CN104537387A - Method and system for classifying automobile types based on neural network - Google Patents
Method and system for classifying automobile types based on neural network Download PDFInfo
- Publication number
- CN104537387A CN104537387A CN201410790026.0A CN201410790026A CN104537387A CN 104537387 A CN104537387 A CN 104537387A CN 201410790026 A CN201410790026 A CN 201410790026A CN 104537387 A CN104537387 A CN 104537387A
- Authority
- CN
- China
- Prior art keywords
- classification
- vehicle
- neuron
- image
- classification results
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- 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
Abstract
The invention relates to the field of automobile classification technologies and discloses a method and system for classifying automobile types based on a neural network. The method comprises the steps that a plurality of training samples are collected, and the training samples are classified based on the convolutional neural network, so a classifier containing label results is obtained; when the automobile types are classified, a video image to be detected is read in, a motion object in the image is detected, and sub-block processing is performed on the image according to the motion object; afterwards, classification processing is performed on each block of image through the classifier, so a detection result is obtained. Accordingly, a neural network system can be constructed easily and conveniently to serve as the classifier, the system is trained by using different automobile samples, the system is made to automatically study complex class conditional density of the samples, and therefore the problems caused by class conditional density functions of artificial hypothesis are avoided. Compared with an existing automobile type classification method, the method for classifying the automobile types based on the convolutional neural network has the advantages that the accuracy of classification is improved, and the classification speed is increased.
Description
Technical field
The present invention relates to separation vehicle technical field, particularly utilize the method and system of neural fusion vehicle classification.
Background technology
Car category classification refers to according to different criteria for classifications, and automobile is divided into different car category.Owing to there is various interference, the complicacy of such as surrounding enviroment, the interference of all kinds of factor of weather reason etc., causes vehicle classification to become studying a question of a more complicated.The subject matter that research at present in Vehicle Classification Technique exists is: accuracy of detection is not high, detection speed is too slow, the problems such as calculated amount is too large.Such as: when having a large amount of vehicles through a certain section, the difficulty so detected strengthens, if the supervisory system that real-time is slow, all type of vehicle can not be detected simultaneously, in addition when weather is more severe, such as: heavy rain weather, snowy day, and haze serious time, can not accurately vehicle be classified.
At present, separation vehicle is 8 classes by Chinese automobile criteria for classification:
1, cargo vehicle: minitruck---Ga<=1.8t; Pickup truck---1.8T; Medium truck---6t; Heavy goods vehicle---Ga>14t.
2, cross-country vehicle: puddle jumper---Ga<=5t; Medium off-road vehicle---5t<=Ga<=13t; Heavy off-road vehicle---13t; Extra heavy off road vehicle---Ga>24t.
3, dump truck: light-duty dump truck---Ga<=6t; Medium dumper---6t; Heavy duty dump truck---Ga>14t; Mining dump truck.
4, tractor: semi-trailer towing vehicle; Full-trailer towing vehicle.
5, special purpose vehicle: box-type automobile; Tank car; Crane/lift truck; Stake truck; Special construction vehicle; Special dump truck.
6, passenger vehicle: microbus---L<=3.5m; Light bus---3.5m; Middle bus---7m; Motorbus---L>10m; Extra bus.
7, car: minicar---V<=1L; Subcompact car---1L; Intermediate-size car---1.6L; Intermediate car---2.5L; Sedan limousine---V>4L.
8, semitrailer: light semi-trailer---Ga<=7.1t; Medium semitrailer---7.1t; Heavy semi-trailer---19.5t; Extra heavy semi-trailer---Ga>34t.
How so numerous type of vehicle, carry out effectively fast it and classify accurately, is a permanent research topic.Due to the diversity of type of vehicle, make vehicle classification degree of difficulty larger, in addition the complicacy of surrounding enviroment, there is a large amount of interference, further increase the complexity of classification.And in some specific occasions, the monitoring on such as expressway, the monitoring of crossroad etc., all need the vehicle of any type during judgement vehicle real-time.
For vehicle classification problem, for a long time, although researchist proposes a large amount of vehicle classifications, when their performance of detection method is described, often use different vehicle testing collection, therefore the superior and inferior evaluating of vehicle checking method lacks a unified standard.Various vehicle checking method is done to the contrast in performance, the development of vehicle testing techniques will be promoted undoubtedly, but, this purpose be reached and just must set up a set of the recognized standard.And in fact, the property complicated and changeable of environment residing for vehicle, it is more difficult for wanting to set up comprehensive, an effective vehicle detection test pattern storehouse.When evaluating vehicle detection performance, introduce four indexs: detect accuracy (correct.rate), false alarm rate (false.alarm-rate), detection speed (detectingspeed) and robustness (robustness).
Detect accuracy, be also precision, the number of vehicles be properly detected exactly is divided by the number of vehicles comprised in original image.Detect accuracy higher, illustrate that the ability to accept of detection system to vehicle is stronger.
Detection speed, major applications field needs to detect vehicle in real time online, as vehicle tracking, programmable vision monitoring etc.Under the prerequisite that verification and measurement ratio and false drop rate reach satisfied, The faster the better for detection speed.At present, vehicle testing techniques is also very unripe, and the uncertain factor affecting vehicle detection result is a lot, such as the friendship and background etc. of attitude.Although these factors do not form too large obstacle for the vision system of the mankind, but certain challenge is just proposed to existing vehicle detecting system, because for vehicle detecting system, it just can only can obtain good Detection results under certain restrictive condition, and needs to improve in detection speed.
In actual applications, because great majority are all towards real-time process, this requires that vehicle detecting algorithm is convenient to realize, and precision wants high, and has detection speed faster.Current vehicle detecting algorithm can't process any environment, illumination preferably and the change condition such as block, and in accuracy of detection, detection speed aspect Shortcomings.
The domestic research to vehicle classification problem is a lot, and the personnel of multiple university, research institution have put in the middle of the research in this field of vehicle classification, and also achieve certain achievement in research.
