CN106023154A - Multi-temporal SAR image change detection method based on dual-channel convolutional neural network (CNN) - Google Patents

Multi-temporal SAR image change detection method based on dual-channel convolutional neural network (CNN) Download PDF

Info

Publication number
CN106023154A
CN106023154A CN201610301686.7A CN201610301686A CN106023154A CN 106023154 A CN106023154 A CN 106023154A CN 201610301686 A CN201610301686 A CN 201610301686A CN 106023154 A CN106023154 A CN 106023154A
Authority
CN
China
Prior art keywords
cnn
training
image block
image
sar image
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201610301686.7A
Other languages
Chinese (zh)
Other versions
CN106023154B (en
Inventor
李映
刘韬
徐隆浩
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.)
Chongqing Commercial Service Technology Co.,Ltd.
Original Assignee
Northwestern Polytechnical University
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 Northwestern Polytechnical University filed Critical Northwestern Polytechnical University
Priority to CN201610301686.7A priority Critical patent/CN106023154B/en
Publication of CN106023154A publication Critical patent/CN106023154A/en
Application granted granted Critical
Publication of CN106023154B publication Critical patent/CN106023154B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/0002—Inspection of images, e.g. flaw detection
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10032—Satellite or aerial image; Remote sensing
    • G06T2207/10044—Radar image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20081—Training; Learning

Landscapes

  • Engineering & Computer Science (AREA)
  • Quality & Reliability (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Image Analysis (AREA)

Abstract

The invention relates to a multi-temporal SAR image change detection method based on a dual-channel convolutional neural network (CNN). A depth neural network usually comprises three models which are respectively a stack self-encoder, a depth confidence network and the CNN. According to the invention, a training mode of the CNN is used, two CNN models of the same structure are constructed first of all, a part of sample points are randomly selected, change detection results of these sample points are manually calibrated, these sample points and calibration detection results are taken as training data for training the CNN, and then final change detection results are obtained by performing change detection on unclassified samples by use of the trained CNN.

