CN105590102A - Front car face identification method based on deep learning - Google Patents

Front car face identification method based on deep learning Download PDF

Info

Publication number
CN105590102A
CN105590102A CN201511006944.0A CN201511006944A CN105590102A CN 105590102 A CN105590102 A CN 105590102A CN 201511006944 A CN201511006944 A CN 201511006944A CN 105590102 A CN105590102 A CN 105590102A
Authority
CN
China
Prior art keywords
car
image
face
neural network
convolutional neural
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
CN201511006944.0A
Other languages
Chinese (zh)
Inventor
舒泓新
蔡晓东
陈超村
王爱华
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.)
CHINACCS INFORMATION INDUSTRY Co Ltd
Original Assignee
CHINACCS INFORMATION INDUSTRY 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 CHINACCS INFORMATION INDUSTRY Co Ltd filed Critical CHINACCS INFORMATION INDUSTRY Co Ltd
Priority to CN201511006944.0A priority Critical patent/CN105590102A/en
Publication of CN105590102A publication Critical patent/CN105590102A/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/50—Context or environment of the image
    • G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
    • G06V20/58—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
    • G06V20/584—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads of vehicle lights or traffic lights
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • General Engineering & Computer Science (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Multimedia (AREA)
  • Image Analysis (AREA)

Abstract

The invention discloses a front car face identification method based on deep learning, concretely comprising: performing modularized pretreatment on a front car front image, wherein each car image obtains five corresponding car face local image modules; constructing and training a convolutional neural network module, and utilizing global car face characteristics extracted through the convolutional neural network module to train a softmax classifier; obtaining the front car face image of the car image to be identified, and performing modularized pretreatment on the front car face image; and inputting the front car face image into a trained convolutional neural network module to obtain the global car face characteristics, and utilizing the trained softmax classifier to perform classification and identification. The method employs a convolutional neural network deep learning algorithm framework to first perform partitioning characteristic extraction on a car face, and then fuses characteristics; the method has the accuracy higher than a traditional classification method, and plays a good role in fake-licensed car detection and suspect car tracking and searching.

