CN105184286A - Vehicle detection method and detection device - Google Patents
Vehicle detection method and detection device Download PDFInfo
- Publication number
- CN105184286A CN105184286A CN201510685237.2A CN201510685237A CN105184286A CN 105184286 A CN105184286 A CN 105184286A CN 201510685237 A CN201510685237 A CN 201510685237A CN 105184286 A CN105184286 A CN 105184286A
- Authority
- CN
- China
- Prior art keywords
- detection window
- unique point
- unit
- feature
- 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.)
- Pending
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
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/19—Recognition using electronic means
- G06V30/192—Recognition using electronic means using simultaneous comparisons or correlations of the image signals with a plurality of references
- G06V30/194—References adjustable by an adaptive method, e.g. learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/08—Detecting or categorising vehicles
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Databases & Information Systems (AREA)
- Image Analysis (AREA)
- Traffic Control Systems (AREA)
Abstract
The embodiment of the invention provides a vehicle detection method, which comprises the following steps: obtaining a to-be-detected image including vehicles; determining M detection windows in the to-be-detected image; carrying out feature point marking on images in regions in which the M detection windows are located to obtain N feature points; carrying out aligning treatment on the N feature points; carrying out feature extraction employing each feature point in N feature points after aligning treatment as the center; forming P feature vectors by the extracted features; carrying out classifying treatment on the P feature vectors to eliminate K detection windows in the M detection windows; and taking the regions corresponding to M to K detection windows, except for the Kth detection window in the M detection windows as target regions. The embodiment of the invention further provides a detection device. Through the vehicle detection method and the detection device, the number of error windows can be reduced, so that the vehicle detection accuracy is improved.
Description
Technical field
The embodiment of the present invention relates to intelligent monitoring technology field, is specifically related to a kind of vehicle checking method and pick-up unit.
Background technology
Vehicle testing techniques is an important component part of intelligent transportation system, and the quality of its testing result directly has influence on locating and tracking and the subsequent processes such as identification and behavior understanding of target.
In prior art, although the vehicle testing techniques in intelligent transportation system is developed rapidly, but, on the one hand, even if due to the automobile video frequency image that obtains on Different periods, different sections of highway illumination, angle, block and vary and same section, the video image collection change that the same period obtains also is very greatly; On the other hand, the kind of vehicle own is extensive, and the outward appearance in class and between class and difference in size significantly, thus, cause vehicle detection accurately lower, reduce Consumer's Experience.
Summary of the invention
Embodiments provide a kind of vehicle checking method and pick-up unit, to improving the accuracy rate of vehicle detection.
Embodiment of the present invention first aspect provides a kind of vehicle checking method, comprising:
Obtain the image to be detected comprising vehicle;
Determine M detection window in described image to be detected, described M be greater than 1 integer;
Carry out unique point mark to the image in described M detection window region and obtain N number of unique point, wherein, described M and described N is the integer being greater than 1;
Registration process is carried out to described N number of unique point;
Feature extraction is carried out centered by each unique point in N number of unique point after described registration process;
Feature composition P the proper vector extracted described, wherein, described P be greater than 1 integer;
A described P proper vector is carried out classification process to get rid of K detection window in a described M detection window, wherein, described K is the integer being more than or equal to 1 and being less than described M;
Using region corresponding for M-K detection window in a described M detection window except a described K detection window as target area.
In conjunction with the first aspect of the embodiment of the present invention, in the first possible embodiment of first aspect, described M the detection window determined in described image to be detected comprises:
The cascade detectors preset is adopted to detect to described image to be detected M the detection window obtaining described image to be detected.
In conjunction with the first aspect of the embodiment of the present invention, in the embodiment that the second of first aspect is possible, describedly unique point mark is carried out to the image in described M detection window region obtain N number of unique point and comprise:
Meanshape unique point mark is carried out to a described M detection window, to obtain N number of unique point.
In conjunction with the first aspect of the embodiment of the present invention, in the third possible embodiment of first aspect, describedly registration process carried out to described N number of unique point comprise:
The characteristic matching function of training in advance is adopted to carry out feature extraction to described N number of unique point, to obtain N number of partial binary feature;
The overall linear regression model (LRM) of training in advance is adopted to carry out overall linear regression processing to described N number of partial binary feature.
In conjunction with the first aspect of the embodiment of the present invention, in the 4th kind of possible embodiment of first aspect, described a described P proper vector is carried out classification process comprise with K the detection window got rid of in a described M detection window:
Scale invariant features transform SIFT is carried out to a described P unique point, to obtain P SIFT feature point;
By described P SIFT feature point composition target feature vector;
Default sorter is adopted to carry out classification process to get rid of K detection window in a described M detection window to described target feature vector.
Embodiment of the present invention second aspect provides a kind of pick-up unit, comprising:
Acquiring unit, for obtaining the image to be detected comprising vehicle;
First determining unit, for determining M detection window in the image to be detected that described acquiring unit gets, described M be greater than 1 integer;
Indexing unit, carry out unique point mark for the image in M detection window region determining described first determining unit and obtain N number of unique point, wherein, described M and described N is the integer being greater than 1;
Alignment unit, carries out registration process for the N number of unique point marked described indexing unit;
First extraction unit, for carrying out feature extraction centered by each unique point in the N number of unique point after described alignment unit registration process;
First component units, for described first extraction unit is extracted feature composition P proper vector, wherein, described P be greater than 1 integer;
Taxon, for P proper vector of described first component units composition is carried out classification process to get rid of K detection window in a described M detection window, wherein, described K is the integer being more than or equal to 1 and being less than described M;
Second determining unit, for using region corresponding for M-K detection window in a described M detection window except K the detection window that described taxon is got rid of as target area.