Description

Multidate SAR image change detection based on dual pathways convolutional neural networks
Technical field
The invention belongs to multidate SAR image change detection, relate to a kind of based on dual pathways convolutional neural networks Multidate SAR image change detection.
Background technology
The common technology of multidate SAR image change-detection is first to take the logarithm respectively to enter by the SAR image of two phases again Row difference, then carries out two classification in units of pixel by difference diagram；Or first SAR image is split, then with segmentation After object as ultimate unit, it is carried out two classification.In the prior art, extract which feature and classify, generally by It is manually set, frequently can lead to many information that image itself contains and be left in the basket so that testing result is the most accurate.
Deep neural network independently extracts characteristics of image by the way of successively training, and efficiently solves and artificially extracts spy The problem that during levying, information is omitted, what convolutional neural networks therein had enriched has been applied in image procossing.But, mesh Front in terms of multidate SAR image change-detection, use convolutional neural networks algorithm not yet to obtain good testing result, fail Effectively play deep neural network advantage in terms of processing big data.
Summary of the invention
Solve the technical problem that
In place of the deficiencies in the prior art, the present invention propose a kind of based on dual pathways convolutional neural networks many time Phase SAR image change detection, by convolutional neural networks Technology application to multidate SAR image change-detection field to solve Prior art detects the problem that degree of accuracy is the most high.
Technical scheme
A kind of multidate SAR image change detection based on dual pathways convolutional neural networks, it is characterised in that step As follows:
Step 1: to t1,t2Two SAR image to be detected of two moment same place acquisitions have carried out registration, radiation school Just and geometric correction process；Data after processing are carried out feature extraction, and extracting method is: extract t at random1,t2Moment two width In SAR figure, the image block of n × n centered by each pixel is as primitive character, and in extraction primitive character, 1-5% is as training Sample, then extract 1% conduct checking sample set of total number of samples amount at random, remaining sample is as test sample collection；
Step 2: the CNN model of two identical topological structure of structure, combines the result that each CNN convolutional layer exports Come, then add logistic regression grader, obtain a new CNN model, this model is used the training data that upper step is extracted It is trained；
The input of described model is the image block of two n × n, and wherein first image block is t1Moment SAR image carries N × n the image block taken, second image block is t2N × n the image block extracted in time chart picture；
The input of first CNN is from t1The image block extracted in moment SAR image；The input of second CNN is from t2Time Carve the image block extracted in SAR image；
Step 3, new CNN model is trained:
Convolution kernel parameter in each CNN in new CNN model is adjusted, using sum of squares function as generation during training Valency function, uses BP algorithm amendment convolution kernel and biasing；
With the checking sample set in step 1, new CNN model is verified, prevents from training the generation of Expired Drugs, Training is both needed to the error of output checking collection result every time；
When the error of checking collection result is become, from the trend tapered into, the trend become larger, representative model extensive Reduced capability, iteration ends has been trained；Iterations is within 800 times；
Step 4: all primitive characters of image are inputted in the new CNN model trained sample to be tested is classified Detection, obtains final change-detection result figure.
Beneficial effect
A kind of based on dual pathways convolutional neural networks the multidate SAR image change detection that the present invention proposes, deeply Degree neutral net generally includes three models, is respectively as follows: stacking own coding device, degree of depth confidence network, convolutional neural networks.This The training method of bright use convolutional neural networks (Convolutional Neural Network, CNN), first two structures of structure Identical CNN model, random choose part sample point, and artificially demarcate the change-detection result of these sample points, by these samples Point and demarcation testing result are as the training data training CNN.Re-use the CNN trained and unfiled sample is changed inspection Survey and obtain final change-detection result.
Accompanying drawing explanation
Fig. 1 is the composition structured flowchart of the dual pathways CNN detection network that the present invention proposes
Fig. 2 is the detection method flow chart that the present invention proposes
Fig. 3 present invention propose in the mode that merges of two CNN
Detailed description of the invention
In conjunction with embodiment, accompanying drawing, the invention will be further described:
Embodiment specifically comprises the following steps that
Step 1, choose training sample
If moment t1,t2Be two the most in the same time, on the same band to same place obtain respectively at t1,t2The two of moment Width SAR image X1,X2, and X1,X2Pass through registration and radiant correction and geometric correction.
If X1,X2The size of two width SAR image is p × q, to each coordinate points of two width images (i, j),Wherein n represents extracted respectively in two width images centered by this coordinate points Individual n × n neighborhood as a pair corresponding image block, the pixel of such as 9 × 9, by the image block of the two n × n (referred to as: figure As block to)WithAs a sample Tr (i, primitive character j), Respectively as two corresponding input signals of CNN, in real change testing result figure, (i, pixel value j) is as this sample for coordinate points This teacher signal.
The sub-fraction of the random total number of samples amount extracting two width image pixels, such as 1%-5% is as training sample set, Wherein X1,X2Middle got should be corresponding pixel.Another fraction of extraction total number of samples amount the most at random, such as 1% As checking sample set, wherein X1,X2Middle got should be corresponding pixel.Remaining sample is as test sample collection.
Step 2, the network structure of structure dual pathways CNN model
Common CNN is made up of multiple convolutional layers and pond layer (i.e. pooling layer), carries out CNN in this model Improving, described CNN simply used convolutional layer.Convolutional layer is, by different convolution kernels, every layer of input is carried out convolution behaviour Making, for different convolution kernels, the response intensity of input picture can difference.Just can obtain through excitation function again after convolution terminates Input to next layer.In the present invention use excitation function f (x) be limit linear unit (Rectified Linear Units, ReLu), its mathematic(al) representation is:
F (x)=max (0, x)
Constructing two CNN with identical topological structure, i.e. two the CNN numbers of plies are identical, and the function of corresponding every layer is identical, right The each convolutional layer answered has the convolution kernel of same size, same number.In the present invention, the excitation function of each layer is ReLu, last convolutional layer does not use excitation function.In the present embodiment, in each CNN, the number of convolutional layer is 5, each layer The convolution kernel size of convolutional layer is 3 × 3, and every layer of convolution kernel quantity is successively decreased afterwards for being first incremented by, no more than 100 convolution kernels.
To X1,X2In certain coordinate points (i, j), whereinNetwork defeated Entering is the image block of two n × n of extraction in step 1, and first CNN only processes t1That extracts in moment SAR image is former Beginning featureSecond CNN only locates reason t2The primitive character extracted in moment SAR image
After last convolutional layer of two CNN, add a full articulamentum, two CNN are coupled together, and passes through The result extracted is merged (merging mode as shown in Figure 3) by this full articulamentum, then by logistic regression grader to SAR Image detects.
Step 3, network is trained
The primitive character of training sample set is sent into CNN according to the output of logistic regression grader, image is classified, Calculate the error of network class result and training sample set teacher signal.Because training sample is more, so using in batches (batch) training.During piecemeal, the order of all sample sets is upset at random, more every m sample is put together as one batch.When calculating error, definition cost function c is sum of squares function:Wherein, m represents batch's Size, typically takes 20-100 image block pair, tiRepresent that i-th image block is to corresponding teacher signal, ziRepresent through network operations The testing result value of rear output i-th image block pair.
Dual pathways CNN model is trained by the back-propagation algorithm using error, calculation cost function c pair Weights W1, convolution kernel W2And the partial derivative of biasing b, then to weights, convolution kernel be biased into Row sum-equal matrix:Wherein subscript old represents old Value, subscript n ew represents new value, and η is learning rate.η=0.01 in the present embodiment.
Wherein the error of last layer of logistic regression layer passes full articulamentum back, and full articulamentum is divided into two parts transmission error To two CNN；In full articulamentum, the node from first CNN model passes error back first CNN model, from second Error is passed back second CNN model by the node of CNN model.
Step 4, choose checking collection verify
The checking set pair training result using step 1 to extract is verified, it should be noted that complete at network training Before, checking collection will not change, and requires that checking collection is misaligned with the sample of training set.The effect of checking collection is to prevent instruction Practice the situation of over-fitting.After having trained each time, the error of checking collection sample can be tried to achieve.
Step 5, judge end condition
Owing to the error of training sample set can be gradually reduced along with the increase of iterations, the error meeting of checking sample set First reducing and increase afterwards, in the present invention, choosing training end condition is: when the error of checking collection starts from becoming by being gradually reduced When being gradually increased, it is believed that whole network has begun to over-fitting, now get final product deconditioning；Otherwise return step 3.Generally iteration Number of times is within 800 times.
Step 6, classification
After training terminates, it is possible to utilize the detection network handles test sample based on CNN trained originally to detect, obtain Final change-detection result figure.