Description

Front vehicle face identification method based on deep learning
Technical Field
The invention relates to the field of image processing, in particular to a front vehicle face identification method based on deep learning.
Background
The requirements on the robustness and reliability of a traffic monitoring system in the current intelligent traffic system are higher and higher, the fake-licensed vehicles are identified in the traffic system so as to prevent traffic violation and even crime evasion, and the combination of vehicle face identification and license plate identification is an effective method for identifying the fake-licensed vehicles; at present, the traditional car face identification method comprises the following two methods: according to the vehicle logo recognition result, a vehicle brand classification method is carried out, and classification recognition is carried out according to the texture features of the front vehicle face; the former method classifies the vehicle types by searching and positioning the area of the vehicle logo in the image and then identifying the vehicle types by the image mode, and the former method can classify the vehicles of common brands, but the former method lacks the capability of accurately classifying the vehicles of the same brand and different types. In the latter method, due to the differences in the layout and shape of the heat sinks and the lamps of the front images of vehicles of different brands and vehicles of the same brand and different types, three technical problems exist, namely, the front image is difficult to accurately position and intercept, the geometric distortion of the front image has certain influence on the recognition rate, the types to be classified are more, and the design of the classifier is relatively complex.
Disclosure of Invention
The invention provides a front vehicle face identification method based on deep learning, aiming at the problems of low identification accuracy and complex classifier of the traditional vehicle face identification.
In order to achieve the above object, the present invention provides a method for recognizing a front vehicle face based on deep learning, which is characterized in that the recognition method specifically comprises:
step S101: reading vehicle pictures acquired by a traffic gate, acquiring front vehicle face images, and performing modular preprocessing on the front vehicle face images to obtain 5 corresponding vehicle face local image modules for each vehicle picture;
step S102: constructing a convolutional neural network, inputting a car face local image module into the convolutional neural network for training to obtain a trained convolutional neural network model, and training a Softmax classifier by using global car face features extracted through the convolutional neural network model; the convolutional neural network comprises 5 convolutional pooling layers, 1 local feature fusion layer and 2 full-connection layers;
step S103: acquiring a front car face image of a picture of a car to be identified, and performing modular preprocessing on the front car face image to obtain 5 car face local image modules; and inputting the vehicle model into a trained convolutional neural network model to obtain global vehicle face characteristics, and performing classification and identification by using a trained Softmax classifier to finally obtain the vehicle model corresponding to the vehicle picture to be identified.
The constructed convolutional neural network comprises 5 convolutional pooling layers, 1 local feature fusion layer and 2 full-connection layers; the convolution pooling layer is used for extracting and keeping the features, the local feature fusion layer is used for fusing the block local features into integral features, and the full connection layer is used for mapping the extracted features onto a feature vector; the number of filters of the convolutional layers in the 5 convolutional pooling layers is 96, 128, 256 and 1024 respectively, the parameter initialization adopts random initialization, and the pooling method of the pooling layers adopts maximum pooling.
In step S101 and step S103, the acquiring of the front vehicle face image specifically includes: the method comprises the steps of carrying out front car face positioning on a car picture by adopting LBP characteristics and an Adaboost classifier, and then obtaining a front car face image by using a classical rectangle method; the classical rectangle method is to take the position of the license plate as a reference and intercept the image of the front car face according to the proportion, and the proportion is determined by adopting a classical empirical value.
The Adaboost algorithm is specifically as follows:
wherein h isjA value representing a simple classifier; thetajIs a threshold value; p is a radical ofjIndicating directions of unequal signs, only taking；fj(x) Representing a characteristic value;
the final strong classifier is:(2)
wherein,，；to representThe error of the classifier is determined by the error of the classifier,the weak classifier representing the smallest error.
In step S101 and step S103, performing modular preprocessing on the front car face image specifically includes: the method comprises the steps of firstly carrying out gray level conversion on a front car face image, converting an RGB three-channel front car face image into a single-channel gray level image, and then partitioning the gray level image according to texture features to obtain 5 car face local image modules.
The specific algorithm of the Softmax classifier is as follows:
the assumed function of Softmax is:
wherein y represents a class label and can take k different values;a training set is represented that represents the training set,and j represents a category of the content,a probability value representing the probability of the category j,the parameters that represent the model are then used,means to normalize the probability distribution such that the sum of all probabilities is 1;
will be provided withBy oneIs represented by the matrix of (a) which is to beObtained by row listing, as follows:
wherein,all model parameters are represented, T represents a transposed matrix, and K represents a dimension;
the cost function for softmax is:
wherein,representing a weight attenuation term, n representing the number of categories, and m representing the number corresponding to a certain category;
the derivative of (c) is:
by minimizingAnd obtaining a Softmax model.
The invention has the beneficial effects that: the method for classifying and identifying the car face by the aid of the deep learning algorithm framework of the convolutional neural network to perform blocking extraction on the car face and feature fusion is performed on the car face, accuracy rate of the method is superior to that of a traditional classification method, and the method is applied to detection of fake-licensed cars and vehicle tracking and searching of suspects to give good benefits.