In conjunction with the second aspect of the embodiment of the present invention, in the first possible embodiment of second aspect, described first determining unit specifically for:
The cascade detectors preset is adopted to detect to described image to be detected M the detection window obtaining described image to be detected.
In conjunction with the second aspect of the embodiment of the present invention, in the embodiment that the second of second aspect is possible, described indexing unit specifically for:
Meanshape unique point mark is carried out to a described M detection window, to obtain N number of unique point.
In conjunction with the second aspect of the embodiment of the present invention, in the third possible embodiment of second aspect, described alignment unit comprises:
Second extraction unit, for adopting the characteristic matching function of training in advance to carry out feature extraction to N number of unique point that described indexing unit marks, to obtain N number of partial binary feature;
Processing unit, carries out overall linear regression processing for adopting the overall linear regression model (LRM) of training in advance to N number of partial binary feature that described second extraction unit extracts.
In conjunction with the second aspect of the embodiment of the present invention, in the 4th kind of possible embodiment of second aspect, described taxon comprises:
Converter unit, for carrying out Scale invariant features transform SIFT, to obtain P SIFT feature point to P unique point of described first component units composition;
Second component units, for converting P the SIFT feature point composition target feature vector obtained by described converter unit;
Classification subelement, carries out classification process to get rid of K detection window in a described M detection window for adopting the target feature vector of default sorter to described second component units composition.
Adopt the embodiment of the present invention, there is following beneficial effect:
The image to be detected comprising vehicle is obtained by the embodiment of the present invention; Determine M detection window in described image to be detected, described M be greater than 1 integer; Carry out unique point mark to the image in described M detection window region and obtain N number of unique point, wherein, described M and described N is the integer being greater than 1; Registration process is carried out to described N number of unique point; Feature extraction is carried out centered by each unique point in N number of unique point after described registration process; Feature composition P the proper vector extracted described, wherein, described P be greater than 1 integer; A described P proper vector is carried out classification process to get rid of K detection window in a described M detection window, wherein, described K is the integer being more than or equal to 1 and being less than described M; Using region corresponding for M-K detection window in a described M detection window except a described K detection window as target area.Implement, in the process of the embodiment of the present invention, the number of wrong window can be reduced, thus, improve the accuracy rate of vehicle detection.
Term " first ", " second ", " the 3rd " and " the 4th " etc. in instructions of the present invention and claims and described accompanying drawing are for distinguishing different object, instead of for describing particular order.In addition, term " comprises " and " having " and their any distortion, and intention is to cover not exclusive comprising.Such as contain the process of series of steps or unit, method, system, product or equipment and be not defined in the step or unit listed, but also comprise the step or unit do not listed alternatively, or also comprise alternatively for other intrinsic step of these processes, method, product or equipment or unit.
The pick-up unit of vehicle detection described by the embodiment of the present invention can comprise video matrix, watch-dog etc., and said apparatus is only citing, and non exhaustive, including but not limited to above-mentioned detection device.
Refer to Fig. 1, the first embodiment schematic flow sheet of the method for a kind of target detection provided for the embodiment of the present invention.The method of the target detection described in the present embodiment, comprises the following steps:
S101, obtain and comprise the image to be detected of vehicle.
In specific implementation, image to be detected can be an independent image, or image to be detected can be a frame or the multiple image of video, wherein, can comprise one or more vehicle in image to be detected.Alternatively, according to different mode classifications, vehicle can be divided into different classes.This vehicle can include but are not limited to: compact car, micro-vehicle, compact vehicle, medium vehicle, senior vehicle, deluxe carmodel, three-box car type, CDV vehicle, MPV vehicle, SUV etc.This vehicle also can be: Audi Q7, BMW X3, Audi A8, Audi A4, Audi A4L etc.Again such as, such as the car of east wind class just has 25 classes: as east wind well-to-do level-K07-2006, east wind well-to-do level-K17-2009, east wind well-to-do level-V07S-2011, east wind well-to-do level-V29-2012, east wind well-to-do level-scene-2014, east wind demeanour-Oudin-2007, east wind demeanour-general visitor-2010, east wind demeanour-Yumsun-2007, east wind demeanour-Yumsun-2009, east wind demeanour-sharp black horse pick up-2007, east wind demeanour-sharp black horse pick up-2013, east wind bearing-A60-2012 etc.
S102, M the detection window determined in described image to be detected, described M be greater than 1 integer.
In specific implementation, M the detection window that pick-up unit can adopt default cascade detectors to treat detected image to carry out detecting to obtain this image to be detected.
Such as, a cascade detectors can be used to carry out detection to obtain detection window to training dataset, and this training dataset is the image in image library, guarantees a higher vehicle detection accuracy rate by arranging some threshold values.Preferably, cascade detectors can be as follows:
Wherein, C
iwhat represent is i-th Weak Classifier.What x represented is proper vector (as HOG feature), f representative classification score.C
ican export a classification results according to the sorting technique of self to x, be such as a car or be not a car, and f can export a corresponding threshold value, be less than the situation of corresponding threshold value at any one f, sample corresponding to proper vector just can be rejected.Further, be not that the window of a car will be rejected after the judgement through front several Weak Classifier usually, be not used as judgement below, thus, speed is very fast.
S103, the image in described M detection window region carried out to unique point mark and obtain N number of unique point, wherein, described M and described N is the integer being greater than 1.