Claims (1)

1. a multidate SAR image change detection based on dual pathways convolutional neural networks, it is characterised in that step is such as Under:
Step 1: to t1,t2Two the moment same place obtain two SAR image to be detected carried out registration, radiant correction and Geometric correction processes；Data after processing are carried out feature extraction, and extracting method is: extract t at random1,t2Moment two width SAR schemes In the image block of n × n centered by each pixel as primitive character, extract in primitive character 1-5% as training sample, Extracting 1% conduct checking sample set of total number of samples amount the most at random, remaining sample is as test sample collection；
Step 2: the CNN model of two identical topological structure of structure, combines the result that each CNN convolutional layer exports, then Adding logistic regression grader, obtain a new CNN model, the training data using upper step to extract on this model is instructed Practice；
The input of described model is the image block of two n × n, and wherein first image block is t1The n extracted in moment SAR image × n image block, second image block is t2N × n the image block extracted in time chart picture；
The input of first CNN is from t1The image block extracted in moment SAR image；The input of second CNN is from t2Moment The image block extracted in SAR image；
Step 3, new CNN model is trained:
Convolution kernel parameter in each CNN in new CNN model is adjusted, using sum of squares function as cost letter during training Number, uses BP algorithm amendment convolution kernel and biasing；
With the checking sample set in step 1, new CNN model is verified, prevent from training the generation of Expired Drugs, every time Training is both needed to the error of output checking collection result；
When the error of checking collection result is become, from the trend tapered into, the trend become larger, the generalization ability of representative model Weakening, iteration ends has been trained；Iterations is within 800 times；
Step 4: all primitive characters of image are inputted in the new CNN model trained and sample to be tested is carried out classification inspection Survey, obtain final change-detection result figure.
CN201610301686.7A 2016-05-09 2016-05-09 Multidate SAR image change detection based on binary channels convolutional neural networks Active CN106023154B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201610301686.7A CN106023154B (en) 2016-05-09 2016-05-09 Multidate SAR image change detection based on binary channels convolutional neural networks

