CN109389152A - A kind of fining recognition methods of the vertical pendant object of transmission line of electricity - Google Patents
A kind of fining recognition methods of the vertical pendant object of transmission line of electricity Download PDFInfo
- Publication number
- CN109389152A CN109389152A CN201811004041.2A CN201811004041A CN109389152A CN 109389152 A CN109389152 A CN 109389152A CN 201811004041 A CN201811004041 A CN 201811004041A CN 109389152 A CN109389152 A CN 109389152A
- Authority
- CN
- China
- Prior art keywords
- interest
- image
- transmission line
- fining
- electricity
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 30
- 230000005540 biological transmission Effects 0.000 title claims abstract description 23
- 230000005611 electricity Effects 0.000 title claims abstract description 22
- 230000003044 adaptive effect Effects 0.000 claims abstract description 13
- 239000000284 extract Substances 0.000 claims abstract description 8
- 230000008569 process Effects 0.000 claims description 12
- 239000011159 matrix material Substances 0.000 claims description 11
- 239000004744 fabric Substances 0.000 claims description 3
- 238000001514 detection method Methods 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 3
- 230000009466 transformation Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000007689 inspection Methods 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 208000025274 Lightning injury Diseases 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 235000013350 formula milk Nutrition 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 210000005036 nerve Anatomy 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012876 topography Methods 0.000 description 1
- 230000032258 transport Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
- G06V10/443—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components by matching or filtering
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Image Analysis (AREA)
Abstract
It hangs down the invention discloses a kind of transmission line of electricity and falls the fining recognition methods of object, include the following steps: step 1, read initial background image, initial background image is modeled according to mixed Gauss model, generates background model;Step 2 reads images to be recognized, determines area-of-interest by calculus of finite differences;Step 3 extracts region of interest area image, and carries out adaptive median filter processing;Step 4 identifies vertical pendant object using multiple dimensioned AdaBoost algorithm;The present invention utilizes the methods of mixed Gauss model, calculus of finite differences and perspective transform, extract area-of-interest, and combine multiple dimensioned AdaBoost algorithm, not only solving lack of training samples causes to train bad problem, and it can allow network that can learn to more structure features, to when analyzing image data to be identified, effectively complex background can be split, and significant the accuracy for improving automatic identification.
Description
Technical field
The present invention relates to technical field of image processing, and in particular to a kind of fining identification side of the vertical pendant object of transmission line of electricity
Method.
Background technique
China's NETWORK STRUCTURE PRESERVING POWER SYSTEM is complicated, and energy centre and load center distance are remote, and it is defeated over long distances to generally require high voltage
Electricity.Overhead transmission line the topography and geomorphology through place it is ever-changing, some local natural environments are more severe, in addition wind and rain,
Lightning stroke etc. influences, and is easy to cause transmission line of electricity the dangerous situations such as stranded, foreign matter attachment occur, transports to the safety and stability of electric system
Row causes greatly to threaten.
Currently, manual inspection, unmanned plane inspection or fixed point video acquisition are mainly passed through to the detection of the vertical pendant object of transmission line of electricity,
Transmission line of electricity image to be detected information, then manual analysis image information are obtained, judges and marks.This manual detection mode
The time of consuming and the memory of occupancy be all it is huge, efficiency is lower, and cannot find the problem in time, it is therefore necessary to study
Automatic identification detection technique.
In recent years, target detection technique is applied to by emerging line walking technology using robot vision technology as discipline background
Transmission line safety operation detection in, by integrated application image procossing, pattern-recognition and nerve network system etc. it is a variety of before
Along property theory, the area-of-interest in image is automatically extracted, and then the vertical pendant object on transmission line of electricity is detected or identified.But
It is that due to the background information of transmission line of electricity is complicated, pendant species of hanging down are various etc., the accuracy rate for causing detection to identify is not high.
Summary of the invention
The purpose of the present invention is to overcome the shortcomings of the existing technology with it is insufficient, a kind of transmission line of electricity is provided and is hung down the fine of pendant object
Change recognition methods, this method extracts area-of-interest using the methods of mixed Gauss model, calculus of finite differences and perspective transform, and ties
Multiple dimensioned AdaBoost algorithm is closed, so that automatic identification can be effectively improved when analyzing image data to be identified
Accuracy.