The human hairs such as Jia Limin, Lee Hai Jian of Beijing Jiaotong University understand a kind of frequency domain spectra energy vehicle type classification method based on geomagnetic sensor.A kind of frequency domain spectra energy vehicle type classification method based on geomagnetic sensor of this disclosure of the invention, belongs to transport information and detects and signal processing technology field.The method comprises the following steps: step 1, utilize geomagnetic sensor to extract vehicle waveform signal, and to the standardization of each vehicle waveform, recycling fast connect obtains the frequency spectrum of each vehicle; Step 2, analysis vehicle frequency spectrum are in the distribution characteristics of different frequency domain section, and definition M kind spectrum energy model, classifies to different automobile types; After step 3, the optimum frequency domain section utilizing optimum frequency domain section lookup algorithm to obtain being applicable to different automobile types criteria for classification or the combination of optimum frequency domain section, provide vehicle classification result; Step 4, analysis vehicle classification result, select the spectrum energy model needed in M kind spectrum energy model.This invention provides optimum capacity model and optimum frequency domain section or combination, for the vehicle classification engineer applied of geomagnetic sensor provides theoretical foundation and application directs.
The human hairs such as Meng Huadong, Zhang Hao of Tsing-Hua University understand a kind of vehicle type classification method based on single-frequency continuous wave radar, and this invention, based on the vehicle type classification method of single-frequency continuous wave radar, belongs to intelligent traffic vehicle detection field.The method, using the time domain radar signal of single-frequency continuous wave radar as input, is carried out time frequency analysis and is obtained time dependent radar return Doppler frequency spectrum figure; By hough transform, Doppler image is mapped as vehicle scattering center location parameter image; Through feature extraction; Karhunen-Loeve (K-L) screening and compression obtain feature samples, utilize Fisher criterion to carry out sample classification, obtain the vehicle classification of vehicle.Use multiple two type sorter to carry out pairwise classification respectively for polymorphic type situation, vote according to classifier result, obtain vehicle classification result.Single-frequency continuous wave radar is simple and reliable for structure, with low cost.The present invention utilize single-frequency continuous wave radar achieve vehicle detection, test the speed while carry out vehicle classification, in practical application, the average correct classification rate that vehicle is divided three classes is reached 94.8%.
The human hairs such as the Jiang Yong oak of Ningbo Haivision Intelligence System Co., Ltd. understand a kind of vehicle type classification method based on video, and this inventive method comprises the following steps: step 1, on road, precalculated position arranges surveyed area and gathers video to described surveyed area; Step 2, in the video collected, isolate the vehicle target prospect of motion; Step 3, to classify according to the vehicle target prospect of vehicle geometric properties to isolated motion.By the enforcement to technical solution of the present invention, following technique effect can be obtained: automatically complete vehicle cab recognition, save manpower; Recognition speed is fast, and accuracy of identification is high, improves the efficiency of vehicle cab recognition.
The Song Yingliang of Guangzhou Turing Information Technology Co., Ltd., the human hairs such as Yang Qinghai understand a kind of automatic vehicle type classification system, this utility model discloses automatic vehicle type classification system, and it comprises infrared launcher, infrared receiving device, ground induction coil, some groups of pressure transducers and signal transacting MCU; Described infrared launcher, infrared receiving device are arranged on the entrance left and right sides, track respectively; Described some groups of pressure transducers are arranged on track, and vertical with track; Described ground induction coil is embedded in below track, and with track with wide; Described signal transacting MCU is connected with ground induction coil, some groups of pressure transducers and infrared receiving device respectively.This system has not only played the accurate judgement of infrared detection sorting technique to vehicle two-dimensional silhouette, also uses the advantage that pressure transducer efficiently judges axletree distance and axletree number, increases substantially the accuracy of vehicle classification.
But still there is certain deficiency in existing vehicle type classification method, prior art haves much room for improvement and improves in the precision or speed of classification.
Summary of the invention
In view of this, be necessary, for Problems existing in existing vehicle classification, a kind of method and system utilizing neural fusion vehicle classification to be provided, the speed of nicety of grading and classification can be improved.
In order to achieve the above object, this invention takes following technical scheme:
Utilize a method for neural fusion vehicle classification, it comprises the steps:
A, collect several training samples, utilize convolutional neural networks to classify to described training sample, obtain the sorter comprising label result;
B, when classifying to vehicle, reading in video image to be measured, the moving target in detected image, according to moving target, piecemeal process being carried out to image; Re-use described sorter to carry out classification process to every block image and draw testing result.
Described utilizes in the method for neural fusion vehicle classification, and described step B comprises:
B1, read in video image, in time moving object having been detected, extract moving object region;
The block of B2, employing fixed size carries out piecemeal process to this moving object region;
B3, utilize convolutional neural networks to carry out classification to obtain classification results.
Described utilizes in the method for neural fusion vehicle classification, specifically comprises in described step B2:
With the block of n*n size, piecemeal process is carried out to moving object region, then move a pixel successively, obtain some pictures, then convergent-divergent is carried out to described some pictures, be converted to the picture that pixel value is 32*32 size; Wherein, n is natural number, and its span is between 50--70.
Described utilizes in the method for neural fusion vehicle classification, and in described step B, classification results comprises: for represent car first kind classification results, for represent passenger vehicle Equations of The Second Kind classification results, for represent truck the 3rd class classification results, for representing the 4th class classification results of offroad vehicle and the 5th class classification results for representing non-vehicle.
Described utilizes in the method for neural fusion vehicle classification, and described steps A comprises:
The convolution kernel of A1, employing fixed size removes each neuron in perception input picture and carries out first to each neuron to be biased process, obtains first volume lamination;
A2, the neuron of first volume lamination to be divided into groups, sue for peace respectively to often organizing neuron, and carry out after the first weighting, second is biased process, using sigmoid function as the activation function of convolutional network to the neuron after summation, obtain fisrt feature mapping graph, i.e. the first down-sampling layer;
A3, fisrt feature mapping graph carried out to process of convolution and obtain volume Two lamination;
A4, the neuron of volume Two lamination being divided into groups, suing for peace respectively to often organizing neuron, and carry out after the second weighting, the 3rd is biased process, uses sigmoid activation function to ask thresholding, obtaining second feature mapping graph S to the neuron after summation
x+1, i.e. the second down-sampling layer;
A5, each neuron in second feature mapping graph and the neuron in input picture are connected to form neural network export.
