CN104143090B - A kind of automobile door opening method based on recognition of face - Google Patents

A kind of automobile door opening method based on recognition of face Download PDF

Info

Publication number
CN104143090B
CN104143090B CN201410370100.3A CN201410370100A CN104143090B CN 104143090 B CN104143090 B CN 104143090B CN 201410370100 A CN201410370100 A CN 201410370100A CN 104143090 B CN104143090 B CN 104143090B
Authority
CN
China
Prior art keywords
mrow
msubsup
face
sphere
sparse
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.)
Active
Application number
CN201410370100.3A
Other languages
Chinese (zh)
Other versions
CN104143090A (en
Inventor
潘正祥
冯庆祥
蔡正富
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.)
Airmate Electrical Shenzhen Co Ltd
Shenzhen Graduate School Harbin Institute of Technology
Original Assignee
Airmate Electrical Shenzhen Co Ltd
Shenzhen Graduate School Harbin Institute of Technology
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 Airmate Electrical Shenzhen Co Ltd, Shenzhen Graduate School Harbin Institute of Technology filed Critical Airmate Electrical Shenzhen Co Ltd
Priority to CN201410370100.3A priority Critical patent/CN104143090B/en
Publication of CN104143090A publication Critical patent/CN104143090A/en
Application granted granted Critical
Publication of CN104143090B publication Critical patent/CN104143090B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Abstract

The present invention relates to automotive field, more particularly to a kind of automobile door opening method based on recognition of face.The present invention provides a kind of automobile door opening method based on recognition of face, comprises the following steps：1st, image or video flowing containing face are gathered using mobile phone camera, face information is then passed to face processing and identifier, face processing and identifier pre-process to face；2nd, face is classified using sparse feature centre of sphere grader；3rd, facial image is matched and identified according to classification results.The beneficial effects of the invention are as follows having used sparse feature centre of sphere grader to carry out differentiation certification to face information, car owner only needs to realize automobile door opening, the automobile and other people can not open the door face close to mobile phone camera and by certification can.