In specific implementation, pick-up unit can adopt meanshape to carry out initialization to M detection window.Such as, window initialization 1 shape that pick-up unit detects cascade detectors, this shape is the unique point of a group echo, such as 27 points in vehicle window angle, vehicle body angle, door, car light etc.
S103, registration process is carried out to described N number of unique point.
In specific implementation, pick-up unit can carry out registration process to N number of unique point.Alternatively, pick-up unit can adopt the characteristic matching function of training in advance to carry out feature extraction to N number of unique point, to obtain N number of partial binary feature; Then, the overall linear regression model (LRM) of training in advance is adopted to carry out overall linear regression processing to this N number of partial binary feature.
Illustrate down, first, utilize characteristic matching function phi
tto N number of feature point extraction partial binary feature.Characteristic matching function is made up of one group of local feature adaptation function, following formula:
φ
t＝[φ
t 1,φ
t 2,...,φ
t L]
In model training, study φ
t lregressive object be the real shape increment demarcated
wherein, l is the integer being more than or equal to L, according to following formula to φ
t llearn:
Wherein, i travels through all training samples, π
lfrom vector
extract two elements (2l-1,2l),
2 dimension side-play amounts of l reference point in i-th training sample demarcated.
Then, overall linear regression w is learnt
tby local random forest after having learnt characteristic matching function, we not only obtain binary features φ
t l, have also been obtained local regression and export w
t l.But this method abandons these local of acquiring and exports w
t l, but to connect these binary features be a global characteristics mapping function φ
t, learn an overall linear mapping w by the objective function minimized below simultaneously
t:
Wherein,
regressive object,
about to w
tregular terms, λ is controlling elements, for controlling the intensity of regular terms.Because when characteristic dimension is higher, regularization is necessary.Because above-mentioned feature processed is highly sparse, two coordinate descent can be used to process so large-scale sparse linear systems.
Feature extraction is carried out centered by each unique point in S104, N number of unique point after described registration process.
In specific implementation, pick-up unit can carry out feature extraction to centered by the unique point i in the N number of unique point after registration process, and wherein, unique point i is one in N number of unique point.Such as, pick-up unit can carry out Scale invariant features transform (English: Scale-invariantfeaturetransform, abbreviation: SIFT), thus can obtain multiple SIFT feature to the arbitrary unique point in the N number of unique point after registration process.
S105, feature composition P the proper vector extracted described, wherein, described P be greater than 1 integer.
In specific implementation, pick-up unit can will obtain feature composition P proper vector, this P be greater than 1 integer.
S106, a described P proper vector carried out classification process to get rid of K detection window in a described M detection window, wherein, described K is the integer being more than or equal to 1 and being less than described M.
S107, using region corresponding for M-K detection window in a described M detection window except a described K detection window as target area.
In specific implementation, after eliminating K detection window in M detection window, M-K remaining detection window, closer to the position of vehicle, more accurately can be determined the position of vehicle, can identify the model of this vehicle according to the vehicle detected further according to remaining M-K.
The image to be detected comprising vehicle is obtained by the embodiment of the present invention; Determine M detection window in described image to be detected, described M be greater than 1 integer; Carry out unique point mark to the image in described M detection window region and obtain N number of unique point, wherein, described M and described N is the integer being greater than 1; Registration process is carried out to described N number of unique point; Feature extraction is carried out centered by each unique point in N number of unique point after described registration process; Feature composition P the proper vector extracted described, wherein, described P be greater than 1 integer; A described P proper vector is carried out classification process to get rid of K detection window in a described M detection window, wherein, described K is the integer being more than or equal to 1 and being less than described M; Using region corresponding for M-K detection window in a described M detection window except a described K detection window as target area.Implement, in the process of the embodiment of the present invention, the number of wrong window can be reduced, thus, improve the accuracy rate of vehicle detection.
Referring to Fig. 2, is the second embodiment schematic flow sheet of a kind of vehicle checking method that the embodiment of the present invention provides.Vehicle checking method described in the present embodiment, comprises the following steps:
S201, obtain and comprise the image to be detected of vehicle.
In specific implementation, image to be detected can be an independent image, or image to be detected can be a frame or the multiple image of video, wherein, can comprise one or more vehicle in image to be detected.Alternatively, according to different mode classifications, vehicle can be divided into different classes.This vehicle can include but are not limited to: compact car, micro-vehicle, compact vehicle, medium vehicle, senior vehicle, deluxe carmodel, three-box car type, CDV vehicle, MPV vehicle, SUV etc.This vehicle also can be: Audi Q7, BMW X3, Audi A8, Audi A4, Audi A4L etc.Again such as, such as the car of east wind class just has 25 classes: as east wind well-to-do level-K07-2006, east wind well-to-do level-K17-2009, east wind well-to-do level-V07S-2011, east wind well-to-do level-V29-2012, east wind well-to-do level-scene-2014, east wind demeanour-Oudin-2007, east wind demeanour-general visitor-2010, east wind demeanour-Yumsun-2007, east wind demeanour-Yumsun-2009, east wind demeanour-sharp black horse pick up-2007, east wind demeanour-sharp black horse pick up-2013, east wind bearing-A60-2012 etc.
S202, M the detection window determined in described image to be detected, described M be greater than 1 integer.
In specific implementation, the cascade detectors that employing is preset treats M the detection window that detected image carries out detecting to obtain this image to be detected.