Drawings
FIG. 1 is a flow chart of a method according to an embodiment of the present invention.
Fig. 2 is a schematic diagram of the technical framework of the present invention.
Fig. 3 is a schematic diagram of the structure of a convolutional neural network.
Fig. 4 is a schematic diagram of the Adaboost algorithm.
Detailed Description
The invention adopts a feature extraction main algorithm flow of a convolutional neural network, provides a front vehicle face recognition method based on deep learning, and simultaneously positions the front vehicle face by using the position relation of a license plate and the vehicle face on the basis of a mature license plate recognition technology; through vehicle image preprocessing, the vehicle face is added into a deep learning model for training, and finally classification and comparison of the vehicle face are realized.
The embodiment of the invention provides a mixed local feature extraction and comparison method based on deep learning, the flow of which is shown in fig. 1 and 2, and the method comprises the following steps:
step S101: reading vehicle pictures acquired by a traffic gate, acquiring front vehicle face images, and performing modular preprocessing on the front vehicle face images to obtain 5 corresponding vehicle face local image modules for each vehicle picture;
step S102: constructing a convolutional neural network, inputting a car face local image module into the convolutional neural network for training to obtain a trained convolutional neural network model, and training a Softmax classifier by using global car face features extracted through the convolutional neural network model;
step S103: acquiring a front car face image of a picture of a car to be identified, and performing modular preprocessing on the front car face image to obtain 5 car face local image modules; and inputting the vehicle model into a trained convolutional neural network model to obtain global vehicle face characteristics, and performing classification and identification by using a trained Softmax classifier to finally obtain the vehicle model corresponding to the vehicle picture to be identified.
In step S102, a convolutional neural network (see fig. 3) is constructed, which includes 5 convolutional pooling layers, 1 local feature fusion layer, and 2 full-link layers; the convolution layer is used for extracting the features, the pooling layer is used for keeping the features so as not to lose too many bottom-layer features, the local feature fusion layer is used for fusing the block local features into overall features, and the full-connection layer is used for mapping the extracted features to a feature vector; the number of filters of the 5 convolutional layers is respectively 96, 128, 256 and 1024, the parameter initialization adopts random initialization, and the pooling layer pooling method adopts maximum pooling. The number of output feature vectors of a convolutional neural network feature layer (namely, a first layer of fully-connected layer) is 1024, the number of output nodes of a last layer of fully-connected layer is 320, and the ownership weight of the convolutional neural network is randomly initialized; when the deep learning network features are extracted, each local feature is extracted in a blocking mode, and the rear connection feature fusion layer fuses 5 blocking local features to form global car face features, so that the global and local features are considered, and the precision is improved.
In step S101 and step S103, the acquiring of the front vehicle face image specifically includes: the method comprises the steps of carrying out front car face positioning on a car picture by adopting LBP characteristics and an Adaboost classifier, and then obtaining a front car face image by using a classical rectangle method; the classical rectangle method is to take the position of the license plate as a reference, intercept the front vehicle face image according to the proportion, and determine the proportion by adopting a classical empirical value.
Referring to fig. 4, Adaboost is formed by connecting a plurality of weak classifiers into a strong classifier, so as to detect the car face, and the basic principle of the Adaboost algorithm is expressed as follows:
wherein h isjA value representing a simple classifier; thetajIs a threshold value; p is a radical ofjIndicating directions of unequal signs, only taking；fj(x) The characteristic value is represented.
The final strong classifier is:(3)
wherein,the weak classifier representing the smallest error is selected,to representThe error of the classifier is determined by the error of the classifier,，。
in step S101 and step S103, the module-based preprocessing of the front car face image specifically includes: carrying out gray level conversion on a front vehicle face image, converting an RGB three-channel front vehicle face image into a single-channel gray level image, and partitioning the gray level image according to texture characteristics to obtain 5 vehicle face local image modules; the formula for converting the color front car face image to gray scale is as follows:
f(i，j)＝0.2999R+0.587G+0.114B
where f (i, j) is the gray value of the pixel at the image coordinate (i, j) after graying, and R, G, B is distributed as the three components of the color image RGB.
In step S104, the Softmax specific algorithm is:
the assumed function of Softmax is:
wherein y represents a class label and can take k different values;a training set is represented that represents the training set,and j represents a category of the content,a probability value representing the probability of the category j,the parameters that represent the model are then used,means to normalize the probability distribution such that the sum of all probabilities is 1;
will be provided withBy oneIs represented by the matrix of (a) which is to beObtained by row listing, as follows:
wherein,all model parameters are represented, T represents a transposed matrix, and K represents a dimension;
the cost function for softmax is:
wherein,representing a weight attenuation term, n representing the number of categories, and m representing the number corresponding to a certain category;
the derivative of (c) is:
by minimizingAnd obtaining a Softmax model.
According to the invention, local features are extracted through local feature training and then are fused into integral car face features, the input car face images are directly classified and identified by using a deep learning training model obtained through training, and car face images which are the same as the target car face can be found from thousands of car face libraries through feature comparison of the input target car face images, so that good benefits of tracking and identification can be achieved in practical application.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.