Described utilizes in the method for neural fusion vehicle classification, and described steps A 1 comprises:
A11, adopt trainable wave filter f
xthe image of convolution one input obtains convolution feature map;
A12, convolution feature map is added a biased b
x, obtain first volume lamination C
x.
Described utilizes in the method for neural fusion vehicle classification, and described steps A 2 comprises:
A21, a pixel is become to the summation of neighborhood every in first volume lamination four pixels obtain scalar W
x+1;
A22, to scalar W
x+1weighting, the biased b of increase
x+1process;
A23, use sigmoid function, as the activation function of convolutional network, obtain the fisrt feature mapping graph S reducing four times
x+1.
Utilize a system for neural fusion vehicle classification, it comprises:
Processing unit, for collecting several training samples, utilizing convolutional neural networks to classify to described training sample, obtaining the sorter comprising label result;
Output unit, for when classifying to vehicle, reading in the moving target in video image to be measured, detected image, carrying out piecemeal process according to moving target to image, and uses described sorter to carry out classification process to every block image to draw testing result.
Described utilizes in the system of neural fusion vehicle classification, and described output unit comprises further:
Extracting subelement, for reading in video image, in time moving object having been detected, extracting moving object region;
Piecemeal subelement, carries out piecemeal process for adopting the block of fixed size to this moving object region;
Classification subelement, carries out classification for utilizing convolutional neural networks and obtains classification results.
Described utilizes in the system of neural fusion vehicle classification, and in described output unit, classification results comprises: for represent car first kind classification results, for represent passenger vehicle Equations of The Second Kind classification results, for represent truck the 3rd class classification results, for representing the 4th class classification results of offroad vehicle and the 5th class classification results for representing non-vehicle.
Beneficial effect: the present invention utilizes the method and system of neural fusion vehicle classification, by collecting several training samples, convolutional neural networks is utilized to classify to described training sample, obtain the sorter comprising label result, when classifying to vehicle, read in video image to be measured, the moving target in detected image, according to moving target, piecemeal process is carried out to image; Re-use described sorter to carry out classification process to every block image and draw testing result, thus can be easy construct nerve network system as sorter, different vehicle sample is used to train this system, allow the class sigma-t of system automatic learning sample complexity, avoid the problem that artificial hypothesis class conditional density function brings.The present invention is based on convolutional neural networks to be relative to the method advantage of existing vehicle classification aspect, improve the precision of classification, and be improved in the speed of classification.
Accompanying drawing explanation
Fig. 1 is the process flow diagram that the present invention utilizes the method for neural fusion vehicle classification.
Fig. 2 is the schematic diagram that the present invention utilizes neural network in the method for neural fusion vehicle classification.
Fig. 3 is the process schematic that the present invention utilizes step S100 in the method for neural fusion vehicle classification.
Fig. 4 is the structured flowchart that the present invention utilizes neural fusion Vehicle Classification System.
Embodiment
The invention provides the method and system utilizing neural fusion vehicle classification, under complex environment, various vehicle is carried out effectively and classifies accurately, thus improve the precision of classification and the speed of classification.The present invention can be applied to a large amount of occasions, traffic surveillance and control system, security device system etc.In vehicle classification, have higher precision, faster in detection speed, this is vital for the system of some real-time, has huge application prospect.
As shown in Figure 1, the method for neural fusion vehicle classification that utilizes provided by the invention comprises the steps:
S100, collect several training samples, utilize convolutional neural networks to classify to described training sample, obtain the sorter comprising label result;
S200, when classifying to vehicle, reading in video image to be measured, the moving target in detected image, according to moving target, piecemeal process being carried out to image; Re-use described sorter to carry out classification process to every block image and draw testing result.
Wherein, step S100 is training process, has collected 250,000 samples: comprising 50,000 car pictures, 50,000 passenger vehicle samples pictures, 50,000 trucies, 50,000 offroad vehicle pictures and 50,000 non-vehicle pictures when training; Then this 250,000 pictures is classified through convolutional neural networks, obtain label result, namely comprise: passenger vehicle, truck, car, offroad vehicle, non-vehicle.
Step S200 is test process, is used for testing the neural network that uses whether reliable for the precision of vehicle classification, speed.Its process comprises: read in video image, is detected by the moving target detecting method of the video read in based on optical flow approach, and Detection and Extraction go out area-of-interest and image is carried out to piecemeal, sorter classification, draws testing result.The present invention, by carrying out piecemeal process to the image after motion detection, can accelerate the speed of testing like this, has vital effect for the reduction test duration.
Particularly, described step S200 comprises: read in video image, in time moving object having been detected, extracts moving object region; Afterwards, the block of fixed size is adopted to carry out piecemeal process to this moving object region; Again, utilize convolutional neural networks to carry out classification and obtain classification results.
Wherein, when carrying out piecemeal to image, using the block of n*n size to carry out piecemeal process to moving object region, then moving a pixel successively, obtain some pictures, then convergent-divergent is carried out to described some pictures, be converted to the picture that pixel value is 32*32 size; Wherein, n is natural number, and its span is between 50-70.Then, the picture these obtained is as input, utilize convolutional neural networks to classify, the result of classification comprises: for represent car first kind classification results, for represent passenger vehicle Equations of The Second Kind classification results, for represent truck the 3rd class classification results, for representing the 4th class classification results of offroad vehicle and the 5th class classification results for representing non-vehicle.
Convolutional neural networks is also an emphasis of the present invention, and convolutional neural networks (CNN) is the one of artificial neural network, has become the study hotspot of current speech analysis and field of image recognition.Its weights shared network structure makes it more to be similar to biological neural network, reduces the complexity of network model, decreases the quantity of weights.