Such as, a cascade detectors can being used to carry out detection to obtain detection window to training dataset, guaranteeing a higher vehicle detection accuracy rate by arranging some threshold values.Preferably, cascade detectors can be as follows:
Wherein, C
iwhat represent is i-th Weak Classifier.What x represented is proper vector (as HOG feature), f representative classification score.C
ican export a classification results according to the sorting technique of self to x, be such as a car or be not a car, and f can export a corresponding threshold value, be less than the situation of corresponding threshold value at any one f, sample corresponding to proper vector just can be rejected.Further, be not that the window of a car will be rejected after the judgement through front several Weak Classifier usually, be not used as judgement below, thus, speed is very fast.
S203, the image in described M detection window region carried out to unique point mark and obtain N number of unique point, wherein, described M and described N is the integer being greater than 1.
In specific implementation, pick-up unit can adopt meanshape to carry out initialization to M detection window.Such as, window initialization 1 shape that pick-up unit detects cascade detectors, this shape is the unique point of a group echo, such as 27 points in vehicle window angle, vehicle body angle, door, car light etc.
S204, registration process is carried out to described N number of unique point.
In specific implementation, pick-up unit can carry out registration process to N number of unique point.Alternatively, pick-up unit can adopt the N number of unique point of characteristic matching function of training in advance to carry out feature extraction, to obtain N number of partial binary feature; Then, the overall linear regression model (LRM) of training in advance is adopted to carry out overall linear regression processing to this N number of partial binary feature.
Feature extraction is carried out centered by each unique point in S205, N number of unique point after described registration process.
In specific implementation, pick-up unit can carry out feature extraction to centered by the unique point i in the N number of unique point after registration process, and wherein, unique point i is one in N number of unique point.Such as, pick-up unit can carry out Scale invariant features transform to the arbitrary unique point in the N number of unique point after registration process, thus can obtain multiple SIFT feature.
S206, feature composition P the proper vector extracted described, wherein, described P be greater than 1 integer.
S207, Scale invariant features transform SIFT is carried out to a described P unique point, to obtain P SIFT feature point.
In specific implementation, pick-up unit can carry out Scale invariant features transform in alignment P unique point, thus can obtain P SIFT feature.
S208, by described P SIFT feature point composition target feature vector.
S209, employing are preset sorter and are carried out classification process to get rid of K detection window in a described M detection window to described target feature vector, and wherein, described K is the integer being more than or equal to 1 and being less than described M.
In specific implementation, default sorter can be the cascade classifier that multiple sorter carries out cascade, and the cascade classifier that alternatively, default sorter can be histograms of oriented gradients, support vector machine forms and default sorter can be convolutional neural networks.Preferably, default sorter can be sorter after SVM.Pick-up unit can adopt default sorter to carry out classification process to get rid of K detection window in a described M detection window to P target feature vector, and wherein, K is the integer being more than or equal to 1 and being less than M.
S210, using region corresponding for M-K detection window in a described M detection window except a described K detection window as target area.
In specific implementation, after eliminating K detection window in M detection window, M-K remaining detection window, closer to the position of vehicle, more accurately can be determined the position of vehicle, can identify the model of this vehicle according to the vehicle detected further according to remaining M-K.
The image to be detected comprising vehicle is obtained by the embodiment of the present invention; Determine M detection window in described image to be detected, described M be greater than 1 integer; Carry out unique point mark to the image in described M detection window region and obtain N number of unique point, wherein, described M and described N is the integer being greater than 1; Registration process is carried out to described N number of unique point; Feature extraction is carried out centered by each unique point in N number of unique point after described registration process; Feature composition P the proper vector extracted described, wherein, described P be greater than 1 integer; Scale invariant features transform SIFT is carried out to a described P unique point, to obtain P SIFT feature point; By described P SIFT feature point composition target feature vector; Adopt default sorter to carry out classification process to get rid of K detection window in a described M detection window to described target feature vector, wherein, described K is the integer being more than or equal to 1 and being less than described M; Using region corresponding for M-K detection window in a described M detection window except a described K detection window as target area.Implement, in the process of the embodiment of the present invention, the number of wrong window can be reduced, thus, improve the accuracy rate of vehicle detection.
Referring to Fig. 3, is the first example structure schematic diagram of a kind of pick-up unit that the embodiment of the present invention provides.Pick-up unit described in the present embodiment, comprise: acquiring unit 301, first determining unit 302, indexing unit 303, alignment unit 304, first extraction unit 305, first component units 306, taxon 307 and the second determining unit 308, specific as follows:
Acquiring unit 301, for obtaining the image to be detected comprising vehicle.
First determining unit 302, for determining M detection window in the image to be detected that described acquiring unit 301 gets, described M be greater than 1 integer.
As a kind of possible embodiment, the first determining unit 302 specifically for:
The cascade detectors preset is adopted to detect to described image to be detected M the detection window obtaining described image to be detected.
Indexing unit 303, carry out unique point mark for the image in M detection window region determining described first determining unit 302 and obtain N number of unique point, wherein, described M and described N is the integer being greater than 1.
As a kind of possible embodiment, indexing unit 303 specifically for:
Meanshape unique point mark is carried out to a described M detection window, to obtain N number of unique point.
Alignment unit 304, carries out registration process for the N number of unique point marked described indexing unit 303.
First extraction unit 305, for carrying out feature extraction centered by each unique point in the N number of unique point after described alignment unit 304 registration process.
First component units 306, for described first extraction unit 305 is extracted feature composition P proper vector, wherein, described P be greater than 1 integer.
Taxon 307, P the proper vector for being formed by described first component units 306 carries out classification process to get rid of K detection window in a described M detection window, and wherein, described K is the integer being more than or equal to 1 and being less than described M.