Description

A kind of automobile door opening method based on recognition of face
Technical field
The present invention relates to automotive field, more particularly to a kind of automobile door opening method based on recognition of face.
Background technology
Automobile door opening system is a subsystem among current onboard system, its main function be used for it is antitheft and conveniently Car owner.
Automobile door opening mode main at present mainly has two kinds：One kind is automobile key, and another kind is Intelligent key.This two Kind of door-opening mode, often can be because of leaving behind key, or loses key and allow driver to have a headache very much.
Grader occupies critically important position in onboard system PRS.Nearest neighbor classifier (NN) and recently Center grader (NM) is two more famous graders.As the extension of nearest neighbor classifier, nearest characteristic face grader (NFP) it is suggested.It can be represented because belonging to the sample of a specific target class by the linear subspaces of this class, based on upper The thought in face, linear regression grader (LRC) are suggested.Linear regression grader can be regarded as prolonging for nearest center grader Stretch.After linear regression grader is suggested, some improved methods are suggested, including based on kernel-LRC, LDA-LRC, PCA-LRC etc..
Different from linear regression grader, the grader (SRC) based on rarefaction representation employs the model of all classes to surveying Sample is originally classified.After the proposition of SRC graders, some other improved methods is suggested, as two benches test sample is dilute Dredge and represent (TPTSSR), the grader (CRC) represented based on cooperation, regular error robust coding grader (RRC) and loosen cooperation Presentation class device (RCR).
In fact, the L2 coefficients that linear regression grader can be considered as on space-like represent.Coefficient represents grader It is by carrying out solving L1-norm minimization problems to the model of all classes, then according to test sample and each class subspace The distance between predicted vector is classified to test sample.But the distance between test sample and predicted vector may not It is a method weighed well.So grader and nearest characteristic face grader are represented based on coefficient, it is proposed that sparse Feature centre of sphere grader (SFSC) grader.
The content of the invention
For defect present in prior art or deficiency, the technical problems to be solved by the invention are：A kind of base is provided In the automobile door opening method of recognition of face, this method can be realized in the situation (or other certification articles) without key The enabling of automobile, driver is thoroughly broken away from will be with the worry of many articles, and the extension for also improving nearest feature line is not smart True and computation complexity problem.
The technical scheme that the present invention takes comprises the following steps to provide a kind of automobile door opening method based on recognition of face
Step 1：Image or video flowing containing face are gathered using mobile phone camera, then face information transmission is given people Face processing and identifier, face processing and identifier pre-process to face；
Step 2：Face is classified using sparse feature centre of sphere grader；
Step 21；Assuming that each class at least three samples, propose a new distance metric, the feature centre of sphere is made to measure, Feature centre of sphere measurement refers to test sample and the feature centre of sphereBetween Euclidean distance, it can be calculated as
In formulaIt is tetrahedronInscribed sphere the centre of sphere, the feature centre of sphereIt can be calculated as
In formulaWithRefer to triangle respectivelyWithArea；
Step 22：The coefficient weights of each class are calculated, coefficient represents that the model X that grader constructs all classes owns for accumulation The q- dimensional vectors of class, it can be expressed as
Then we are standardized the model X of all classes, so as to produce a unit vector；Assuming that test vector is X, we can solve L1-norm minimization problems and are：
G=argming||g||1Subject to Xg=x (4)
After we obtain sparse coefficient vector using formula (3) and (4), we will calculate the sparse system of each class Number and ScFor
Because Sc1 is less than, and it may be negative, so we are ScPositive number is changed to, so as to be used as c classes Sparse weights, it can be calculated as
wc=1-sc (6)
Using the sparse weights of each class, sparse features centre of sphere grader will calculate test sample x to three training sampleWithSparse weighting the feature centre of sphere measurement, it can be calculated as
After we obtain the feature centre of sphere measurement of all sparse weightings, these distances are carried out ascending order and arranged by we Sequence, and each distance is corresponding with a class label, and last sparse features centre of sphere grader, which will assign to test sample, to be possessed In the class of minimum range, it can be represented as