Described step S100 comprises: the convolution kernel of a1, employing fixed size removes each neuron in perception input picture and carry out first to each neuron to be biased process, obtains first volume lamination; A2, the neuron of first volume lamination to be divided into groups, sue for peace respectively to often organizing neuron, and carry out after the first weighting, second is biased process, using sigmoid function as the activation function of convolutional network to the neuron after summation, obtain fisrt feature mapping graph, i.e. the first down-sampling layer; A3, fisrt feature mapping graph carried out to process of convolution and obtain volume Two lamination; A4, the neuron of volume Two lamination to be divided into groups, sue for peace respectively to often organizing neuron, and carry out after the second weighting, the 3rd is biased process, using sigmoid activation function to ask thresholding to the neuron after summation, obtain second feature mapping graph Sx+1, i.e. the second down-sampling layer; A5, each neuron in second feature mapping graph and the neuron in input picture are connected to form neural network export.
Wherein, described step a1 is convolution process, and it comprises: adopt trainable wave filter f
xthe image (first stage is the image inputted, and the stage has below been exactly convolution feature map) of convolution one input obtains convolution feature map, convolution feature map is added a biased bx, obtains first volume lamination C
x.
Described step a2 is that sub-sampling procedures comprises: four the pixel summations of every neighborhood become a pixel and obtain scalar W
x+1, then by scalar W
x+1weighting, then increase biased b
x+11, then pass through a sigmoid activation function as the activation function of convolutional network, produce the fisrt feature mapping graph S that is probably reduced four times
x+1.Do convolution algorithm so can be regarded as from a plane to the mapping of next plane, S-layer can regard fuzzy filter as, plays the effect of Further Feature Extraction.Between hidden layer and hidden layer, spatial resolution is successively decreased, and the number of planes contained by every layer increases progressively, and can be used for like this detecting more characteristic information.
Below in conjunction with Fig. 2 and Fig. 3, convolution process and sub-sampling procedures are described in detail:
Each neuron (i.e. each pixel) in perception input picture is removed with the convolution kernel of a fixed size, feature map is produced at first volume lamination C1 after convolution, afterwards, four pixels often organized in feature map are sued for peace again, weighted value, be biased, obtained the feature map of the first down-sampling layer S2 by a Sigmoid function; These map obtain volume Two lamination C3 through convolution again; This hierarchical structure again with the first down-sampling layer S2 is the same produces the second down-sampling layer S4; Again each feature map of the second down-sampling layer S4 is connected with each neuron in convolutional layer C afterwards, the generation of over-fitting can be prevented like this.Finally, these pixel values are rasterized, and connect into a vector and be input to traditional neural network, exported.
Usually, convolutional layer C is feature extraction layer, and the convolution kernel be made up of weights with removes each feature map of one deck before perception, and this has just extracted the feature of image, and generates the feature map of this convolutional layer; S layer is down-sampling layer, and each computation layer of network is made up of multiple Feature Mapping, and each Feature Mapping is a plane, and in plane, all neuronic weights are equal.Feature Mapping structure adopts affects the little sigmoid function of kernel function as the activation function of convolutional network, makes Feature Mapping have shift invariant.It should be noted that the convolution kernel used at every one deck is duplicate, so just reach the effect that weights are shared, the complexity of whole network is reduced greatly.
Convolutional neural networks of the present invention has 7 layers (not comprising input layer input), every layer all comprise can training parameter (connection weight), and each layer has multiple feature map, each feature Map extracts a kind of feature of input by a kind of convolution kernel, and then each feature Map has multiple neuron.
In the present invention, setting input picture is 32*32 size, first volume lamination C1 is made up of 6 feature map, in feature map, each neuron is connected with the neighborhood of 5*5 in input, the size of feature map is 28*28, first volume lamination C1 has (28*28+1) * 6=4710 can training parameter (comprising weights and bias), described first volume lamination C1 with have with input layer that 5*5*6*32*32=153600 is individual to be connected.
S2 layer is a down-sampling layer, and namely the first down-sampling layer, comprises the feature map of 6 14*14 sizes, and each unit in feature map is connected with the 2*2 neighborhood of corresponding feature map in first volume lamination C1.4 of each unit of first down-sampling layer S2 inputs are added, and being multiplied by one can training parameter, adds one and can train biased, then calculate result by sigmoid function.Coefficient and the biased nonlinear degree that control sigmoid function can be trained.The 2*2 receptive field of each unit is not overlapping, and therefore in the first down-sampling layer S2, the size of each characteristic pattern is 1/4 (row and column each 1/2) of characteristic pattern size in C1.First down-sampling layer S2 layer has (14*14+1) * 6=1020 can training parameter, has 6*28*28*5*5=117600 to be connected with first volume lamination C1 layer.
Volume Two lamination C3 is also a convolutional layer, and it to be deconvoluted the first down-sampling layer S2 by the convolution kernel of 5*5 equally, and the feature map obtained only has 10*10 neuron, the corresponding a kind of convolution kernel of each feature map, so it has 16 kinds of different convolution kernels.Here should be noted that a bit: each feature map in volume Two lamination C3 is all 6 or the several feature map that are connected in the first down-sampling layer S2, represents that the feature map of this layer is the various combination of the feature map that last layer extracts.
Second down-sampling layer S4 is also a down-sampling layer, it is made up of the feature map of 16 5*5 sizes, each unit in feature map is connected with the 2*2 neighborhood of individual features figure in volume Two lamination C3, the same with the connection between first volume lamination C1 and the first down-sampling layer S2.Second down-sampling layer S4 layer has 16*5*5+16=416 can training parameter, and its and volume Two lamination C3 layer one have that 10*10*5*5*16=65000 is individual to be connected.
Finally, the second down-sampling layer S4 layer is connected entirely with convolutional layer C, and this convolutional layer is made up of neuron one by one, and the present invention adopts 70 neurons, and each the feature map in the second down-sampling layer S4 is connected entirely with each neuron of this convolutional layer.Finally, be entirely connected by 70 of convolutional layer each labels of neuron and output layer, the object adding a convolutional layer is, prevents the situation of over-fitting from occurring.Finally export and obtain H
w,b(X).
Like this, be input as the picture of a 32*32 size, first volume lamination C1 has the feature map of 6 28*28 sizes, first down-sampling layer S2 has the feature map of 6 14*14 sizes, volume Two lamination C3 has the feature map of 16 10*10 sizes, second down-sampling layer S4 has the feature map of 16 5*5 sizes, convolutional layer has 70 neurons, and last output layer has five labels: car (representing with 1), passenger vehicle (representing with 2), truck (representing with 3), offroad vehicle (representing with 4), non-vehicle (representing with 5).