Claims (6)

1. The front vehicle face identification method based on deep learning is characterized by comprising the following steps of:
step S101: reading vehicle pictures acquired by a traffic gate, acquiring front vehicle face images, and performing modular preprocessing on the front vehicle face images to obtain 5 corresponding vehicle face local image modules for each vehicle picture;
step S102: constructing a convolutional neural network, inputting a car face local image module into the convolutional neural network for training to obtain a trained convolutional neural network model, and training a Softmax classifier by using global car face features extracted through the convolutional neural network model; the convolutional neural network comprises 5 convolutional pooling layers, 1 local feature fusion layer and 2 full-connection layers;
step S103: acquiring a front car face image of a picture of a car to be identified, and performing modular preprocessing on the front car face image to obtain 5 car face local image modules; and inputting the vehicle model into a trained convolutional neural network model to obtain global vehicle face characteristics, and performing classification and identification by using a trained Softmax classifier to finally obtain the vehicle model corresponding to the vehicle picture to be identified.
2. A front vehicle face recognition method based on deep learning according to claim 1, wherein the convolutional neural network comprises 5 convolutional pooling layers, 1 local feature fusion layer and 2 full-link layers; the convolution pooling layer is used for extracting and keeping the features, the local feature fusion layer is used for fusing the block local features into integral features, and the full connection layer is used for mapping the extracted features onto a feature vector; the number of filters of the convolutional layers in the 5 convolutional pooling layers is 96, 128, 256 and 1024 respectively, the parameter initialization adopts random initialization, and the pooling method of the pooling layers adopts maximum pooling.
3. A method for recognizing a front vehicle face based on deep learning according to claim 1, wherein in step S101 and step S103, the obtaining of the front vehicle face image specifically includes: and (3) carrying out front car face positioning on the car picture by adopting LBP (local binary pattern) characteristics and an Adaboost classifier, and then obtaining a front car face image by using a classical rectangle method.
4. A method for recognizing a front vehicle face based on deep learning according to claim 1, wherein in step S101 and step S103, performing modular preprocessing on the front vehicle face image specifically comprises: the method comprises the steps of firstly carrying out gray level conversion on a front car face image, converting an RGB three-channel front car face image into a single-channel gray level image, and then partitioning the gray level image according to texture features to obtain 5 car face local image modules.
5. A front vehicle face recognition method based on deep learning according to claim 3, wherein the Adaboost algorithm is specifically as follows:
wherein h isjA value representing a simple classifier; thetajIs a threshold value; p is a radical ofjIndicating directions of unequal signs, only taking；fj(x) Representing a characteristic value;
the final strong classifier is:(2)
wherein,，；to representThe error of the classifier is determined by the error of the classifier,the weak classifier representing the smallest error.
6. A front vehicle face recognition method based on deep learning as claimed in claim 1, wherein the specific algorithm of the Softmax classifier is as follows:
the assumed function of Softmax is:
wherein y represents a class label and can take k different values;a training set is represented that represents the training set,and j represents a category of the content,a probability value representing the probability of the category j,the parameters that represent the model are then used,means to normalize the probability distribution such that the sum of all probabilities is 1;
will be provided withBy oneIs represented by the matrix of (a) which is to beObtained by row listing, as follows:
wherein,all model parameters are represented, T represents a transposed matrix, and K represents a dimension;
the cost function for Softmax is:
wherein,representing a weight attenuation term, n representing the number of categories, and m representing the number corresponding to a certain category;
the derivative of (c) is:
by minimizingAnd obtaining a Softmax model.
CN201511006944.0A 2015-12-30 2015-12-30 Front car face identification method based on deep learning Pending CN105590102A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201511006944.0A CN105590102A (en) 2015-12-30 2015-12-30 Front car face identification method based on deep learning

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201511006944.0A CN105590102A (en) 2015-12-30 2015-12-30 Front car face identification method based on deep learning

Publications (1)

Publication Number Publication Date
CN105590102A true CN105590102A (en) 2016-05-18

Family

ID=55929672

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201511006944.0A Pending CN105590102A (en) 2015-12-30 2015-12-30 Front car face identification method based on deep learning

Country Status (1)

Country Link
CN (1) CN105590102A (en)