The purpose of the invention is achieved by the following technical solution:
A kind of fining recognition methods of the vertical pendant object of transmission line of electricity, includes the following steps:
Step 1 reads initial background image, is modeled according to mixed Gauss model to initial background image, generates background mould
Type;
Step 2 reads images to be recognized, determines area-of-interest by calculus of finite differences;
Step 3 extracts region of interest area image, and carries out adaptive median filter processing;
Step 4 identifies vertical pendant object using multiple dimensioned AdaBoost algorithm.
Preferably, detailed process is as follows for the step 1:
Initial background image is read first, then by each pixel of initial background image respectively with by K Gauss point
The mixed Gauss model that cloth is constituted models, and generates background model, it may be assumed that
WhereinIndicate the estimated value of the weight coefficient of i-th of Gaussian Profile in moment t mixed Gauss model;η indicates high
This distribution probability density function;xj=[xjB xjG xjR] indicate value of the pixel j in t moment, xjB、xjGAnd xjRRespectively indicate B,
G, the pixel value in tri- channels R;WithRespectively indicate the mean value of i-th of Gaussian Profile in moment t mixed Gauss model
Vector sum covariance matrix.
Preferably, detailed process is as follows for the step 2:
Images to be recognized and background model work are subtracted into operation, and taken absolute value, it may be assumed that | I (xj)-P(xj) |, wherein I (xj) be
The pixel value of images to be recognized;If | I (xj)-P(xj) | > T, then the pixel is point-of-interest, and wherein T is adjustable threshold value;
The minimum rectangle that all point-of-interests finally can be surrounded with one determines area-of-interest.
Preferably, detailed process is as follows for the step 3:
Area-of-interest is split from images to be recognized using perspective transform, and it is emerging to be transformed into upright rectangle sense
Then interesting image is filtered region of interest area image with adaptive median filter algorithm.
Preferably, detailed process is as follows for the step 4:
(1) it collects transmission line of electricity and hangs down and fall the image data of object, construct training input matrix;
(2) Weak Classifier is then cascaded, and more in the multiple dimensioned upper multiple Weak Classifiers of training using training input matrix
Strong classifier is obtained by interative computation on the basis of new Weak Classifier weight;
(3) according to strong classifier, to step 3, treated that image carries out hangs down pendant object identification.
The present invention have compared with prior art it is below the utility model has the advantages that
The present invention extracts area-of-interest, and combine using the methods of mixed Gauss model, calculus of finite differences and perspective transform
Multiple dimensioned AdaBoost algorithm, not only solving lack of training samples causes to train bad problem, and can allow network that can learn
More structure features are practised, to can effectively carry out to complex background when analyzing image data to be identified
Segmentation, and significant the accuracy for improving automatic identification.
Detailed description of the invention
Fig. 1 is flow chart of the invention.
Specific embodiment
Present invention will now be described in further detail with reference to the embodiments and the accompanying drawings, but embodiments of the present invention are unlimited
In this.
The fining recognition methods of pendant object as shown in Figure 1, a kind of transmission line of electricity hangs down, includes the following steps:
Step 1 reads initial background image, is modeled according to mixed Gauss model to initial background image, generates background mould
Type;
Wherein, detailed process is as follows for the step 1:
Initial background image is read first, then by each pixel of initial background image respectively with by K Gauss point
The mixed Gauss model that cloth is constituted models, and is that the color that pixel is presented is considered stochastic variable Ψ, then each
The pixel value of video frame images obtained by moment t=1 ..., T is the sampled value of stochastic variable Ψ, so that background model is generated,
That is:
WhereinIndicate the estimated value of the weight coefficient of i-th of Gaussian Profile in moment t mixed Gauss model;η indicates high
This distribution probability density function;xj=[xjB xjG xjR] indicate value of the pixel j in t moment, xjB、xjGAnd xjRRespectively indicate B,
G, the pixel value in tri- channels R;WithRespectively indicate the mean value of i-th of Gaussian Profile in moment t mixed Gauss model
Vector sum covariance matrix.
Step 2, read images to be recognized, by calculus of finite differences by images to be recognized and background model in certain threshold value
Different zones are marked, so that it is determined that area-of-interest;
Wherein, detailed process is as follows for the step 2:
Images to be recognized and background model work are subtracted into operation, and taken absolute value, it may be assumed that | I (xj)-P(xj) |, wherein I (xj) be
The pixel value of images to be recognized;If | I (xj)-P(xj) | > T, then the pixel is point-of-interest, and wherein T is adjustable threshold value;
The minimum rectangle that all point-of-interests finally can be surrounded with one determines area-of-interest.