Second determining unit 308, for using region corresponding for M-K detection window in a described M detection window except K the detection window that described taxon 307 is got rid of as target area.
As a kind of possible embodiment, as shown in Figure 3 b, the alignment unit 304 of the pick-up unit described in Fig. 3 a can comprise: the second extraction unit 3041 and processing unit 3042, specific as follows:
Second extraction unit 3041, for adopting the characteristic matching function of training in advance to carry out feature extraction to N number of unique point that described indexing unit 303 marks, to obtain N number of partial binary feature.
Processing unit 3042, carries out overall linear regression processing for adopting the overall linear regression model (LRM) of training in advance to N number of partial binary feature that described second extraction unit 3041 extracts.
As a kind of possible embodiment, as shown in Figure 3 c, the taxon 307 of the pick-up unit described in Fig. 3 a comprises:
Converter unit 3071, carries out Scale invariant features transform SIFT, to obtain P SIFT feature point for P the unique point formed described first component units 306.
Second component units 3072, for converting P the SIFT feature point composition target feature vector obtained by described converter unit 3071.
Classification subelement 3073, carries out classification process to get rid of K detection window in a described M detection window for adopting default sorter to the target feature vector that described second component units 3072 forms.
The image to be detected comprising vehicle can be obtained by the pick-up unit described by the embodiment of the present invention; Determine M detection window in described image to be detected, described M be greater than 1 integer; Carry out unique point mark to the image in described M detection window region and obtain N number of unique point, wherein, described M and described N is the integer being greater than 1; Registration process is carried out to described N number of unique point; Feature extraction is carried out centered by each unique point in N number of unique point after described registration process; Feature composition P the proper vector extracted described, wherein, described P be greater than 1 integer; A described P proper vector is carried out classification process to get rid of K detection window in a described M detection window, wherein, described K is the integer being more than or equal to 1 and being less than described M; Using region corresponding for M-K detection window in a described M detection window except a described K detection window as target area.Implement, in the process of the embodiment of the present invention, the number of wrong window can be reduced, thus, improve the accuracy rate of vehicle detection.
Referring to Fig. 4, is the second example structure schematic diagram of a kind of pick-up unit that the embodiment of the present invention provides.Pick-up unit described in the present embodiment, comprising: at least one input equipment 1000; At least one output device 2000; At least one processor 3000, such as CPU; With storer 4000, above-mentioned input equipment 1000, output device 2000, processor 3000 are connected by bus 5000 with storer 4000.
Wherein, above-mentioned input equipment 1000 specifically can be physical button, touch-screen or contact panel.
Above-mentioned output device 2000 specifically can be display screen.
Above-mentioned storer 4000 can be high-speed RAM storer, also can be non-labile storer (non-volatilememory), such as magnetic disk memory.Above-mentioned storer 4000 is for storing batch processing code, and above-mentioned input equipment 1000, output device 2000 and processor 3000, for calling the program code stored in storer 4000, perform and operate as follows:
Above-mentioned processor 3000, for:
Obtain the image to be detected comprising vehicle;
Determine M detection window in described image to be detected, described M be greater than 1 integer;
Carry out unique point mark to the image in described M detection window region and obtain N number of unique point, wherein, described M and described N is the integer being greater than 1;
Registration process is carried out to described N number of unique point;
Feature extraction is carried out centered by each unique point in N number of unique point after described registration process;
Feature composition P the proper vector extracted described, wherein, described P be greater than 1 integer;
A described P proper vector is carried out classification process to get rid of K detection window in a described M detection window, wherein, described K is the integer being more than or equal to 1 and being less than described M;
Using region corresponding for M-K detection window in a described M detection window except a described K detection window as target area.
As a kind of possible embodiment, above-mentioned processor 3000 determines that M detection window in described image to be detected comprises:
The cascade detectors preset is adopted to detect to described image to be detected M the detection window obtaining described image to be detected.
As a kind of possible embodiment, above-mentioned processor 3000 carries out unique point mark to the image in described M detection window region and obtains N number of unique point and comprise:
Meanshape unique point mark is carried out to a described M detection window, to obtain N number of unique point.
As a kind of possible embodiment, above-mentioned processor 3000 carries out registration process to described N number of unique point and comprises:
The characteristic matching function of training in advance is adopted to carry out feature extraction to described N number of unique point, to obtain N number of partial binary feature;
The overall linear regression model (LRM) of training in advance is adopted to carry out overall linear regression processing to described N number of partial binary feature.
As a kind of possible embodiment, a described P proper vector is carried out classification process and is comprised with K the detection window got rid of in a described M detection window by above-mentioned processor 3000:
Scale invariant features transform SIFT is carried out to a described P unique point, to obtain P SIFT feature point;
By described P SIFT feature point composition target feature vector;
Default sorter is adopted to carry out classification process to get rid of K detection window in a described M detection window to described target feature vector.
In specific implementation, input equipment 1000 described in the embodiment of the present invention, output device 2000 and processor 3000 can perform the first embodiment of a kind of vehicle checking method that the embodiment of the present invention provides and the implementation described in the second embodiment, also can perform the implementation of the terminal described in the first embodiment of a kind of terminal that the embodiment of the present invention provides, not repeat them here.