Cited By (40)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106250832A (en) * 2016-07-25 2016-12-21 华南理工大学 A kind of national recognition methods based on integrated convolutional neural networks
CN106529446A (en) * 2016-10-27 2017-03-22 桂林电子科技大学 Vehicle type identification method and system based on multi-block deep convolutional neural network
CN106529461A (en) * 2016-11-07 2017-03-22 湖南源信光电科技有限公司 Vehicle model identifying algorithm based on integral characteristic channel and SVM training device
CN106548145A (en) * 2016-10-31 2017-03-29 北京小米移动软件有限公司 Image-recognizing method and device
CN106557579A (en) * 2016-11-28 2017-04-05 中通服公众信息产业股份有限公司 A kind of vehicle model searching system and method based on convolutional neural networks
CN106570516A (en) * 2016-09-06 2017-04-19 国网重庆市电力公司电力科学研究院 Obstacle recognition method using convolution neural network
CN106611156A (en) * 2016-11-03 2017-05-03 桂林电子科技大学 Pedestrian recognition method and system capable of self-adapting to deep space features
CN106647758A (en) * 2016-12-27 2017-05-10 深圳市盛世智能装备有限公司 Target object detection method and device and automatic guiding vehicle following method
CN106682649A (en) * 2017-01-24 2017-05-17 成都容豪电子信息科技有限公司 Vehicle type recognition method based on deep learning
CN106709528A (en) * 2017-01-10 2017-05-24 深圳大学 Method and device of vehicle reidentification based on multiple objective function deep learning
CN106803090A (en) * 2016-12-05 2017-06-06 中国银联股份有限公司 A kind of image-recognizing method and device
CN106934378A (en) * 2017-03-16 2017-07-07 山东建筑大学 A kind of dazzle light identifying system and method based on video depth study
CN107067027A (en) * 2017-03-08 2017-08-18 同济大学 A kind of target identification method based on smooth multi-instance learning
CN107133578A (en) * 2017-04-19 2017-09-05 华南理工大学 A kind of facial expression recognizing method transmitted based on file and system
CN107368886A (en) * 2017-02-23 2017-11-21 奥瞳系统科技有限公司 Based on the nerve network system for reusing small-scale convolutional neural networks module
CN107516061A (en) * 2016-06-17 2017-12-26 北京市商汤科技开发有限公司 A kind of image classification method and system
WO2017220032A1 (en) * 2016-06-24 2017-12-28 平安科技（深圳）有限公司 Vehicle license plate classification method and system based on deep learning, electronic apparatus, and storage medium
WO2018017319A1 (en) * 2016-07-22 2018-01-25 Nec Laboratories America, Inc. Liveness detection for antispoof face recognition
CN108009579A (en) * 2017-11-29 2018-05-08 合肥寰景信息技术有限公司 Special vehicle detection and identifying system based on deep learning
CN108052861A (en) * 2017-11-08 2018-05-18 北京卓视智通科技有限责任公司 A kind of nerve network system and the model recognizing method based on the nerve network system
CN108171276A (en) * 2018-01-17 2018-06-15 百度在线网络技术（北京）有限公司 For generating the method and apparatus of information
CN108229302A (en) * 2017-11-10 2018-06-29 深圳市商汤科技有限公司 Feature extracting method, device, computer program, storage medium and electronic equipment
CN108334892A (en) * 2017-12-26 2018-07-27 新智数字科技有限公司 A kind of model recognizing method, device and equipment based on convolutional neural networks
CN108806792A (en) * 2017-05-03 2018-11-13 金波 Deep learning facial diagnosis system
CN108874911A (en) * 2018-05-28 2018-11-23 广西师范学院 Suspect's position predicting method based on regional environment Yu crime dramas data
CN109100648A (en) * 2018-05-16 2018-12-28 上海海事大学 Ocean current generator impeller based on CNN-ARMA-Softmax winds failure fusion diagnosis method
CN109145928A (en) * 2017-06-16 2019-01-04 杭州海康威视数字技术股份有限公司 It is a kind of based on the headstock of image towards recognition methods and device
CN109359666A (en) * 2018-09-07 2019-02-19 佳都新太科技股份有限公司 A kind of model recognizing method and processing terminal based on multiple features fusion neural network
CN109886978A (en) * 2019-02-20 2019-06-14 贵州电网有限责任公司 A kind of end-to-end warning information recognition methods based on deep learning
CN110781931A (en) * 2019-10-14 2020-02-11 国家广播电视总局广播电视科学研究院 Ultrahigh-definition film source conversion curve detection method for local feature extraction and fusion
CN110826411A (en) * 2019-10-10 2020-02-21 电子科技大学 Vehicle target rapid identification method based on unmanned aerial vehicle image
CN110889428A (en) * 2019-10-21 2020-03-17 浙江大搜车软件技术有限公司 Image recognition method and device, computer equipment and storage medium
CN111126173A (en) * 2019-12-04 2020-05-08 玉林师范学院 High-precision face detection method
CN111340026A (en) * 2020-03-05 2020-06-26 苏州科达科技股份有限公司 Training method of vehicle annual payment identification model and vehicle annual payment identification method
CN111368852A (en) * 2018-12-26 2020-07-03 沈阳新松机器人自动化股份有限公司 Article identification and pre-sorting system and method based on deep learning and robot
CN113298021A (en) * 2021-06-11 2021-08-24 宿州学院 Mining area transport vehicle head and tail identification method and system based on convolutional neural network
CN113344040A (en) * 2021-05-20 2021-09-03 深圳索信达数据技术有限公司 Image classification method and device, computer equipment and storage medium
US11113513B2 (en) 2019-02-19 2021-09-07 Fujitsu Limited Apparatus and method for training classification model and apparatus for classifying with classification model
CN107871103B (en) * 2016-09-23 2021-10-19 北京眼神科技有限公司 Face authentication method and device
US11961335B1 (en) 2020-06-26 2024-04-16 Harris County Toll Road Authority Dual mode electronic toll road system