Step 3 extracts region of interest area image, and carries out adaptive median filter processing;
Wherein, detailed process is as follows for the step 3:
Area-of-interest is split from images to be recognized using perspective transform, and it is emerging to be transformed into upright rectangle sense
Then interesting image is filtered region of interest area image with adaptive median filter algorithm.
Specifically, perspective transform is interested by four datum marks of the rectangle in images to be recognized and upright rectangle
Determined by four datum marks of image, and four datum marks of upright rectangle image of interest are adjustable;
Transformation for mula of the perspective transform for each group of corresponding position are as follows:
Wherein, x ' and y ' is the datum mark coordinate of upright rectangle image of interest;S is change of scale;X and y is wait know
The datum mark coordinate of rectangle in other image;For perspective transformation matrix;
After calculating perspective transformation matrix, the coordinate that the rectangle in images to be recognized surrounds successively is substituted into, then it is convertible
Obtain upright rectangle image of interest.
The window size S range of adaptive median filter algorithm is adjustable, is [Smin,Smax];Adapt to median filtering algorithm tool
Body step are as follows:
Step A: the intermediate value Z of all pixels value in adaptive median filter algorithm scope is calculatedmed;
Step B: if Zmin<Zmed<Zmax, go to step E;
Step C: increase the window size of adaptive median filter algorithm, if S≤Smax, go to step A;
Step D: output Zmed;
Step E: if Zmin<Zxy<Zmax, export Zxy;
Step F: output Zmed;
Wherein ZminFor the minimum value of all pixels value in adaptive median filter algorithm scope;ZmaxFor adaptive intermediate value
The maximum value of all pixels value in filtering algorithm scope;ZxyThe value of a pixel is arranged for y row xth in image.
Step 4 is trained study using image data of the multiple dimensioned AdaBoost algorithm to the vertical pendant object of transmission line of electricity,
It by training multiple Weak Classifiers on multiple images scale, then cascades Weak Classifier and constitutes strong classifier, thus to vertical pendant
Object carries out fining identification.
Wherein, detailed process is as follows for the step 4:
(1) it collects transmission line of electricity and hangs down and fall the image data of object, construct training input matrix;
(2) Weak Classifier is then cascaded, and more in the multiple dimensioned upper multiple Weak Classifiers of training using training input matrix
Strong classifier is obtained by interative computation on the basis of new Weak Classifier weight;
(3) according to strong classifier, to step 3, treated that image carries out hangs down pendant object identification.
The present invention includes reading initial background image, is modeled according to mixed Gauss model to initial background image, generates back
Scape model, then reads images to be recognized, determines area-of-interest by calculus of finite differences;Region of interest area image is extracted later, and
Carry out adaptive median filter processing;Vertical pendant object is identified using multiple dimensioned AdaBoost algorithm.
The present invention extracts area-of-interest, and combine using the methods of mixed Gauss model, calculus of finite differences and perspective transform
Multiple dimensioned AdaBoost algorithm, not only solving lack of training samples causes to train bad problem, and can allow network that can learn
More structure features are practised, to can effectively carry out to complex background when analyzing image data to be identified
Segmentation, and significant the accuracy for improving automatic identification.
Above-mentioned is the preferable embodiment of the present invention, but embodiments of the present invention are not limited by the foregoing content,
His any changes, modifications, substitutions, combinations, simplifications made without departing from the spirit and principles of the present invention, should be
The substitute mode of effect, is included within the scope of the present invention.
Claims (5)
- The fining recognition methods of pendant object 1. a kind of transmission line of electricity hangs down, which is characterized in that include the following steps:Step 1 reads initial background image, is modeled according to mixed Gauss model to initial background image, generates background model;Step 2 reads images to be recognized, determines area-of-interest by calculus of finite differences;Step 3 extracts region of interest area image, and carries out adaptive median filter processing;Step 4 identifies vertical pendant object using multiple dimensioned AdaBoost algorithm.
- The fining recognition methods of pendant object 2. transmission line of electricity according to claim 1 hangs down, which is characterized in that the step 1 Detailed process is as follows:Initial background image is read first, then by each pixel of initial background image respectively with by K Gaussian Profile structure At mixed Gauss model model, generate background model, it may be assumed thatWhereinIndicate the estimated value of the weight coefficient of i-th of Gaussian Profile in moment t mixed Gauss model;η indicates Gauss point Cloth probability density function;xj=[xjB xjG xjR] indicate value of the pixel j in t moment, xjB、xjGAnd xjRRespectively indicate B, G, R The pixel value in three channels;WithRespectively indicate the mean value of i-th of Gaussian Profile in moment t mixed Gauss model to Amount and covariance matrix.
