1) same fingerprint judges
Compare newly gathering fingerprint with original registered fingerprint, if after same piece of fingerprint, perform following steps and determine and be No needs carries out fingerprint replacing；Replace if it is not, then do not do fingerprint；
2) raw data list is built
Consider the indices that fingerprint is evaluated, select corresponding fingerprint evaluation index so that it is can synthetically reflect fingerprint Quality, by be evaluated fingerprint (1,2 ..., n) and evaluation index (1,2 ..., m) be arranged in n row m row initial data sum of ranks ratio Matrix, n and m is the integer more than 0；The described fingerprint being evaluated is for newly gathering fingerprint, original registered fingerprint, i.e. n=2；
Described initial data sum of ranks has 2 pieces of fingerprints than the row representative in matrix and participates in overall merit, wherein sets the 1st row and represents new Gathering fingerprint, set the 2nd row and represent original registered fingerprint, the row in matrix represent m evaluation index, arbitrary unit in matrix Element is RijRepresent row's order result of the jth index of i-th piece of fingerprint；
3) compile order, compose power
Respectively to newly gather fingerprint, original registered fingerprint different evaluation index row order, different evaluation indexes is carried out compose power: Estimator can improve the weight coefficient of corresponding index for the attention degree of a certain index along with the change of desired value；
4) calculate respectively newly gather fingerprint, the sum of ranks of original registered fingerprint compares RSR
RSR i = Σ j = 1 m W j R ij n × m
In formula, n is the fingerprint being evaluated, and m is evaluation index, WjFor weight coefficient；
RijIt it is the rank of the i-th row jth column element；
5) if newly gathering the comprehensive evaluation result RSR of fingerprint1It is better than the RSR of original registered fingerprint2Comprehensive evaluation result, then use The new image gathering fingerprint replaces the image of original registered fingerprint as existing registered fingerprint.
2. as claimed in claim 1 based on sum of ranks than the registered fingerprint replacement method of method, it is characterised in that select fingerprint image Quality and fingerprint characteristic number are as evaluation index；Described evaluation index m=2；
Described step 2) structure raw data list be
A, calculates respectively and newly gathers fingerprint, the mass fraction of original registered fingerprint
Quality evaluation for fingerprint image method is utilized to calculate the mass fraction of the fingerprint image newly gathering fingerprint, original registered fingerprint respectively, Record the mass fraction of each finger fingerprint image；
B, calculates respectively and newly gathers fingerprint, the Characteristic Number of original registered fingerprint
Utilize Finger print characteristic abstract algorithm to extract respectively and newly gather fingerprint, the fingerprint characteristic of original registered fingerprint, record each finger fingerprint The number of feature；
Building initial data sum of ranks than matrix is R 2 × 2 = R 11 R 12 R 21 R 22 ; Row representative in matrix has 2 pieces of fingerprints and participates in comprehensively commenting Valency, wherein sets the 1st row representative and newly gathers fingerprint, set the 2nd row and represent original registered fingerprint, and the row in matrix represent 2 Individual evaluation index, the 1st is classified as and singly refers to fingerprint quality, and the 2nd is classified as fingerprint characteristic number；In matrix, either element is RijRepresent Row's order result of the jth index of i piece of fingerprint.
3. as claimed in claim 2 based on sum of ranks than the registered fingerprint replacement method of method, it is characterised in that described picture quality Weight W1=1, weight W of described fingerprint characteristic number2=1.
4. as claimed in claim 2 based on sum of ranks than the registered fingerprint replacement method of method, it is characterised in that described picture quality Weight W1=0.8, weight W of described fingerprint characteristic number2=1.
5. as claimed in claim 2 based on sum of ranks than the registered fingerprint replacement method of method, it is characterised in that described new collection refers to The comprehensive evaluation result RSR of stricture of vagina1It is better than the RSR of original registered fingerprint2Comprehensive evaluation result, for the new overall merit knot gathering fingerprint Really RSR1Value more than the comprehensive evaluation result RSR of original registered fingerprint2Value.
CN201510289626.3A 2015-05-29 2015-05-29 A kind of based on sum of ranks than the registered fingerprint replacement method of method Pending CN106295475A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201510289626.3A CN106295475A (en) 2015-05-29 2015-05-29 A kind of based on sum of ranks than the registered fingerprint replacement method of method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201510289626.3A CN106295475A (en) 2015-05-29 2015-05-29 A kind of based on sum of ranks than the registered fingerprint replacement method of method

Publications (1)

Publication Number Publication Date
CN106295475A true CN106295475A (en) 2017-01-04

Family

ID=57655990

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201510289626.3A Pending CN106295475A (en) 2015-05-29 2015-05-29 A kind of based on sum of ranks than the registered fingerprint replacement method of method

Country Status (1)