As shown in Figure 2, in fig. 2, Input size is the picture of 32*32; C1 is first volume lamination, and the feature map6 of total 28*28 is individual; S2 is the first down-sampling layer, and the feature map6 of total 14*14 is individual; C3 is volume Two lamination, and the feature map16 of total 10*10 is individual; S4 is the second down-sampling layer, and the feature map16 of total 5*5 size is individual; Convolutional layer has 70 neurons, is entirely connected with the second down-sampling layer S4 above; Last one deck is Output layer (output layer), is entirely be connected with convolutional layer above, exports and is Hw, b (X).
To sum up, the neural metwork training sorter that the present invention is provided by the embodiment of step S100, the main flow that neural network is used for pattern-recognition is supervised learning, and unsupervised learning is more for cluster analysis.For the pattern-recognition having supervision, classification due to arbitrary sample is known, the distribution of sample in space is no longer divide according to its NATURAL DISTRIBUTION tendency, but a kind of suitable space-division method to be looked for according to the separation degree between the distribution of similar sample in space and inhomogeneity sample, or find a classification boundaries, inhomogeneity sample is laid respectively in different regions.This just needs a long-time and learning process for complexity, and constantly adjustment is in order to divide the position of the classification boundaries of sample space, and the least possible sample is divided in non-homogeneous region.
Convolutional network is a kind of mapping being input to output in itself, it can learn the mapping relations between a large amount of constrained input, and without any need for the accurate mathematical expression formula between input and output, as long as trained convolutional network by known pattern, network just has the mapping ability between inputoutput pair.What convolutional network performed is Training, thus its sample set be by shape as: the vector of (input vector, desirable output vector) is to forming.All these vectors are right, should be all to derive from the actual " RUN " result that network is about to the system of simulation.They can gather from actual motion system.Before starting training, all power all should carry out initialization by some different little random numbers, the random number of such as distribution between [0,1]." little random number " is used for ensureing that network can not enter state of saturation because weights are excessive, thus causes failure to train; " difference " is used for ensureing that network can normally learn.In fact, if with identical several deinitialization weight matrixs, then have symmetry, cause the convolution kernel of every one deck all identical, then network impotentia study.
As shown in Figure 4, the present invention is also corresponding provides a kind of system utilizing neural fusion vehicle classification, and it comprises:
Processing unit 10, for collecting several training samples, utilizing convolutional neural networks to classify to described training sample, obtaining the sorter comprising label result;
Output unit 20, for when classifying to vehicle, reading in the moving target in video image to be measured, detected image, carrying out piecemeal process according to moving target to image, and uses described sorter to carry out classification process to every block image to draw testing result.
Wherein, in the present embodiment, described output unit 20 comprises further:
Extracting subelement 201, for reading in video image, in time moving object having been detected, extracting moving object region;
Piecemeal subelement 202, carries out piecemeal process for adopting the block of fixed size to this moving object region;
Classification subelement 203, carries out classification for utilizing convolutional neural networks and obtains classification results.
Further, in described output unit 20, classification results comprises: for represent car first kind classification results, for represent passenger vehicle Equations of The Second Kind classification results, for represent truck the 3rd class classification results, for representing the 4th class classification results of offroad vehicle and the 5th class classification results for representing non-vehicle.
The above embodiment only have expressed several embodiment of the present invention, and it describes comparatively concrete and detailed, but therefore can not be interpreted as the restriction to the scope of the claims of the present invention.It should be pointed out that for the person of ordinary skill of the art, without departing from the inventive concept of the premise, can also make some distortion and improvement, these all belong to protection scope of the present invention.Therefore, the protection domain of patent of the present invention should be as the criterion with claims.
Claims (10)
1. utilize a method for neural fusion vehicle classification, it is characterized in that: comprise the steps:
A, collect several training samples, utilize convolutional neural networks to classify to described training sample, obtain the sorter comprising label result;
B, when classifying to vehicle, reading in video image to be measured, the moving target in detected image, according to moving target, piecemeal process being carried out to image; Re-use described sorter to carry out classification process to every block image and draw testing result.
2. the method utilizing neural fusion vehicle classification according to claim 1, is characterized in that, described step B comprises:
B1, read in video image, in time moving object having been detected, extract moving object region;
The block of B2, employing fixed size carries out piecemeal process to this moving object region;
B3, utilize convolutional neural networks to carry out classification to obtain classification results.
3. the method utilizing neural fusion vehicle classification according to claim 2, is characterized in that, specifically comprises in described step B2:
With the block of n*n size, piecemeal process is carried out to moving object region, then move a pixel successively, obtain some pictures, then convergent-divergent is carried out to described some pictures, be converted to the picture that pixel value is 32*32 size; Wherein, n is natural number, and its span is between 50--70.
4. the method utilizing neural fusion vehicle classification according to claim 2, is characterized in that: in described step B, classification results comprises: for represent car first kind classification results, for represent passenger vehicle Equations of The Second Kind classification results, for represent truck the 3rd class classification results, for representing the 4th class classification results of offroad vehicle and the 5th class classification results for representing non-vehicle.
5. the method utilizing neural fusion vehicle classification according to claim 1, is characterized in that: described steps A comprises:
The convolution kernel of A1, employing fixed size removes each neuron in perception input picture and carries out first to each neuron to be biased process, obtains first volume lamination;
A2, the neuron of first volume lamination to be divided into groups, sue for peace respectively to often organizing neuron, and carry out after the first weighting, second is biased process, using sigmoid function as the activation function of convolutional network to the neuron after summation, obtain fisrt feature mapping graph, i.e. the first down-sampling layer;
A3, fisrt feature mapping graph carried out to process of convolution and obtain volume Two lamination;
A4, the neuron of volume Two lamination being divided into groups, suing for peace respectively to often organizing neuron, and carry out after the second weighting, the 3rd is biased process, uses sigmoid activation function to ask thresholding, obtaining second feature mapping graph S to the neuron after summation
x+1, i.e. the second down-sampling layer;
A5, each neuron in second feature mapping graph and the neuron in input picture are connected to form neural network export.