- The fining recognition methods of pendant object 3. transmission line of electricity according to claim 1 hangs down, which is characterized in that the step 2 Detailed process is as follows:Images to be recognized and background model work are subtracted into operation, and taken absolute value, it may be assumed that | I (xj)-P(xj) |, wherein I (xj) it is wait know The pixel value of other image;If | I (xj)-P(xj) | > T, then the pixel is point-of-interest, and wherein T is adjustable threshold value;Finally The minimum rectangle that all point-of-interests can be surrounded with one determines area-of-interest.
- The fining recognition methods of pendant object 4. transmission line of electricity according to claim 1 hangs down, which is characterized in that the step 3 Detailed process is as follows:Area-of-interest is split from images to be recognized using perspective transform, and is transformed into upright rectangle figure interested Then picture is filtered region of interest area image with adaptive median filter algorithm.
- The fining recognition methods of pendant object 5. transmission line of electricity according to claim 1 hangs down, which is characterized in that the step 4 Detailed process is as follows:(1) it collects transmission line of electricity and hangs down and fall the image data of object, construct training input matrix;(2) Weak Classifier is then cascaded in the multiple dimensioned upper multiple Weak Classifiers of training using training input matrix, and weak updating Strong classifier is obtained by interative computation on the basis of classifier weight;(3) according to strong classifier, to step 3, treated that image carries out hangs down pendant object identification.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811004041.2A CN109389152B (en) | 2018-08-30 | 2018-08-30 | Refined identification method for power transmission line falling object |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811004041.2A CN109389152B (en) | 2018-08-30 | 2018-08-30 | Refined identification method for power transmission line falling object |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109389152A true CN109389152A (en) | 2019-02-26 |
CN109389152B CN109389152B (en) | 2022-05-13 |
Family
ID=65418689
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811004041.2A Expired - Fee Related CN109389152B (en) | 2018-08-30 | 2018-08-30 | Refined identification method for power transmission line falling object |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109389152B (en) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109977962A (en) * | 2019-03-21 | 2019-07-05 | 国网山东省电力公司经济技术研究院 | A kind of Cable's Fault hidden danger automatic identifying method and system |
CN113486865A (en) * | 2021-09-03 | 2021-10-08 | 国网江西省电力有限公司电力科学研究院 | Power transmission line suspended foreign object target detection method based on deep learning |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20100266207A1 (en) * | 2009-04-21 | 2010-10-21 | ArcSoft ( Hangzhou) Multimedia Technology Co., Ltd | Focus enhancing method for portrait in digital image |
CN101923652A (en) * | 2010-07-23 | 2010-12-22 | 华中师范大学 | Pornographic picture identification method based on joint detection of skin colors and featured body parts |
US20100322534A1 (en) * | 2009-06-09 | 2010-12-23 | Colorado State University Research Foundation | Optimized correlation filters for signal processing |
US20120002868A1 (en) * | 2010-07-01 | 2012-01-05 | Loui Alexander C | Method for fast scene matching |
CN103473791A (en) * | 2013-09-10 | 2013-12-25 | 惠州学院 | Method for automatically recognizing abnormal velocity event in surveillance video |
CN103646241A (en) * | 2013-12-30 | 2014-03-19 | 中国科学院自动化研究所 | Real-time taxi identification method based on embedded system |
CN104463195A (en) * | 2014-11-08 | 2015-03-25 | 沈阳工业大学 | Printing style digital recognition method based on template matching |
CN107103303A (en) * | 2017-04-27 | 2017-08-29 | 昆明理工大学 | A kind of pedestrian detection method based on GMM backgrounds difference and union feature |
CN107133974A (en) * | 2017-06-02 | 2017-09-05 | 南京大学 | The vehicle type classification method that Gaussian Background modeling is combined with Recognition with Recurrent Neural Network |
-
2018
- 2018-08-30 CN CN201811004041.2A patent/CN109389152B/en not_active Expired - Fee Related
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20100266207A1 (en) * | 2009-04-21 | 2010-10-21 | ArcSoft ( Hangzhou) Multimedia Technology Co., Ltd | Focus enhancing method for portrait in digital image |