Step 3：Facial image is matched and identified according to classification results.
The beneficial effects of the invention are as follows：The mobile phone carried with using car owner, it becomes possible to open the door, make driver thoroughly break away from To be saved with the worry of many articles, because everyone face information is different, car owner only needs face to lean on Nearly mobile phone camera and automobile door opening is realized by certification can, and other people can not open the door automobile, add automobile Security.
Brief description of the drawings
Fig. 1 is automobile door opening method characteristic centre of sphere measurement schematic diagram of the present invention based on recognition of face.
Embodiment
The present invention is further described for explanation and embodiment below in conjunction with the accompanying drawings.
Automobile door opening system of the invention based on recognition of face is the mobile phone camera by car owner, to facial image or is regarded Frequency is acquired.After perfect person's face image or video information is gathered, this new automobile door opening system can use the sparse of proposition Feature centre of sphere grader carry out to face information certification, to judge whether to be car owner in enabling automobile.
Driver is acquired using a mobile phone camera to facial image or video, then the face figure being collected Picture or delivery of video are to face processing and identifier.When face processing and identifier receive the facial image that collector piece transmits or After video, it can carry out some pretreatments to facial image or vision signal, followed by use it is proposed that sparse feature ball Heart grader carries out classification certification to it, after the system of in-car is by certification, will open car door.
As shown in figure 1, the present invention proposes a kind of automobile door opening method based on recognition of face, two steps are included： One step, sparse feature centre of sphere grader assume each class at least 3 samples, and this is similar with nearest characteristic face.Sparse Feature centre of sphere grader is measured instead of using characteristic face, proposes a new distance metric, makes the feature centre of sphere measure.The feature centre of sphere Measurement refers to test sample and the feature centre of sphereBetween Euclidean distance, it can be calculated as
WhereinIt is tetrahedronInscribed sphere the centre of sphere feature centre ofs sphereIt can be calculated as
WhereinWithRefer to triangle respectivelyWithArea；
After the first the end of the step, we start second step to calculate the coefficient weights of each class.Coefficient represents grader structure The model X for making all classes is the q- dimensional vectors for accumulating all classes, and it can be expressed as
Then we are standardized the model X of all classes, so as to produce a unit vector.Assuming that test vector is X, we can solve L1-norm minimization problems and are:
G=argming||g||1Subject to Xg=x (4)
For formula (4), it is understood that the purpose of rarefaction representation grader is to produce a preferable parameter vectorSo the class for possessing greatest coefficient sum represents most like class.Situation based on more than, We will be weighted value using sparse coefficient to feature centre of sphere measurement.When we are sparse using formula (3) and (4) After number vector obtains, we will calculate the sparse coefficient and s of each classcFor
Because Sc1 is less than, and it may be negative, so we are ScPositive number is changed to, so as to be used as c classes Sparse weights, it can be calculated as
wc=1-sc (6)
Using the sparse weights of each class, sparse features centre of sphere grader will calculate test sample x to three training sampleWithSparse weighting the feature centre of sphere measurement, it can be calculated as
After we obtain the feature centre of sphere measurement of all sparse weightings, these distances are carried out ascending order and arranged by we Sequence, and each distance is corresponding with a class label, and last sparse features centre of sphere grader, which will assign to test sample, to be possessed In the class of minimum range, it can be represented as
3rd step, facial image is matched and identified according to classification results.
Above content is to combine specific preferred embodiment further description made for the present invention, it is impossible to is assert The specific implementation of the present invention is confined to these explanations.For general technical staff of the technical field of the invention, On the premise of not departing from present inventive concept, some simple deduction or replace can also be made, should all be considered as belonging to the present invention's Protection domain.