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201610301686.7A CN106023154B (en) 2016-05-09 2016-05-09 Multidate SAR image change detection based on binary channels convolutional neural networks

Publications (2)

Publication Number Publication Date
CN106023154A true CN106023154A (en) 2016-10-12
CN106023154B CN106023154B (en) 2019-03-29

Family

ID=57100096

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201610301686.7A Active CN106023154B (en) 2016-05-09 2016-05-09 Multidate SAR image change detection based on binary channels convolutional neural networks

Country Status (1)

Country Link
CN (1) CN106023154B (en)

Cited By (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107154054A (en) * 2017-05-22 2017-09-12 西安电子科技大学 Polarization SAR remote sensing imagery change detection method based on depth confidence network
CN107239795A (en) * 2017-05-19 2017-10-10 西安电子科技大学 SAR image change detecting system and method based on sparse self-encoding encoder and convolutional neural networks
CN107256396A (en) * 2017-06-12 2017-10-17 电子科技大学 Ship target ISAR characteristics of image learning methods based on convolutional neural networks
CN107590782A (en) * 2017-08-21 2018-01-16 西北工业大学 A kind of spissatus minimizing technology of high-resolution optical image based on full convolutional network
CN107944483A (en) * 2017-11-17 2018-04-20 西安电子科技大学 Classification of Multispectral Images method based on binary channels DCGAN and Fusion Features
CN108596108A (en) * 2018-04-26 2018-09-28 中国科学院电子学研究所 Method for detecting change of remote sensing image of taking photo by plane based on the study of triple semantic relation
CN108846829A (en) * 2018-05-23 2018-11-20 平安科技（深圳）有限公司 Diseased region recognition methods and device, computer installation and readable storage medium storing program for executing
CN109086753A (en) * 2018-10-08 2018-12-25 新疆大学 Traffic sign recognition method, device based on binary channels convolutional neural networks
CN109558806A (en) * 2018-11-07 2019-04-02 北京科技大学 The detection method and system of high score Remote Sensing Imagery Change
CN112187266A (en) * 2020-09-29 2021-01-05 中国电子科技集团公司第三十六研究所 Nonlinear correction method and device for analog-to-digital converter and electronic equipment
CN112215834A (en) * 2020-10-22 2021-01-12 深圳大学 Power battery welding quality detection method and system based on convolutional neural network

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103810699A (en) * 2013-12-24 2014-05-21 西安电子科技大学 SAR (synthetic aperture radar) image change detection method based on non-supervision depth nerve network
CN103955702A (en) * 2014-04-18 2014-07-30 西安电子科技大学 SAR image terrain classification method based on depth RBF network
CN104102919A (en) * 2014-07-14 2014-10-15 同济大学 Image classification method capable of effectively preventing convolutional neural network from being overfit
CN104408481A (en) * 2014-12-05 2015-03-11 西安电子科技大学 Deep wavelet neural network-based polarimetric SAR (synthetic aperture radar) image classification method
US20150125049A1 (en) * 2013-11-04 2015-05-07 Facebook, Inc. Systems and methods for facial representation
CN104732243A (en) * 2015-04-09 2015-06-24 西安电子科技大学 SAR target identification method based on CNN
CN104850845A (en) * 2015-05-30 2015-08-19 大连理工大学 Traffic sign recognition method based on asymmetric convolution neural network
CN105320965A (en) * 2015-10-23 2016-02-10 西北工业大学 Hyperspectral image classification method based on spectral-spatial cooperation of deep convolutional neural network