US20100322534A1 (en) * | 2009-06-09 | 2010-12-23 | Colorado State University Research Foundation | Optimized correlation filters for signal processing |
US20120002868A1 (en) * | 2010-07-01 | 2012-01-05 | Loui Alexander C | Method for fast scene matching |
CN101923652A (en) * | 2010-07-23 | 2010-12-22 | 华中师范大学 | Pornographic picture identification method based on joint detection of skin colors and featured body parts |
CN103473791A (en) * | 2013-09-10 | 2013-12-25 | 惠州学院 | Method for automatically recognizing abnormal velocity event in surveillance video |
CN103646241A (en) * | 2013-12-30 | 2014-03-19 | 中国科学院自动化研究所 | Real-time taxi identification method based on embedded system |
CN104463195A (en) * | 2014-11-08 | 2015-03-25 | 沈阳工业大学 | Printing style digital recognition method based on template matching |
CN107103303A (en) * | 2017-04-27 | 2017-08-29 | 昆明理工大学 | A kind of pedestrian detection method based on GMM backgrounds difference and union feature |
CN107133974A (en) * | 2017-06-02 | 2017-09-05 | 南京大学 | The vehicle type classification method that Gaussian Background modeling is combined with Recognition with Recurrent Neural Network |
Non-Patent Citations (1)
Title |
---|
王相海: "基于感兴趣区域AdaBoost分类器的视频车辆检测研究", 《辽宁师范大学学报(自然科学版)》 * |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109977962A (en) * | 2019-03-21 | 2019-07-05 | 国网山东省电力公司经济技术研究院 | A kind of Cable's Fault hidden danger automatic identifying method and system |
CN109977962B (en) * | 2019-03-21 | 2022-03-15 | 国网山东省电力公司经济技术研究院 | Method and system for automatically identifying fault hidden danger of optical cable |
CN113486865A (en) * | 2021-09-03 | 2021-10-08 | 国网江西省电力有限公司电力科学研究院 | Power transmission line suspended foreign object target detection method based on deep learning |
Also Published As
Publication number | Publication date |
---|---|
CN109389152B (en) | 2022-05-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109522793B (en) | Method for detecting and identifying abnormal behaviors of multiple persons based on machine vision | |
CN107563381B (en) | Multi-feature fusion target detection method based on full convolution network | |
CN110059694A (en) | The intelligent identification Method of lteral data under power industry complex scene | |
CN114022432B (en) | Insulator defect detection method based on improved yolov5 | |
CN109101865A (en) | A kind of recognition methods again of the pedestrian based on deep learning | |
CN106503742B (en) | A kind of visible images insulator recognition methods | |
CN105740945B (en) | A kind of people counting method based on video analysis | |
CN106022345B (en) | A kind of high voltage isolator state identification method based on Hough forest | |
CN104050481B (en) | Multi-template infrared image real-time pedestrian detection method combining contour feature and gray level | |
CN109446922B (en) | Real-time robust face detection method | |
CN107092884B (en) | Rapid coarse-fine cascade pedestrian detection method | |
CN107610128A (en) | The method for inspecting and device of a kind of oil level indicator | |
CN107392237B (en) | Cross-domain foundation cloud picture classification method based on migration visual information | |
CN101908153B (en) | Method for estimating head postures in low-resolution image treatment | |
CN105138983B (en) | The pedestrian detection method divided based on weighting block model and selective search | |
CN113409252B (en) | Obstacle detection method for overhead transmission line inspection robot | |
CN106485273A (en) | A kind of method for detecting human face based on HOG feature and DNN grader | |
CN105069745A (en) | face-changing system based on common image sensor and enhanced augmented reality technology and method | |
CN109359549A (en) | A kind of pedestrian detection method based on mixed Gaussian and HOG_LBP | |
CN104794449A (en) | Gait energy image acquisition method based on human body HOG (histogram of oriented gradient) features and identity identification method | |
CN109117717A (en) | A kind of city pedestrian detection method | |
CN105426924A (en) | Scene classification method based on middle level features of images | |
CN109389152A (en) | A kind of fining recognition methods of the vertical pendant object of transmission line of electricity | |
CN110188646B (en) | Human ear identification method based on fusion of gradient direction histogram and local binary pattern | |
CN111241979A (en) | Real-time obstacle detection method based on image feature calibration |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20220513 |
|
CF01 | Termination of patent right due to non-payment of annual fee |