Claims (1)

  1. A kind of 1. automobile door opening method based on recognition of face, it is characterised in that：Comprise the following steps
    Step 1：Image or video flowing containing face are gathered using mobile phone camera, then face information is passed at face Reason and identifier, face processing and identifier pre-process to face；
    Step 2：Face is classified using sparse feature centre of sphere grader；
    Step 21；Assuming that each class at least three samples, propose a distance metric, the feature centre of sphere is made to measure, the feature centre of sphere Measurement refers to test sample and the feature centre of sphereBetween Euclidean distance, be calculated as
    <mrow> <msub> <mi>d</mi> <mrow> <mi>F</mi> <mi>S</mi> <mi>C</mi> </mrow> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>,</mo> <mover> <mrow> <msubsup> <mi>x</mi> <mi>i</mi> <mi>c</mi> </msubsup> <msubsup> <mi>x</mi> <mi>j</mi> <mi>c</mi> </msubsup> <msubsup> <mi>x</mi> <mi>k</mi> <mi>c</mi> </msubsup> </mrow> <mo>&amp;OverBar;</mo> </mover> <mo>)</mo> </mrow> <mo>=</mo> <mo>|</mo> <mo>|</mo> <mi>x</mi> <mo>-</mo> <msubsup> <mi>x</mi> <mi>o</mi> <mrow> <mi>i</mi> <mi>j</mi> <mi>k</mi> <mo>,</mo> <mi>c</mi> </mrow> </msubsup> <mo>|</mo> <mo>|</mo> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>1</mn> <mo>)</mo> </mrow> </mrow>
    In formulaIt is tetrahedronInscribed sphere the centre of sphere, the feature centre of sphereIt can be calculated as
    <mrow> <msubsup> <mi>x</mi> <mi>o</mi> <mrow> <mi>i</mi> <mi>j</mi> <mi>k</mi> <mo>,</mo> <mi>c</mi> </mrow> </msubsup> <mo>=</mo> <mfrac> <mrow> <msubsup> <mi>s</mi> <mrow> <mi>i</mi> <mi>j</mi> <mi>k</mi> </mrow> <mi>c</mi> </msubsup> <mo>&amp;times;</mo> <mi>x</mi> <mo>+</mo> <msubsup> <mi>s</mi> <mrow> <mi>y</mi> <mi>j</mi> <mi>k</mi> </mrow> <mi>c</mi> </msubsup> <mo>&amp;times;</mo> <msubsup> <mi>x</mi> <mi>i</mi> <mi>c</mi> </msubsup> <mo>+</mo> <msubsup> <mi>s</mi> <mrow> <mi>y</mi> <mi>i</mi> <mi>k</mi> </mrow> <mi>c</mi> </msubsup> <mo>&amp;times;</mo> <msubsup> <mi>x</mi> <mi>j</mi> <mi>c</mi> </msubsup> <mo>+</mo> <msubsup> <mi>s</mi> <mrow> <mi>y</mi> <mi>i</mi> <mi>j</mi> </mrow> <mi>c</mi> </msubsup> <mo>&amp;times;</mo> <msubsup> <mi>x</mi> <mi>k</mi> <mi>c</mi> </msubsup> </mrow> <mrow> <msubsup> <mi>s</mi> <mrow> <mi>i</mi> <mi>j</mi> <mi>k</mi> </mrow> <mi>c</mi> </msubsup> <mo>+</mo> <msubsup> <mi>s</mi> <mrow> <mi>y</mi> <mi>j</mi> <mi>k</mi> </mrow> <mi>c</mi> </msubsup> <mo>+</mo> <msubsup> <mi>s</mi> <mrow> <mi>y</mi> <mi>i</mi> <mi>k</mi> </mrow> <mi>c</mi> </msubsup> <mo>+</mo> <msubsup> <mi>s</mi> <mrow> <mi>y</mi> <mi>i</mi> <mi>j</mi> </mrow> <mi>c</mi> </msubsup> </mrow> </mfrac> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>2</mn> <mo>)</mo> </mrow> </mrow>
    In formulaWithRefer to triangle respectivelyWithArea, WithThree training samples are represented respectively；
    Step 22：The coefficient weights of each class are calculated, coefficient represents that the model X that grader constructs all classes is to accumulate all classes Q- dimensional vectors, the coefficient weights of each class are expressed as
    <mrow> <mi>X</mi> <mo>=</mo> <mo>&amp;lsqb;</mo> <mo>...</mo> <mo>,</mo> <msubsup> <mi>x</mi> <mi>i</mi> <mi>c</mi> </msubsup> <mo>,</mo> <mo>...</mo> <mo>&amp;rsqb;</mo> <mo>,</mo> <mi>c</mi> <mo>=</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mo>...</mo> <mo>,</mo> <mi>M</mi> <mo>,</mo> <mi>i</mi> <mo>=</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mo>...</mo> <mo>,</mo> <mi>N</mi> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>3</mn> <mo>)</mo> </mrow> </mrow>