6. the method utilizing neural fusion vehicle classification according to claim 5, is characterized in that: described steps A 1 comprises:
A11, adopt trainable wave filter f
xthe image of convolution one input obtains convolution feature map;
A12, convolution feature map is added a biased b
x, obtain first volume lamination C
x.
7. the method utilizing neural fusion vehicle classification according to claim 5, is characterized in that: described steps A 2 comprises:
A21, a pixel is become to the summation of neighborhood every in first volume lamination four pixels obtain scalar W
x+1;
A22, to scalar W
x+1weighting, the biased b of increase
x+1process;
A23, use sigmoid function, as the activation function of convolutional network, obtain the fisrt feature mapping graph S reducing four times
x+1.
8. utilize a system for neural fusion vehicle classification, it is characterized in that: comprising:
Processing unit, for collecting several training samples, utilizing convolutional neural networks to classify to described training sample, obtaining the sorter comprising label result;
Output unit, for when classifying to vehicle, reading in the moving target in video image to be measured, detected image, carrying out piecemeal process according to moving target to image, and uses described sorter to carry out classification process to every block image to draw testing result.
9. the system utilizing neural fusion vehicle classification according to claim 8, is characterized in that, described output unit comprises further:
Extracting subelement, for reading in video image, in time moving object having been detected, extracting moving object region;
Piecemeal subelement, carries out piecemeal process for adopting the block of fixed size to this moving object region;
Classification subelement, carries out classification for utilizing convolutional neural networks and obtains classification results.
10. the system utilizing neural fusion vehicle classification according to claim 9, is characterized in that: in described output unit, classification results comprises: for represent car first kind classification results, for represent passenger vehicle Equations of The Second Kind classification results, for represent truck the 3rd class classification results, for representing the 4th class classification results of offroad vehicle and the 5th class classification results for representing non-vehicle.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410790026.0A CN104537387A (en) | 2014-12-16 | 2014-12-16 | Method and system for classifying automobile types based on neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201410790026.0A CN104537387A (en) | 2014-12-16 | 2014-12-16 | Method and system for classifying automobile types based on neural network |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN104537387A true CN104537387A (en) | 2015-04-22 |
Family
ID=52852907
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201410790026.0A Pending CN104537387A (en) | 2014-12-16 | 2014-12-16 | Method and system for classifying automobile types based on neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104537387A (en) |
Cited By (26)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104966104A (en) * | 2015-06-30 | 2015-10-07 | 孙建德 | Three-dimensional convolutional neural network based video classifying method |
| CN105069472A (en) * | 2015-08-03 | 2015-11-18 | 电子科技大学 | Vehicle detection method based on convolutional neural network self-adaption |
| CN105550701A (en) * | 2015-12-09 | 2016-05-04 | 福州华鹰重工机械有限公司 | Real-time image extraction and recognition method and device |
| CN105938560A (en) * | 2016-03-23 | 2016-09-14 | 吉林大学 | Convolutional-neural-network-based vehicle model refined classification system |
| CN106156750A (en) * | 2016-07-26 | 2016-11-23 | 浙江捷尚视觉科技股份有限公司 | A kind of based on convolutional neural networks to scheme to search car method |
| CN106203330A (en) * | 2016-07-08 | 2016-12-07 | 西安理工大学 | A kind of vehicle classification method based on convolutional neural networks |
| CN106886023A (en) * | 2017-02-27 | 2017-06-23 | 中国人民解放军理工大学 | A kind of Radar Echo Extrapolation method based on dynamic convolutional neural networks |
| CN106934319A (en) * | 2015-12-29 | 2017-07-07 | 北京大唐高鸿软件技术有限公司 | People's car objective classification method in monitor video based on convolutional neural networks |
| CN107025642A (en) * | 2016-01-27 | 2017-08-08 | 百度在线网络技术（北京）有限公司 | Vehicle's contour detection method and device based on cloud data |
| CN107292333A (en) * | 2017-06-05 | 2017-10-24 | 浙江工业大学 | A kind of rapid image categorization method based on deep learning |
| CN108052861A (en) * | 2017-11-08 | 2018-05-18 | 北京卓视智通科技有限责任公司 | A kind of nerve network system and the model recognizing method based on the nerve network system |
| CN108229523A (en) * | 2017-04-13 | 2018-06-29 | 深圳市商汤科技有限公司 | Image detection, neural network training method, device and electronic equipment |
| CN108268882A (en) * | 2016-12-30 | 2018-07-10 | 南京烽火软件科技有限公司 | A kind of Internet picture scene classification method and its system |
| CN108267724A (en) * | 2018-01-19 | 2018-07-10 | 中国人民解放军陆军装甲兵学院 | A kind of unknown object recognition methods of radar target recognition |
| CN108764314A (en) * | 2018-05-17 | 2018-11-06 | 北京邮电大学 | A kind of structural data sorting technique, device, electronic equipment and storage medium |
| CN109409337A (en) * | 2018-11-30 | 2019-03-01 | 公安部交通管理科学研究所 | Muck vehicle feature identification method based on convolutional neural network |
| CN109670450A (en) * | 2018-12-20 | 2019-04-23 | 天津天地伟业信息系统集成有限公司 | A kind of people's vehicle object detecting method based on video |
| CN109948610A (en) * | 2019-03-14 | 2019-06-28 | 西南交通大学 | A kind of vehicle fine grit classification method in the video based on deep learning |
| WO2019144865A1 (en) * | 2018-01-26 | 2019-08-01 | Mediatek Inc. | Method and apparatus of neural networks with grouping for video coding |