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100278436A1 (en) * 2009-04-30 2010-11-04 Industrial Technology Research Institute Method and system for image identification and identification result output
CN104077577A (en) * 2014-07-03 2014-10-01 浙江大学 Trademark detection method based on convolutional neural network
CN104200224A (en) * 2014-08-28 2014-12-10 西北工业大学 Valueless image removing method based on deep convolutional neural networks
CN104298976A (en) * 2014-10-16 2015-01-21 电子科技大学 License plate detection method based on convolutional neural network
CN104463172A (en) * 2014-12-09 2015-03-25 中国科学院重庆绿色智能技术研究院 Face feature extraction method based on face feature point shape drive depth model
CN104809443A (en) * 2015-05-05 2015-07-29 上海交通大学 Convolutional neural network-based license plate detection method and system
CN105005774A (en) * 2015-07-28 2015-10-28 中国科学院自动化研究所 Face relative relation recognition method based on convolutional neural network and device thereof

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20100278436A1 (en) * 2009-04-30 2010-11-04 Industrial Technology Research Institute Method and system for image identification and identification result output
CN104077577A (en) * 2014-07-03 2014-10-01 浙江大学 Trademark detection method based on convolutional neural network
CN104200224A (en) * 2014-08-28 2014-12-10 西北工业大学 Valueless image removing method based on deep convolutional neural networks
CN104298976A (en) * 2014-10-16 2015-01-21 电子科技大学 License plate detection method based on convolutional neural network
CN104463172A (en) * 2014-12-09 2015-03-25 中国科学院重庆绿色智能技术研究院 Face feature extraction method based on face feature point shape drive depth model
CN104809443A (en) * 2015-05-05 2015-07-29 上海交通大学 Convolutional neural network-based license plate detection method and system
CN105005774A (en) * 2015-07-28 2015-10-28 中国科学院自动化研究所 Face relative relation recognition method based on convolutional neural network and device thereof