    Then we are standardized the model X of all classes, so as to produce a unit vector；Assuming that test vector is x, solution Certainly L1-norm minimization problems are：
    G=arg ming||g||1Subject to Xg=x (4)
    After sparse coefficient vector is obtained using formula (3) and (4), the sparse coefficient and S of each class are calculatedcFor
    <mrow> <msub> <mi>s</mi> <mi>c</mi> </msub> <mo>=</mo> <msubsup> <mi>g</mi> <mn>1</mn> <mi>c</mi> </msubsup> <mo>+</mo> <msubsup> <mi>g</mi> <mn>2</mn> <mi>c</mi> </msubsup> <mo>+</mo> <mo>...</mo> <mo>+</mo> <msubsup> <mi>g</mi> <mi>N</mi> <mi>c</mi> </msubsup> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>5</mn> <mo>)</mo> </mrow> </mrow>
    Sc1 is less than, and it may be negative, ScPositive number is changed to, so as to the sparse weights as c classes, ScAs The sparse weights of c classes are calculated as
    wc=1-sc (6)
    Using the sparse weights of each class, sparse features centre of sphere grader will calculate test sample x to three training sampleWithSparse weighting the feature centre of sphere measurement, test sample x to three training sampleWithIt is sparse plus The feature centre of sphere measurement of power is calculated as
    <mrow> <msub> <mi>d</mi> <mrow> <mi>S</mi> <mi>F</mi> <mi>S</mi> <mi>C</mi> </mrow> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>,</mo> <mover> <mrow> <msubsup> <mi>x</mi> <mi>i</mi> <mi>c</mi> </msubsup> <msubsup> <mi>x</mi> <mi>j</mi> <mi>c</mi> </msubsup> <msubsup> <mi>x</mi> <mi>k</mi> <mi>c</mi> </msubsup> </mrow> <mo>&amp;OverBar;</mo> </mover> <mo>)</mo> </mrow> <mo>=</mo> <mo>|</mo> <mo>|</mo> <mi>x</mi> <mo>-</mo> <msubsup> <mi>x</mi> <mi>o</mi> <mrow> <mi>i</mi> <mi>j</mi> <mi>k</mi> <mo>,</mo> <mi>c</mi> </mrow> </msubsup> <mo>|</mo> <mo>|</mo> <mo>&amp;times;</mo> <msub> <mi>w</mi> <mi>c</mi> </msub> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>7</mn> <mo>)</mo> </mrow> </mrow>
    After the feature centre of sphere measurement for obtaining all sparse weightings, these distances carry out ascending order and are ranked up, and each distance A class label is corresponding with, last sparse features centre of sphere grader will be assigned to test sample in the class for possessing minimum range, It is represented as
    <mrow> <munder> <mrow> <mi>m</mi> <mi>i</mi> <mi>n</mi> </mrow> <mrow> <mi>c</mi> <mo>*</mo> </mrow> </munder> <msub> <mi>d</mi> <mrow> <mi>S</mi> <mi>F</mi> <mi>S</mi> <mi>C</mi> </mrow> </msub> <mrow> <mo>(</mo> <mi>x</mi> <mo>,</mo> <mover> <mrow> <msubsup> <mi>x</mi> <mi>i</mi> <mi>c</mi> </msubsup> <msubsup> <mi>x</mi> <mi>j</mi> <mi>c</mi> </msubsup> <msubsup> <mi>x</mi> <mi>k</mi> <mi>c</mi> </msubsup> </mrow> <mo>&amp;OverBar;</mo> </mover> <mo>)</mo> </mrow> <mo>,</mo> <mi>c</mi> <mo>=</mo> <mn>1</mn> <mo>,</mo> <mn>2</mn> <mo>,</mo> <mo>...</mo> <mo>,</mo> <mi>M</mi> <mo>-</mo> <mo>-</mo> <mo>-</mo> <mrow> <mo>(</mo> <mn>8</mn> <mo>)</mo> </mrow> </mrow>
    Step 3：Facial image is matched and identified according to classification results.
CN201410370100.3A 2014-07-30 2014-07-30 A kind of automobile door opening method based on recognition of face Active CN104143090B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201410370100.3A CN104143090B (en) 2014-07-30 2014-07-30 A kind of automobile door opening method based on recognition of face