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20150125049A1 (en) * 2013-11-04 2015-05-07 Facebook, Inc. Systems and methods for facial representation
CN103810699A (en) * 2013-12-24 2014-05-21 西安电子科技大学 SAR (synthetic aperture radar) image change detection method based on non-supervision depth nerve network
CN103955702A (en) * 2014-04-18 2014-07-30 西安电子科技大学 SAR image terrain classification method based on depth RBF network
CN104102919A (en) * 2014-07-14 2014-10-15 同济大学 Image classification method capable of effectively preventing convolutional neural network from being overfit
CN104408481A (en) * 2014-12-05 2015-03-11 西安电子科技大学 Deep wavelet neural network-based polarimetric SAR (synthetic aperture radar) image classification method
CN104732243A (en) * 2015-04-09 2015-06-24 西安电子科技大学 SAR target identification method based on CNN
CN104850845A (en) * 2015-05-30 2015-08-19 大连理工大学 Traffic sign recognition method based on asymmetric convolution neural network
CN105320965A (en) * 2015-10-23 2016-02-10 西北工业大学 Hyperspectral image classification method based on spectral-spatial cooperation of deep convolutional neural network

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
FANGSHUN LIAO 等: "An Improved Method in Change Detection of Multitemporal Remote Sensing Image", 《ISCIDE 2013》 *
MAOGUO GONG 等: "Change Detection in Synthetic Aperture Radar Images Based on Deep Neural Networks", 《IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS》 *
YAN CHU 等: "Change Detection of Remote Sensing Image Based on Deep Neural Networks", 《2ND INTERNATIONAL CONFERENCE ON ARTIFICIAL INTELLIGENCE AND INDUSTRIAL ENGINEERING (AIIE2016)》 *
赵姣姣: "基于无监督方法的SAR图像变化检测", 《中国优秀硕士学位论文全文数据库 信息科技辑（月刊）》 *

Cited By (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107239795A (en) * 2017-05-19 2017-10-10 西安电子科技大学 SAR image change detecting system and method based on sparse self-encoding encoder and convolutional neural networks
CN107154054A (en) * 2017-05-22 2017-09-12 西安电子科技大学 Polarization SAR remote sensing imagery change detection method based on depth confidence network
CN107256396A (en) * 2017-06-12 2017-10-17 电子科技大学 Ship target ISAR characteristics of image learning methods based on convolutional neural networks
CN107590782A (en) * 2017-08-21 2018-01-16 西北工业大学 A kind of spissatus minimizing technology of high-resolution optical image based on full convolutional network
CN107590782B (en) * 2017-08-21 2020-05-12 西北工业大学 High-resolution optical image thick cloud removing method based on full convolution network
CN107944483A (en) * 2017-11-17 2018-04-20 西安电子科技大学 Classification of Multispectral Images method based on binary channels DCGAN and Fusion Features
CN107944483B (en) * 2017-11-17 2020-02-07 西安电子科技大学 Multispectral image classification method based on dual-channel DCGAN and feature fusion
CN108596108B (en) * 2018-04-26 2021-02-23 中国科学院电子学研究所 Aerial remote sensing image change detection method based on triple semantic relation learning
CN108596108A (en) * 2018-04-26 2018-09-28 中国科学院电子学研究所 Method for detecting change of remote sensing image of taking photo by plane based on the study of triple semantic relation
CN108846829A (en) * 2018-05-23 2018-11-20 平安科技（深圳）有限公司 Diseased region recognition methods and device, computer installation and readable storage medium storing program for executing
CN108846829B (en) * 2018-05-23 2021-03-23 平安科技（深圳）有限公司 Lesion site recognition device, computer device, and readable storage medium
CN109086753A (en) * 2018-10-08 2018-12-25 新疆大学 Traffic sign recognition method, device based on binary channels convolutional neural networks
CN109558806A (en) * 2018-11-07 2019-04-02 北京科技大学 The detection method and system of high score Remote Sensing Imagery Change
CN109558806B (en) * 2018-11-07 2021-09-14 北京科技大学 Method for detecting high-resolution remote sensing image change
CN112187266A (en) * 2020-09-29 2021-01-05 中国电子科技集团公司第三十六研究所 Nonlinear correction method and device for analog-to-digital converter and electronic equipment
CN112187266B (en) * 2020-09-29 2024-03-08 中国电子科技集团公司第三十六研究所 Nonlinear correction method and device of analog-to-digital converter and electronic equipment
CN112215834A (en) * 2020-10-22 2021-01-12 深圳大学 Power battery welding quality detection method and system based on convolutional neural network
CN112215834B (en) * 2020-10-22 2022-04-01 深圳大学 Power battery welding quality detection method and system based on convolutional neural network