Country Link
CN (1) CN106295475A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107169466A (en) * 2017-05-25 2017-09-15 北京东方金指科技有限公司 A kind of palmprint image quality overall evaluation method based on sum of ranks ratio method

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070297653A1 (en) * 2006-06-22 2007-12-27 Rudolf Maarten Bolle Fingerprint representation using localized texture features
CN101231691A (en) * 2007-12-24 2008-07-30 浙江金指码科技有限公司 Learning type intellectual fingerprint identification comparison method adjusted based on fingerprint data quantity
CN101499130A (en) * 2008-01-30 2009-08-05 深圳市普罗巴克科技股份有限公司 Fingerprint recognition method and fingerprint recognition system
CN102708360A (en) * 2012-05-09 2012-10-03 深圳市亚略特生物识别科技有限公司 Method for generating and automatically updating fingerprint template
CN105389566A (en) * 2015-11-13 2016-03-09 广东欧珀移动通信有限公司 Fingerprint identification method, fingerprint template updating method, fingerprint template updating device and mobile terminal
CN106326886A (en) * 2016-11-07 2017-01-11 重庆工商大学 Finger-vein image quality evaluation method and system based on convolutional neural network

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070297653A1 (en) * 2006-06-22 2007-12-27 Rudolf Maarten Bolle Fingerprint representation using localized texture features
CN101231691A (en) * 2007-12-24 2008-07-30 浙江金指码科技有限公司 Learning type intellectual fingerprint identification comparison method adjusted based on fingerprint data quantity
CN101499130A (en) * 2008-01-30 2009-08-05 深圳市普罗巴克科技股份有限公司 Fingerprint recognition method and fingerprint recognition system
CN102708360A (en) * 2012-05-09 2012-10-03 深圳市亚略特生物识别科技有限公司 Method for generating and automatically updating fingerprint template
CN105389566A (en) * 2015-11-13 2016-03-09 广东欧珀移动通信有限公司 Fingerprint identification method, fingerprint template updating method, fingerprint template updating device and mobile terminal
CN106326886A (en) * 2016-11-07 2017-01-11 重庆工商大学 Finger-vein image quality evaluation method and system based on convolutional neural network

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
彭宪等: ""正交试验法优选野菊花的高速分散均质提取工艺"", 《广东药学院学报》 *
曾祥绪: ""网络在线指纹考勤系统可用性措施"", 《计算机应用》 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107169466A (en) * 2017-05-25 2017-09-15 北京东方金指科技有限公司 A kind of palmprint image quality overall evaluation method based on sum of ranks ratio method
CN107169466B (en) * 2017-05-25 2020-03-31 北京东方金指科技有限公司 Palm print image quality comprehensive evaluation method based on rank-sum ratio method

Similar Documents

Publication Publication Date Title
CN104462184B (en) A kind of large-scale data abnormality recognition method based on two-way sampling combination
CN106845717B (en) Energy efficiency evaluation method based on multi-model fusion strategy
Bai et al. Learning backtrackless aligned-spatial graph convolutional networks for graph classification
WO2019233189A1 (en) Method for detecting sensor network abnormal data
Jiang et al. Distance measures for image segmentation evaluation
CN104050247B (en) The method for realizing massive video quick-searching
CN103702416B (en) Semi-supervised learning indoor positioning method based on support vector machine
CN102930533B (en) A kind of semi-supervised Hyperspectral imaging dimension reduction method based on improving K-mean cluster
CN106022509A (en) Power distribution network space load prediction method taking region and load property dual differences into consideration
CN101178703B (en) Failure diagnosis chart clustering method based on network dividing
CN108805213B (en) Power load curve double-layer spectral clustering method considering wavelet entropy dimensionality reduction
CN106934410A (en) The sorting technique and system of data
CN109597757B (en) Method for measuring similarity between software networks based on multidimensional time series entropy
CN103336771A (en) Data similarity detection method based on sliding window
CN102298663A (en) Method for automatically identifying syndrome type in traditional Chinese medical science
CN110533116A (en) Based on the adaptive set of Euclidean distance at unbalanced data classification method
CN109255029A (en) A method of automatic Bug report distribution is enhanced using weighted optimization training set
CN105574265B (en) Entire assembly model quantitative description towards model index
CN103902798B (en) Data preprocessing method
CN106612511A (en) Wireless network throughput evaluation method and device based on support vector machine
Bai et al. Learning aligned-spatial graph convolutional networks for graph classification
Banumathi et al. A novel approach for upgrading Indian education by using data mining techniques
CN102208027B (en) Method for evaluating land utilization spatial pattern based on clearance degree dimension
CN110321350A (en) A kind of existence authentication method and system verified based on data reparation and Active Learning
CN106126523A (en) A kind of counterfeit money Crime Information analyzes system and the method for analysis

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
WD01 Invention patent application deemed withdrawn after publication
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20170104