The image to be detected comprising vehicle can be obtained by the pick-up unit described by the embodiment of the present invention; Determine M detection window in described image to be detected, described M be greater than 1 integer; Carry out unique point mark to the image in described M detection window region and obtain N number of unique point, wherein, described M and described N is the integer being greater than 1; Registration process is carried out to described N number of unique point; Feature extraction is carried out centered by each unique point in N number of unique point after described registration process; Feature composition P the proper vector extracted described, wherein, described P be greater than 1 integer; A described P proper vector is carried out classification process to get rid of K detection window in a described M detection window, wherein, described K is the integer being more than or equal to 1 and being less than described M; Using region corresponding for M-K detection window in a described M detection window except a described K detection window as target area.Implement, in the process of the embodiment of the present invention, the number of wrong window can be reduced, thus, improve the accuracy rate of vehicle detection.
The embodiment of the present invention also provides a kind of computer-readable storage medium, and wherein, this computer-readable storage medium can have program stored therein, and comprises the part or all of step of any one signal processing method recorded in said method embodiment when this program performs.
In the above-described embodiments, the description of each embodiment is all emphasized particularly on different fields, in certain embodiment, there is no the part described in detail, can see the associated description of other embodiments.
It should be noted that, for aforesaid each embodiment of the method, in order to simple description, therefore it is all expressed as a series of combination of actions, but those skilled in the art should know, the present invention is not by the restriction of described sequence of movement, because according to the present invention, some step may can adopt other orders or carry out simultaneously.Secondly, those skilled in the art also should know, the embodiment described in instructions all belongs to preferred embodiment, and involved action and module might not be that the present invention is necessary.
In several embodiments that the application provides, should be understood that, disclosed device, the mode by other realizes.Such as, device embodiment described above is only schematic, the division of such as said units, be only a kind of logic function to divide, actual can have other dividing mode when realizing, such as multiple unit or assembly can in conjunction with or another system can be integrated into, or some features can be ignored, or do not perform.Another point, shown or discussed coupling each other or direct-coupling or communication connection can be by some interfaces, and the indirect coupling of device or unit or communication connection can be electrical or other form.
The above-mentioned unit illustrated as separating component or can may not be and physically separates, and the parts as unit display can be or may not be physical location, namely can be positioned at a place, or also can be distributed in multiple network element.Some or all of unit wherein can be selected according to the actual needs to realize the object of the present embodiment scheme.
In addition, each functional unit in various embodiments of the present invention can be integrated in a processing unit, also can be that the independent physics of unit exists, also can two or more unit in a unit integrated.Above-mentioned integrated unit both can adopt the form of hardware to realize, and the form of SFU software functional unit also can be adopted to realize.
If above-mentioned integrated unit using the form of SFU software functional unit realize and as independently production marketing or use time, can be stored in a computer read/write memory medium.Based on such understanding, the part that technical scheme of the present invention contributes to prior art in essence in other words or all or part of of this technical scheme can embody with the form of software product, this computer software product is stored in a storage medium, comprise all or part of step of some instructions in order to make a computer equipment (can be personal computer, server or the network equipment etc., can be specifically the processor in computer equipment) perform each embodiment said method of the present invention.Wherein, and aforesaid storage medium can comprise: USB flash disk, portable hard drive, magnetic disc, CD, ROM (read-only memory) are (English: Read-OnlyMemory, abbreviation: ROM) or random access memory (English: RandomAccessMemory, abbreviation: RAM) etc. various can be program code stored medium.
The above, above embodiment only in order to technical scheme of the present invention to be described, is not intended to limit; Although with reference to previous embodiment to invention has been detailed description, those of ordinary skill in the art is to be understood that: it still can be modified to the technical scheme described in foregoing embodiments, or carries out equivalent replacement to wherein portion of techniques feature; And these amendments or replacement, do not make the essence of appropriate technical solution depart from the spirit and scope of various embodiments of the present invention technical scheme.
Accompanying drawing explanation
In order to be illustrated more clearly in the technical scheme in the embodiment of the present invention, below the accompanying drawing used required in describing embodiment is briefly described, apparently, accompanying drawing in the following describes is some embodiments of the present invention, for those of ordinary skill in the art, under the prerequisite not paying creative work, other accompanying drawing can also be obtained according to these accompanying drawings.
Fig. 1 is the first embodiment schematic flow sheet of the method for a kind of vehicle detection that the embodiment of the present invention provides;
Fig. 2 is the second embodiment schematic flow sheet of the method for a kind of vehicle detection that the embodiment of the present invention provides;
Fig. 3 a is the first example structure schematic diagram of a kind of pick-up unit that the embodiment of the present invention provides;
Fig. 3 b be a kind of pick-up unit of providing of the embodiment of the present invention the another structural representation of the first embodiment;
Fig. 3 c be a kind of pick-up unit of providing of the embodiment of the present invention the another structural representation of the first embodiment;
Fig. 4 is the second example structure schematic diagram of a kind of pick-up unit that the embodiment of the present invention provides.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, be clearly and completely described the technical scheme in the embodiment of the present invention, obviously, described embodiment is the present invention's part embodiment, instead of whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art, not making the every other embodiment obtained under creative work prerequisite, belong to the scope of protection of the invention.
Below be described in detail respectively.
Claims (10)
1. a vehicle checking method, is characterized in that, comprising:
Obtain the image to be detected comprising vehicle;
Determine M detection window in described image to be detected, described M be greater than 1 integer;
Carry out unique point mark to the image in described M detection window region and obtain N number of unique point, wherein, described M and described N is the integer being greater than 1;
Registration process is carried out to described N number of unique point;
Feature extraction is carried out centered by each unique point in N number of unique point after described registration process;
Feature composition P the proper vector extracted described, wherein, described P be greater than 1 integer;
A described P proper vector is carried out classification process to get rid of K detection window in a described M detection window, wherein, described K is the integer being more than or equal to 1 and being less than described M;
Using region corresponding for M-K detection window in a described M detection window except a described K detection window as target area.