| CN110222641A (en) * | 2019-06-06 | 2019-09-10 | 北京百度网讯科技有限公司 | The method and apparatus of image for identification |
| CN110222748A (en) * | 2019-05-27 | 2019-09-10 | 西南交通大学 | OFDM Radar Signal Recognition method based on the fusion of 1D-CNN multi-domain characteristics |
| CN110555347A (en) * | 2018-06-01 | 2019-12-10 | 杭州海康威视数字技术股份有限公司 | Vehicle target identification method and device with dangerous cargo carrying behavior and electronic equipment |
| CN111178326A (en) * | 2020-01-15 | 2020-05-19 | 山东易构软件技术股份有限公司 | Vehicle type identification method based on T-nearest neighbor spectral clustering under heterogeneous computing model |
| CN111507362A (en) * | 2019-01-30 | 2020-08-07 | 中强光电股份有限公司 | Neural network training method, neural network-based classification method and device |
| CN113822386A (en) * | 2021-11-24 | 2021-12-21 | 苏州浪潮智能科技有限公司 | Image identification method, device, equipment and medium |
| CN115761659A (en) * | 2023-01-09 | 2023-03-07 | 南京隼眼电子科技有限公司 | Recognition model construction method, vehicle type recognition method, electronic device, and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6125194A (en) * | 1996-02-06 | 2000-09-26 | Caelum Research Corporation | Method and system for re-screening nodules in radiological images using multi-resolution processing, neural network, and image processing |
| US20080088887A1 (en) * | 2006-10-11 | 2008-04-17 | Canon Kabushiki Kaisha | Image reading apparatus and method for adjusting same |
| CN101867699A (en) * | 2010-05-25 | 2010-10-20 | 中国科学技术大学 | Real-time tracking method of nonspecific target based on partitioning |
| CN103778432A (en) * | 2014-01-08 | 2014-05-07 | 南京邮电大学 | Human being and vehicle classification method based on deep belief net |
| CN104103033A (en) * | 2014-08-05 | 2014-10-15 | 四川九成信息技术有限公司 | Image real-time processing method |
-
2014
- 2014-12-16 CN CN201410790026.0A patent/CN104537387A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6125194A (en) * | 1996-02-06 | 2000-09-26 | Caelum Research Corporation | Method and system for re-screening nodules in radiological images using multi-resolution processing, neural network, and image processing |
| US20080088887A1 (en) * | 2006-10-11 | 2008-04-17 | Canon Kabushiki Kaisha | Image reading apparatus and method for adjusting same |
| CN101867699A (en) * | 2010-05-25 | 2010-10-20 | 中国科学技术大学 | Real-time tracking method of nonspecific target based on partitioning |
| CN103778432A (en) * | 2014-01-08 | 2014-05-07 | 南京邮电大学 | Human being and vehicle classification method based on deep belief net |
| CN104103033A (en) * | 2014-08-05 | 2014-10-15 | 四川九成信息技术有限公司 | Image real-time processing method |
Non-Patent Citations (3)
| Title |
|---|
| 吴成东 等: "基于图像分块和RBF神经网络的人脸识别方法", 《沈阳建筑大学学报（自然科学版）》 * |
| 周红晓: "基于神经网络的汽车车型识别器的设计", 《工业控制计算机》 * |
| 龚丁禧 等: "基于卷积神经网络的植物叶片分类", 《计算机与现代化》 * |
Cited By (43)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104966104A (en) * | 2015-06-30 | 2015-10-07 | 孙建德 | Three-dimensional convolutional neural network based video classifying method |
| CN104966104B (en) * | 2015-06-30 | 2018-05-11 | 山东管理学院 | A kind of video classification methods based on Three dimensional convolution neutral net |
| CN105069472A (en) * | 2015-08-03 | 2015-11-18 | 电子科技大学 | Vehicle detection method based on convolutional neural network self-adaption |
| CN105069472B (en) * | 2015-08-03 | 2018-07-27 | 电子科技大学 | A kind of vehicle checking method adaptive based on convolutional neural networks |
| CN105550701A (en) * | 2015-12-09 | 2016-05-04 | 福州华鹰重工机械有限公司 | Real-time image extraction and recognition method and device |
| CN105550701B (en) * | 2015-12-09 | 2018-11-06 | 福州华鹰重工机械有限公司 | Realtime graphic extracts recognition methods and device |
| CN106934319A (en) * | 2015-12-29 | 2017-07-07 | 北京大唐高鸿软件技术有限公司 | People's car objective classification method in monitor video based on convolutional neural networks |
| US10229330B2 (en) | 2016-01-27 | 2019-03-12 | Baidu Online Network Technology (Beijing) Co., Ltd. | Method and apparatus for detecting vehicle contour based on point cloud data |
| CN107025642A (en) * | 2016-01-27 | 2017-08-08 | 百度在线网络技术（北京）有限公司 | Vehicle's contour detection method and device based on cloud data |
| CN105938560A (en) * | 2016-03-23 | 2016-09-14 | 吉林大学 | Convolutional-neural-network-based vehicle model refined classification system |
| CN105938560B (en) * | 2016-03-23 | 2017-03-22 | 吉林大学 | Convolutional-neural-network-based vehicle model refined classification system |
| CN106203330A (en) * | 2016-07-08 | 2016-12-07 | 西安理工大学 | A kind of vehicle classification method based on convolutional neural networks |
| CN106156750A (en) * | 2016-07-26 | 2016-11-23 | 浙江捷尚视觉科技股份有限公司 | A kind of based on convolutional neural networks to scheme to search car method |
| CN108268882A (en) * | 2016-12-30 | 2018-07-10 | 南京烽火软件科技有限公司 | A kind of Internet picture scene classification method and its system |
| CN106886023B (en) * | 2017-02-27 | 2019-04-02 | 中国人民解放军理工大学 | A kind of Radar Echo Extrapolation method based on dynamic convolutional neural networks |
| CN106886023A (en) * | 2017-02-27 | 2017-06-23 | 中国人民解放军理工大学 | A kind of Radar Echo Extrapolation method based on dynamic convolutional neural networks |
| CN108229523A (en) * | 2017-04-13 | 2018-06-29 | 深圳市商汤科技有限公司 | Image detection, neural network training method, device and electronic equipment |
| CN107292333A (en) * | 2017-06-05 | 2017-10-24 | 浙江工业大学 | A kind of rapid image categorization method based on deep learning |
| CN107292333B (en) * | 2017-06-05 | 2019-11-29 | 浙江工业大学 | A kind of rapid image categorization method based on deep learning |
| CN108052861A (en) * | 2017-11-08 | 2018-05-18 | 北京卓视智通科技有限责任公司 | A kind of nerve network system and the model recognizing method based on the nerve network system |
| CN108267724A (en) * | 2018-01-19 | 2018-07-10 | 中国人民解放军陆军装甲兵学院 | A kind of unknown object recognition methods of radar target recognition |