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201410370100.3A CN104143090B (en) 2014-07-30 2014-07-30 A kind of automobile door opening method based on recognition of face

Publications (2)

Publication Number Publication Date
CN104143090A CN104143090A (en) 2014-11-12
CN104143090B true CN104143090B (en) 2018-02-02

Family

ID=51852260

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201410370100.3A Active CN104143090B (en) 2014-07-30 2014-07-30 A kind of automobile door opening method based on recognition of face

Country Status (1)

Country Link
CN (1) CN104143090B (en)

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10915769B2 (en) 2018-06-04 2021-02-09 Shanghai Sensetime Intelligent Technology Co., Ltd Driving management methods and systems, vehicle-mounted intelligent systems, electronic devices, and medium
US10970571B2 (en) 2018-06-04 2021-04-06 Shanghai Sensetime Intelligent Technology Co., Ltd. Vehicle control method and system, vehicle-mounted intelligent system, electronic device, and medium
CN108819900A (en) * 2018-06-04 2018-11-16 上海商汤智能科技有限公司 Control method for vehicle and system, vehicle intelligent system, electronic equipment, medium

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7426292B2 (en) * 2003-08-07 2008-09-16 Mitsubishi Electric Research Laboratories, Inc. Method for determining optimal viewpoints for 3D face modeling and face recognition
CN101564328A (en) * 2009-05-07 2009-10-28 杭州电子科技大学 Laptop artificial limb multi-movement-mode identifying method based on support vector data description
CN101833654A (en) * 2010-04-02 2010-09-15 清华大学 Sparse representation face identification method based on constrained sampling

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7426292B2 (en) * 2003-08-07 2008-09-16 Mitsubishi Electric Research Laboratories, Inc. Method for determining optimal viewpoints for 3D face modeling and face recognition
CN101564328A (en) * 2009-05-07 2009-10-28 杭州电子科技大学 Laptop artificial limb multi-movement-mode identifying method based on support vector data description
CN101833654A (en) * 2010-04-02 2010-09-15 清华大学 Sparse representation face identification method based on constrained sampling

Non-Patent Citations (5)

* Cited by examiner, † Cited by third party
Title
基于人脸识别的门禁系统设计;高双喜等;《河北省科学院学报》;20120331;第29卷(第1期);第28-30页 *
基于稀疏表达的人脸识别算法研究;王静;《中国优秀硕士学位论文全文数据库 信息科技辑》;20120715(第07期);I138-2431 *
基于超球球心间距多类支持向量机的滚动轴承故障分类;康守强等;《中国电机工程学报》;20140515;第34卷(第14期);第2319-2325页 *
复杂目标稳定稀疏表达在人脸识别中的应用;赵程程;《中国优秀硕士学位论文全文数据库 信息科技辑》;20140715(第07期);第1.3-1.5节，第4.2.5-4.5节 *
最近特征分类器的研究与改进;冯庆祥;《中国优秀硕士学位论文全文数据库 信息科技辑》;20140415(第04期);第3.8节，第4.1-4.3节 *

Also Published As

Publication number Publication date
CN104143090A (en) 2014-11-12

Similar Documents

Publication Publication Date Title
JP6398979B2 (en) Video processing apparatus, video processing method, and video processing program
CN109063565B (en) Low-resolution face recognition method and device
CN105787482A (en) Specific target outline image segmentation method based on depth convolution neural network
CN107220604A (en) A kind of fall detection method based on video
CN107194346A (en) A kind of fatigue drive of car Forecasting Methodology
US20110190008A1 (en) Systems, methods, and apparatuses for providing context-based navigation services
CN104143090B (en) A kind of automobile door opening method based on recognition of face
CN104091169A (en) Behavior identification method based on multi feature fusion
CN109145982A (en) The personal identification method and device of driver, storage medium, terminal
CN108875907B (en) Fingerprint identification method and device based on deep learning
CN106062871A (en) Training classifiers using selected cohort sample subsets
CN102254394A (en) Antitheft monitoring method for poles and towers in power transmission line based on video difference analysis
Mobark et al. Improving the accuracy of complex activities recognition using accelerometer-embedded mobile phone classifiers
CN110458210A (en) A kind of driver&#39;s state monitoring method and system based on transfer learning
CN103745238A (en) Pantograph identification method based on AdaBoost and active shape model
Guan Realtime multi-person 2d pose estimation using shufflenet
CN103413116A (en) Effective fingerprint direction field calculating method
CN109614987A (en) More disaggregated model optimization methods, device, storage medium and electronic equipment
CN109637126A (en) A kind of traffic object identifying system and its method based on V2X terminal
CN105718896A (en) Intelligent robot with target recognition function
CN111144462A (en) Unknown individual identification method and device for radar signals
Zhang et al. Positioning method of pedestrian dead reckoning based on human activity recognition assistance
CN104463246A (en) Manifold-based linear regression learning method
CN110288026A (en) A kind of image partition method and device practised based on metric relation graphics
CN104228767B (en) A kind of automobile starting method based on palmprint authentication

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