Cited By (56)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107516061A (en) * 2016-06-17 2017-12-26 北京市商汤科技开发有限公司 A kind of image classification method and system
CN107516061B (en) * 2016-06-17 2020-04-07 北京市商汤科技开发有限公司 Image classification method and system
US10528841B2 (en) 2016-06-24 2020-01-07 Ping An Technology (Shenzhen) Co., Ltd. Method, system, electronic device, and medium for classifying license plates based on deep learning
WO2017220032A1 (en) * 2016-06-24 2017-12-28 平安科技（深圳）有限公司 Vehicle license plate classification method and system based on deep learning, electronic apparatus, and storage medium
WO2018017319A1 (en) * 2016-07-22 2018-01-25 Nec Laboratories America, Inc. Liveness detection for antispoof face recognition
CN106250832A (en) * 2016-07-25 2016-12-21 华南理工大学 A kind of national recognition methods based on integrated convolutional neural networks
CN106570516A (en) * 2016-09-06 2017-04-19 国网重庆市电力公司电力科学研究院 Obstacle recognition method using convolution neural network
CN107871103B (en) * 2016-09-23 2021-10-19 北京眼神科技有限公司 Face authentication method and device
CN106529446A (en) * 2016-10-27 2017-03-22 桂林电子科技大学 Vehicle type identification method and system based on multi-block deep convolutional neural network
CN106548145A (en) * 2016-10-31 2017-03-29 北京小米移动软件有限公司 Image-recognizing method and device
CN106611156A (en) * 2016-11-03 2017-05-03 桂林电子科技大学 Pedestrian recognition method and system capable of self-adapting to deep space features
CN106611156B (en) * 2016-11-03 2019-12-20 桂林电子科技大学 Pedestrian identification method and system based on self-adaptive depth space characteristics
CN106529461A (en) * 2016-11-07 2017-03-22 湖南源信光电科技有限公司 Vehicle model identifying algorithm based on integral characteristic channel and SVM training device
CN106557579A (en) * 2016-11-28 2017-04-05 中通服公众信息产业股份有限公司 A kind of vehicle model searching system and method based on convolutional neural networks
CN106557579B (en) * 2016-11-28 2020-08-25 中通服公众信息产业股份有限公司 Vehicle model retrieval system and method based on convolutional neural network
CN106803090A (en) * 2016-12-05 2017-06-06 中国银联股份有限公司 A kind of image-recognizing method and device
CN106647758A (en) * 2016-12-27 2017-05-10 深圳市盛世智能装备有限公司 Target object detection method and device and automatic guiding vehicle following method
CN106709528A (en) * 2017-01-10 2017-05-24 深圳大学 Method and device of vehicle reidentification based on multiple objective function deep learning
CN106682649A (en) * 2017-01-24 2017-05-17 成都容豪电子信息科技有限公司 Vehicle type recognition method based on deep learning
CN107368886A (en) * 2017-02-23 2017-11-21 奥瞳系统科技有限公司 Based on the nerve network system for reusing small-scale convolutional neural networks module
CN107368886B (en) * 2017-02-23 2020-10-02 奥瞳系统科技有限公司 Neural network system based on repeatedly used small-scale convolutional neural network module
CN107067027B (en) * 2017-03-08 2020-06-23 同济大学 Target identification method based on smooth multi-instance learning
CN107067027A (en) * 2017-03-08 2017-08-18 同济大学 A kind of target identification method based on smooth multi-instance learning
CN106934378B (en) * 2017-03-16 2020-04-24 山东建筑大学 Automobile high beam identification system and method based on video deep learning
CN106934378A (en) * 2017-03-16 2017-07-07 山东建筑大学 A kind of dazzle light identifying system and method based on video depth study
CN107133578B (en) * 2017-04-19 2020-05-22 华南理工大学 Facial expression recognition method and system based on file transmission
CN107133578A (en) * 2017-04-19 2017-09-05 华南理工大学 A kind of facial expression recognizing method transmitted based on file and system
CN108806792B (en) * 2017-05-03 2022-01-04 金波 Deep learning face diagnosis system
CN108806792A (en) * 2017-05-03 2018-11-13 金波 Deep learning facial diagnosis system
CN109145928A (en) * 2017-06-16 2019-01-04 杭州海康威视数字技术股份有限公司 It is a kind of based on the headstock of image towards recognition methods and device
CN109145928B (en) * 2017-06-16 2020-10-27 杭州海康威视数字技术股份有限公司 Method and device for identifying vehicle head orientation based on image
CN108052861A (en) * 2017-11-08 2018-05-18 北京卓视智通科技有限责任公司 A kind of nerve network system and the model recognizing method based on the nerve network system
CN108229302A (en) * 2017-11-10 2018-06-29 深圳市商汤科技有限公司 Feature extracting method, device, computer program, storage medium and electronic equipment
CN108009579A (en) * 2017-11-29 2018-05-08 合肥寰景信息技术有限公司 Special vehicle detection and identifying system based on deep learning
CN108334892B (en) * 2017-12-26 2020-11-17 新智数字科技有限公司 Vehicle type identification method, device and equipment based on convolutional neural network
CN108334892A (en) * 2017-12-26 2018-07-27 新智数字科技有限公司 A kind of model recognizing method, device and equipment based on convolutional neural networks
CN108171276B (en) * 2018-01-17 2019-07-23 百度在线网络技术（北京）有限公司 Method and apparatus for generating information
CN108171276A (en) * 2018-01-17 2018-06-15 百度在线网络技术（北京）有限公司 For generating the method and apparatus of information
CN109100648B (en) * 2018-05-16 2020-07-24 上海海事大学 CNN-ARMA-Softmax-based ocean current generator impeller winding fault fusion diagnosis method
CN109100648A (en) * 2018-05-16 2018-12-28 上海海事大学 Ocean current generator impeller based on CNN-ARMA-Softmax winds failure fusion diagnosis method
CN108874911A (en) * 2018-05-28 2018-11-23 广西师范学院 Suspect's position predicting method based on regional environment Yu crime dramas data
CN109359666A (en) * 2018-09-07 2019-02-19 佳都新太科技股份有限公司 A kind of model recognizing method and processing terminal based on multiple features fusion neural network
CN111368852A (en) * 2018-12-26 2020-07-03 沈阳新松机器人自动化股份有限公司 Article identification and pre-sorting system and method based on deep learning and robot
US11113513B2 (en) 2019-02-19 2021-09-07 Fujitsu Limited Apparatus and method for training classification model and apparatus for classifying with classification model
CN109886978A (en) * 2019-02-20 2019-06-14 贵州电网有限责任公司 A kind of end-to-end warning information recognition methods based on deep learning
CN110826411A (en) * 2019-10-10 2020-02-21 电子科技大学 Vehicle target rapid identification method based on unmanned aerial vehicle image
CN110826411B (en) * 2019-10-10 2022-05-03 电子科技大学 Vehicle target rapid identification method based on unmanned aerial vehicle image
CN110781931A (en) * 2019-10-14 2020-02-11 国家广播电视总局广播电视科学研究院 Ultrahigh-definition film source conversion curve detection method for local feature extraction and fusion
CN110781931B (en) * 2019-10-14 2022-03-08 国家广播电视总局广播电视科学研究院 Ultrahigh-definition film source conversion curve detection method for local feature extraction and fusion
CN110889428A (en) * 2019-10-21 2020-03-17 浙江大搜车软件技术有限公司 Image recognition method and device, computer equipment and storage medium
CN111126173A (en) * 2019-12-04 2020-05-08 玉林师范学院 High-precision face detection method
CN111126173B (en) * 2019-12-04 2023-05-26 玉林师范学院 High-precision face detection method
CN111340026A (en) * 2020-03-05 2020-06-26 苏州科达科技股份有限公司 Training method of vehicle annual payment identification model and vehicle annual payment identification method
US11961335B1 (en) 2020-06-26 2024-04-16 Harris County Toll Road Authority Dual mode electronic toll road system
CN113344040A (en) * 2021-05-20 2021-09-03 深圳索信达数据技术有限公司 Image classification method and device, computer equipment and storage medium
CN113298021A (en) * 2021-06-11 2021-08-24 宿州学院 Mining area transport vehicle head and tail identification method and system based on convolutional neural network