Also Published As

Publication number Publication date
CN106023154B (en) 2019-03-29

Similar Documents

Publication Publication Date Title
CN106023154A (en) Multi-temporal SAR image change detection method based on dual-channel convolutional neural network (CNN)
CN105957086B (en) A kind of method for detecting change of remote sensing image based on optimization neural network model
CN105046277B (en) Robust mechanism study method of the feature significance in image quality evaluation
CN111739075B (en) Deep network lung texture recognition method combining multi-scale attention
CN109003240B (en) Image denoising method based on multi-scale parallel CNN
CN110532894A (en) Remote sensing target detection method based on boundary constraint CenterNet
CN104537647B (en) A kind of object detection method and device
CN107492095A (en) Medical image pulmonary nodule detection method based on deep learning
CN109919108A (en) Remote sensing images fast target detection method based on depth Hash auxiliary network
CN108388896A (en) A kind of licence plate recognition method based on dynamic time sequence convolutional neural networks
CN106203625A (en) A kind of deep-neural-network training method based on multiple pre-training
CN107016406A (en) The pest and disease damage image generating method of network is resisted based on production
CN106886801A (en) A kind of image, semantic dividing method and device
CN106295714A (en) A kind of multi-source Remote-sensing Image Fusion based on degree of depth study
CN111507884A (en) Self-adaptive image steganalysis method and system based on deep convolutional neural network
CN106845529A (en) Image feature recognition methods based on many visual field convolutional neural networks
CN107316066A (en) Image classification method and system based on multi-path convolutional neural networks
CN105528638A (en) Method for grey correlation analysis method to determine number of hidden layer characteristic graphs of convolutional neural network
CN105608446A (en) Video stream abnormal event detection method and apparatus
CN109190643A (en) Based on the recognition methods of convolutional neural networks Chinese medicine and electronic equipment
CN108108751A (en) A kind of scene recognition method based on convolution multiple features and depth random forest
CN108009481A (en) A kind of training method and device of CNN models, face identification method and device
Semenenko et al. Browserbite: Accurate cross-browser testing via machine learning over image features
CN107092883A (en) Object identification method for tracing
CN108447057A (en) SAR image change detection based on conspicuousness and depth convolutional network

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
TR01 Transfer of patent right
TR01 Transfer of patent right

Effective date of registration: 20200313

Address after: 400021 2403, 24 / F, No.11, seventh branch road, Panxi, Jiangbei District, Chongqing

Patentee after: Chongqing Commercial Service Technology Co.,Ltd.

Address before: 710072 Xi'an friendship West Road, Shaanxi, No. 127

Patentee before: Northwestern Polytechnical University