2. the method for claim 1, is characterized in that, described M the detection window determined in described image to be detected comprises:
The cascade detectors preset is adopted to detect to described image to be detected M the detection window obtaining described image to be detected.
3. the method for claim 1, is characterized in that, describedly carries out unique point mark to the image in described M detection window region and obtains N number of unique point and comprise:
Meanshape unique point mark is carried out to a described M detection window, to obtain N number of unique point.
4. the method for claim 1, is characterized in that, describedly carries out registration process to described N number of unique point and comprises:
The characteristic matching function of training in advance is adopted to carry out feature extraction to described N number of unique point, to obtain N number of partial binary feature;
The overall linear regression model (LRM) of training in advance is adopted to carry out overall linear regression processing to described N number of partial binary feature.
5. the method for claim 1, is characterized in that, described a described P proper vector is carried out classification process comprise with K the detection window got rid of in a described M detection window:
Scale invariant features transform SIFT is carried out to a described P unique point, to obtain P SIFT feature point;
By described P SIFT feature point composition target feature vector;
Default sorter is adopted to carry out classification process to get rid of K detection window in a described M detection window to described target feature vector.
6. a pick-up unit, is characterized in that, comprising:
Acquiring unit, for obtaining the image to be detected comprising vehicle;
First determining unit, for determining M detection window in the image to be detected that described acquiring unit gets, described M be greater than 1 integer;
Indexing unit, carry out unique point mark for the image in M detection window region determining described first determining unit and obtain N number of unique point, wherein, described M and described N is the integer being greater than 1;
Alignment unit, carries out registration process for the N number of unique point marked described indexing unit;
First extraction unit, for carrying out feature extraction centered by each unique point in the N number of unique point after described alignment unit registration process;
First component units, for described first extraction unit is extracted feature composition P proper vector, wherein, described P be greater than 1 integer;
Taxon, for P proper vector of described first component units composition is carried out classification process to get rid of K detection window in a described M detection window, wherein, described K is the integer being more than or equal to 1 and being less than described M;
Second determining unit, for using region corresponding for M-K detection window in a described M detection window except K the detection window that described taxon is got rid of as target area.
7. pick-up unit as claimed in claim 6, is characterized in that, described first determining unit specifically for:
The cascade detectors preset is adopted to detect to described image to be detected M the detection window obtaining described image to be detected.
8. pick-up unit as claimed in claim 6, is characterized in that, described indexing unit specifically for:
Meanshape unique point mark is carried out to a described M detection window, to obtain N number of unique point.
9. pick-up unit as claimed in claim 6, it is characterized in that, described alignment unit comprises:
Second extraction unit, for adopting the characteristic matching function of training in advance to carry out feature extraction to N number of unique point that described indexing unit marks, to obtain N number of partial binary feature;
Processing unit, carries out overall linear regression processing for adopting the overall linear regression model (LRM) of training in advance to N number of partial binary feature that described second extraction unit extracts.
10. pick-up unit as claimed in claim 6, it is characterized in that, described taxon comprises:
Converter unit, for carrying out Scale invariant features transform SIFT, to obtain P SIFT feature point to P unique point of described first component units composition;
Second component units, for converting P the SIFT feature point composition target feature vector obtained by described converter unit;
Classification subelement, carries out classification process to get rid of K detection window in a described M detection window for adopting the target feature vector of default sorter to described second component units composition.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510685237.2A CN105184286A (en) | 2015-10-20 | 2015-10-20 | Vehicle detection method and detection device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510685237.2A CN105184286A (en) | 2015-10-20 | 2015-10-20 | Vehicle detection method and detection device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105184286A true CN105184286A (en) | 2015-12-23 |
Family
ID=54906352
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510685237.2A Pending CN105184286A (en) | 2015-10-20 | 2015-10-20 | Vehicle detection method and detection device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105184286A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106446927A (en) * | 2016-07-07 | 2017-02-22 | 浙江大学 | Self-paced reinforcement image classification method and system |
| CN106971187A (en) * | 2017-04-12 | 2017-07-21 | 华中科技大学 | A kind of vehicle part detection method and system based on vehicle characteristics point |
| CN107545216A (en) * | 2016-06-29 | 2018-01-05 | 深圳市格视智能科技有限公司 | A kind of vehicle identification method available for power-line patrolling |
| CN109492755A (en) * | 2018-11-07 | 2019-03-19 | 北京旷视科技有限公司 | Image processing method, image processing apparatus and computer readable storage medium |
| CN109615925A (en) * | 2019-01-03 | 2019-04-12 | 上海钧正网络科技有限公司 | Vehicle parking control method, device, system and server based on video monitoring |