Similar Documents

Publication Publication Date Title
CN105590102A (en) Front car face identification method based on deep learning
CN112233097B (en) Road scene other vehicle detection system and method based on space-time domain multi-dimensional fusion
CN106845487B (en) End-to-end license plate identification method
Kim et al. An Efficient Color Space for Deep‐Learning Based Traffic Light Recognition
CN107563372B (en) License plate positioning method based on deep learning SSD frame
CN111767882A (en) Multi-mode pedestrian detection method based on improved YOLO model
Jiao et al. A configurable method for multi-style license plate recognition
CN112241728B (en) Real-time lane line detection method and system for learning context information by adopting attention mechanism
CN110119726B (en) Vehicle brand multi-angle identification method based on YOLOv3 model
CN111191667B (en) Crowd counting method based on multiscale generation countermeasure network
CN111611874B (en) Face mask wearing detection method based on ResNet and Canny
CN107545263B (en) Object detection method and device
Kuang et al. Feature selection based on tensor decomposition and object proposal for night-time multiclass vehicle detection
Lee et al. Available parking slot recognition based on slot context analysis
Li et al. Robust vehicle detection in high-resolution aerial images with imbalanced data
CN110287798B (en) Vector network pedestrian detection method based on feature modularization and context fusion
CN113723377A (en) Traffic sign detection method based on LD-SSD network
CN112464731B (en) Traffic sign detection and identification method based on image processing
Gao et al. Synergizing appearance and motion with low rank representation for vehicle counting and traffic flow analysis
CN105989334A (en) Road detection method based on monocular vision
CN116912485A (en) Scene semantic segmentation method based on feature fusion of thermal image and visible light image
Zhang et al. Adaptive dense pyramid network for object detection in UAV imagery
CN113011308A (en) Pedestrian detection method introducing attention mechanism
Aldahoul et al. A comparison between various human detectors and CNN-based feature extractors for human activity recognition via aerial captured video sequences
Mittal et al. Vehicle detection and traffic density estimation using ensemble of deep learning models

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20160518