| WO2019144865A1 (en) * | 2018-01-26 | 2019-08-01 | Mediatek Inc. | Method and apparatus of neural networks with grouping for video coding |
| GB2585517B (en) * | 2018-01-26 | 2022-12-14 | Mediatek Inc | Method and apparatus of neural networks with grouping for video coding |
| GB2585517A (en) * | 2018-01-26 | 2021-01-13 | Mediatek Inc | Method and apparatus of neural networks with grouping for video coding |
| CN108764314A (en) * | 2018-05-17 | 2018-11-06 | 北京邮电大学 | A kind of structural data sorting technique, device, electronic equipment and storage medium |
| CN110555347A (en) * | 2018-06-01 | 2019-12-10 | 杭州海康威视数字技术股份有限公司 | Vehicle target identification method and device with dangerous cargo carrying behavior and electronic equipment |
| CN109409337A (en) * | 2018-11-30 | 2019-03-01 | 公安部交通管理科学研究所 | Muck vehicle feature identification method based on convolutional neural network |
| CN109409337B (en) * | 2018-11-30 | 2022-05-13 | 公安部交通管理科学研究所 | Muck vehicle feature identification method based on convolutional neural network |
| CN109670450A (en) * | 2018-12-20 | 2019-04-23 | 天津天地伟业信息系统集成有限公司 | A kind of people's vehicle object detecting method based on video |
| CN109670450B (en) * | 2018-12-20 | 2023-07-25 | 天津天地伟业信息系统集成有限公司 | Video-based man-vehicle object detection method |
| CN111507362A (en) * | 2019-01-30 | 2020-08-07 | 中强光电股份有限公司 | Neural network training method, neural network-based classification method and device |
| CN111507362B (en) * | 2019-01-30 | 2023-12-19 | 中强光电股份有限公司 | Neural network training method, neural network-based classification method and device |
| CN109948610B (en) * | 2019-03-14 | 2022-11-18 | 西南交通大学 | Deep learning-based method for classifying fine granularity of vehicles in video |
| CN109948610A (en) * | 2019-03-14 | 2019-06-28 | 西南交通大学 | A kind of vehicle fine grit classification method in the video based on deep learning |
| CN110222748A (en) * | 2019-05-27 | 2019-09-10 | 西南交通大学 | OFDM Radar Signal Recognition method based on the fusion of 1D-CNN multi-domain characteristics |
| CN110222748B (en) * | 2019-05-27 | 2022-12-20 | 西南交通大学 | OFDM radar signal identification method based on 1D-CNN multi-domain feature fusion |
| CN110222641B (en) * | 2019-06-06 | 2022-04-19 | 北京百度网讯科技有限公司 | Method and apparatus for recognizing image |
| CN110222641A (en) * | 2019-06-06 | 2019-09-10 | 北京百度网讯科技有限公司 | The method and apparatus of image for identification |
| CN111178326A (en) * | 2020-01-15 | 2020-05-19 | 山东易构软件技术股份有限公司 | Vehicle type identification method based on T-nearest neighbor spectral clustering under heterogeneous computing model |
| CN113822386A (en) * | 2021-11-24 | 2021-12-21 | 苏州浪潮智能科技有限公司 | Image identification method, device, equipment and medium |
| CN113822386B (en) * | 2021-11-24 | 2022-02-22 | 苏州浪潮智能科技有限公司 | Image identification method, device, equipment and medium |
| CN115761659A (en) * | 2023-01-09 | 2023-03-07 | 南京隼眼电子科技有限公司 | Recognition model construction method, vehicle type recognition method, electronic device, and storage medium |
| CN115761659B (en) * | 2023-01-09 | 2023-04-11 | 南京隼眼电子科技有限公司 | Recognition model construction method, vehicle type recognition method, electronic device, and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104537387A (en) | Method and system for classifying automobile types based on neural network | |
| CN104036323B (en) | A kind of vehicle checking method based on convolutional neural networks | |
| CN107133974B (en) | Gaussian Background models the vehicle type classification method combined with Recognition with Recurrent Neural Network | |
| CN104504395A (en) | Method and system for achieving classification of pedestrians and vehicles based on neural network | |
| Khaing et al. | Development of control system for fruit classification based on convolutional neural network | |
| CN103218831B (en) | A kind of video frequency motion target classifying identification method based on profile constraint | |
| CN105184271A (en) | Automatic vehicle detection method based on deep learning | |
| CN105550701A (en) | Real-time image extraction and recognition method and device | |
| CN108460328A (en) | A kind of fake-licensed car detection method based on multitask convolutional neural networks | |
| CN103198332A (en) | Real-time robust far infrared vehicle-mounted pedestrian detection method | |
| CN104517103A (en) | Traffic sign classification method based on deep neural network | |
| CN102915453B (en) | Real-time feedback and update vehicle detection method | |
| CN106776943A (en) | A kind of vehicle retrieval method based on AutoEncoder and attribute tags | |
| CN105260715B (en) | Towards the toy object detection method in remote location | |
| CN105335716A (en) | Improved UDN joint-feature extraction-based pedestrian detection method | |
| CN102819841B (en) | Global threshold partitioning method for partitioning target image | |
| CN111274886B (en) | Deep learning-based pedestrian red light running illegal behavior analysis method and system | |
| CN112434723B (en) | Day/night image classification and object detection method based on attention network | |
| CN106056102A (en) | Video-image-analysis-based road vehicle type classification method | |
| CN107092884A (en) | Rapid coarse-fine cascade pedestrian detection method | |
| Shangzheng | A traffic sign image recognition and classification approach based on convolutional neural network | |
| Goyal et al. | A neural network based approach for the vehicle classification | |
| Deepan et al. | Road recognition from remote sensing imagery using machine learning | |
| He et al. | Real-time pedestrian warning system on highway using deep learning methods | |
| Ganakwar et al. | Convolutional neural network-VGG16 for road extraction from remotely sensed images |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20150422 |