| CN110415424A (en) * | 2019-06-17 | 2019-11-05 | 众安信息技术服务有限公司 | A kind of authentication method, apparatus, computer equipment and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060210168A1 (en) * | 2005-03-02 | 2006-09-21 | Samsung Electronics Co., Ltd. | Apparatus and method for generating shape model of object and apparatus and method for automatically searching for feature points of object employing the same |
| CN101320520A (en) * | 2008-07-21 | 2008-12-10 | 北京中星微电子有限公司 | Vehicle detection method and equipment thereof |
| CN102663431A (en) * | 2012-04-17 | 2012-09-12 | 北京博研新创数码科技有限公司 | Image matching calculation method on basis of region weighting |
| CN103198323A (en) * | 2013-03-12 | 2013-07-10 | 无锡慧眼电子科技有限公司 | Pedestrian detection method based on curve fitting |
| CN104537359A (en) * | 2015-01-15 | 2015-04-22 | 浙江宇视科技有限公司 | Vehicle object detection method and device |
| CN104598936A (en) * | 2015-02-28 | 2015-05-06 | 北京畅景立达软件技术有限公司 | Human face image face key point positioning method |
-
2015
- 2015-10-20 CN CN201510685237.2A patent/CN105184286A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060210168A1 (en) * | 2005-03-02 | 2006-09-21 | Samsung Electronics Co., Ltd. | Apparatus and method for generating shape model of object and apparatus and method for automatically searching for feature points of object employing the same |
| CN101320520A (en) * | 2008-07-21 | 2008-12-10 | 北京中星微电子有限公司 | Vehicle detection method and equipment thereof |
| CN102663431A (en) * | 2012-04-17 | 2012-09-12 | 北京博研新创数码科技有限公司 | Image matching calculation method on basis of region weighting |
| CN103198323A (en) * | 2013-03-12 | 2013-07-10 | 无锡慧眼电子科技有限公司 | Pedestrian detection method based on curve fitting |
| CN104537359A (en) * | 2015-01-15 | 2015-04-22 | 浙江宇视科技有限公司 | Vehicle object detection method and device |
| CN104598936A (en) * | 2015-02-28 | 2015-05-06 | 北京畅景立达软件技术有限公司 | Human face image face key point positioning method |
Non-Patent Citations (3)
| Title |
|---|
| 《交通信息感知理论与方法汉、英》: "《交通信息感知理论与方法汉、英》", 30 September 2014, 东南大学出版社 * |
| SHAOQING REN,ET AL.: "Face Alignment at 3000 FPS via Regressing Local Binary Features", 《2014 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION》 * |
| 姜谊: "车牌检测及汽车类型分类方法研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107545216A (en) * | 2016-06-29 | 2018-01-05 | 深圳市格视智能科技有限公司 | A kind of vehicle identification method available for power-line patrolling |
| CN106446927A (en) * | 2016-07-07 | 2017-02-22 | 浙江大学 | Self-paced reinforcement image classification method and system |
| CN106446927B (en) * | 2016-07-07 | 2019-05-28 | 浙江大学 | It is a kind of to enhance image classification method and system from step |
| CN106971187A (en) * | 2017-04-12 | 2017-07-21 | 华中科技大学 | A kind of vehicle part detection method and system based on vehicle characteristics point |
| CN106971187B (en) * | 2017-04-12 | 2019-07-09 | 华中科技大学 | A kind of vehicle part detection method and system based on vehicle characteristics point |
| CN109492755A (en) * | 2018-11-07 | 2019-03-19 | 北京旷视科技有限公司 | Image processing method, image processing apparatus and computer readable storage medium |
| CN109492755B (en) * | 2018-11-07 | 2022-03-01 | 北京旷视科技有限公司 | Image processing method, image processing apparatus, and computer-readable storage medium |
| CN109615925A (en) * | 2019-01-03 | 2019-04-12 | 上海钧正网络科技有限公司 | Vehicle parking control method, device, system and server based on video monitoring |
| CN110415424A (en) * | 2019-06-17 | 2019-11-05 | 众安信息技术服务有限公司 | A kind of authentication method, apparatus, computer equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105184286A (en) | Vehicle detection method and detection device | |
| CN108830188B (en) | Vehicle detection method based on deep learning | |
| CN105224951B (en) | A kind of vehicle type classification method and sorter | |
| Liu et al. | Convolutional neural networks-based intelligent recognition of Chinese license plates | |
| CN110399884B (en) | Feature fusion self-adaptive anchor frame model vehicle detection method | |
| CN111274926B (en) | Image data screening method, device, computer equipment and storage medium | |
| CN108009567B (en) | Automatic excrement character distinguishing method combining image color and HOG and SVM | |
| CN108805180B (en) | Target object detection method and device | |
| CN104615986A (en) | Method for utilizing multiple detectors to conduct pedestrian detection on video images of scene change | |
| CN112580734B (en) | Target detection model training method, system, terminal equipment and storage medium | |
| CN112560971A (en) | Image classification method and system for active learning self-iteration | |
| CN111291807A (en) | Fine-grained image classification method and device and storage medium | |
| CN112241736A (en) | Text detection method and device | |
| CN114359624A (en) | Article sorting method and device and storage medium | |
| CN108362227A (en) | Wheel hub detection method, device, system and control device | |
| CN111027601B (en) | Plane detection method and device based on laser sensor | |
| Chaturvedi et al. | Automatic license plate recognition system using surf features and rbf neural network | |
| CN113723431B (en) | Image recognition method, apparatus and computer readable storage medium | |
| CN105160333B (en) | A kind of model recognizing method and identification device | |
| CN114422450B (en) | Network traffic analysis method and device based on multi-source network traffic data | |
| CN104123538A (en) | Network malicious image detection method based on visual word bag | |
| CN111931229B (en) | Data identification method, device and storage medium | |
| CN107895167A (en) | A kind of ir data classifying identification method based on rarefaction representation | |
| CN111127485B (en) | Method, device and equipment for extracting target area in CT image | |
| CN112766387A (en) | Error correction method, device, equipment and storage medium for training data |
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 |
Application publication date: 20151223 |
|
| RJ01 | Rejection of invention patent application after publication |