CN110991377B - Front mesh identification method of automobile safety auxiliary system based on monocular vision neural network - Google Patents
Front mesh identification method of automobile safety auxiliary system based on monocular vision neural network Download PDFInfo
- Publication number
- CN110991377B CN110991377B CN201911263299.9A CN201911263299A CN110991377B CN 110991377 B CN110991377 B CN 110991377B CN 201911263299 A CN201911263299 A CN 201911263299A CN 110991377 B CN110991377 B CN 110991377B
- Authority
- CN
- China
- Prior art keywords
- region
- neural network
- basis function
- layer
- radial basis
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 41
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 40
- 239000013598 vector Substances 0.000 claims abstract description 14
- 238000003062 neural network model Methods 0.000 claims abstract description 9
- 230000011218 segmentation Effects 0.000 claims abstract description 5
- 238000012549 training Methods 0.000 claims description 14
- 238000000605 extraction Methods 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 11
- 238000003709 image segmentation Methods 0.000 claims description 5
- 238000012546 transfer Methods 0.000 claims description 4
- 238000003064 k means clustering Methods 0.000 claims description 3
- 230000006870 function Effects 0.000 description 35
- 238000001514 detection method Methods 0.000 description 13
- 102100034112 Alkyldihydroxyacetonephosphate synthase, peroxisomal Human genes 0.000 description 3
- 101000799143 Homo sapiens Alkyldihydroxyacetonephosphate synthase, peroxisomal Proteins 0.000 description 3
- 238000000848 angular dependent Auger electron spectroscopy Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000001444 catalytic combustion detection Methods 0.000 description 2
- 238000012706 support-vector machine Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 206010039203 Road traffic accident Diseases 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000013527 convolutional neural network Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 210000002364 input neuron Anatomy 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000007500 overflow downdraw method Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- 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
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- 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
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/26—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion
- G06V10/267—Segmentation of patterns in the image field; Cutting or merging of image elements to establish the pattern region, e.g. clustering-based techniques; Detection of occlusion by performing operations on regions, e.g. growing, shrinking or watersheds
-
- 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
- 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)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Multimedia (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Computational Linguistics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Evolutionary Biology (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Probability & Statistics with Applications (AREA)
- Image Analysis (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a method for identifying a front target of an automobile safety auxiliary system based on a radial basis function neural network, which comprises the following steps: acquiring road condition images in front of a vehicle, and carrying out segmentation pretreatment on the road condition images; extracting edges of the preprocessed road condition images, and searching to obtain an interested region; extracting features of the region of interest to obtain edge features and region features corresponding to the region of interest; constructing a radial basis neural network model by taking the edge features and the region features as input layer vectors, and analyzing the input layer vector features in a neural network to obtain output quantity related to a target; according to the output quantity, a corresponding vehicle target is obtained, and the vehicle target is output as a recognition result.
Description
Technical Field
The invention relates to the field of automobile safety auxiliary driving control, in particular to a front destination identification method of an automobile safety auxiliary system based on a monocular vision neural network.
Background
The automobile safety auxiliary driving system (ADAS) is used for rapidly and accurately extracting information such as vehicles or obstacles in front of the system by using radar or machine vision, and can prompt a driver to avoid collision danger or automatically control the vehicles to realize the early warning or collision avoidance function. The system function is not only suitable for expressway driving conditions, but also particularly important for identifying targets such as non-motor vehicles, pedestrians, obstacles and the like in front of an ADAS system and particularly identifying electric bicycles for other driving environments, especially urban road driving environments, because traffic accident rates of the non-motor vehicles (mainly electric bicycles) and the motor vehicles frequently occur and the proportion is large in urban roads.
Currently, there are few studies on classification of motor vehicles and electric bicycles based on a machine vision detection and identification method. Most methods are directed to detection and identification studies of vehicles and are used in auxiliary driving systems. Such as by using linear geometric feature information of the vehicle, symmetry of the vehicle, or computer vision methods employing special hardware such as color CCDs and binocular CCDs, etc. In addition, there are an optical flow-based method, a template matching method, a support vector machine method, a neural network training method, a multi-sensor information fusion method, a vehicle detection and recognition method based on an AdaBoost method and a support vector machine classifier, and a vehicle detection or recognition method based on deep learning and a high-speed-area convolutional neural network. The research method for vehicle detection has the end result that the region of interest of the vehicle is determined, but the region still has the possibility of false detection, and if the detection result in the region of interest can be deeply confirmed, the accuracy of vehicle target identification can be greatly improved, and the false detection rate is reduced to increase the reliability of system identification.
Disclosure of Invention
The invention designs and develops a front mesh identification method of an automobile safety auxiliary system based on a radial basis function neural network, which performs edge extraction and search on preprocessed road condition images to obtain a region of interest as a possible vehicle region, and constructs a radial basis function neural network vehicle identifier by counting edge characteristics and region characteristic parameters of the vehicle to realize classification of the vehicle and the electric bicycle in a detection region.
The technical scheme provided by the invention is as follows:
a method for identifying the front destination of an automobile safety auxiliary system based on a radial basis function neural network, comprising the following steps:
acquiring road condition images in front of a vehicle, and carrying out segmentation pretreatment on the road condition images;
extracting edges of the preprocessed road condition images, and searching to obtain an interested region;
extracting features of the region of interest to obtain edge features and region features corresponding to the region of interest;
constructing a radial basis neural network model by taking the edge features and the region features as input layer vectors, and analyzing the input layer vector features in a neural network to obtain output quantity related to a target;
obtaining a corresponding vehicle target according to the output quantity, and outputting the vehicle target as a recognition result;
wherein the region of interest includes electric bicycles and motor vehicles.
Preferably, the edge features and the region features corresponding to the region of interest include: edge characteristic parameters and region description characteristic parameters formed by independent invariant moment parameters of sub-coefficients of discrete cosine transform.
Preferably, the discrete cosine transform sub-coefficient calculation formula is:
C(k)=|F(k)|/F(1);
wherein C (k) is a discrete cosineThe sub-coefficients of the transform are used,k is the number of discrete sub-coefficients, k=1, 2 …; f (k) =x (k) +jy (k); /> j is the imaginary part of the complex plane n=1, 2,3 … N-1; n is the characteristic point variable of the closed edge curve obtained by edge extraction after image segmentation, N is the number of characteristic points of the closed edge curve obtained by edge extraction after image segmentation, and f (m) =x (m) +jy (m); m is more than or equal to 1 and less than or equal to n, and f (m) is a one-dimensional complex sequence.
Preferably, the independent invariant moment parameter calculation formula is:
wherein ,mu, the coordinates of the central point of the region pq The central moment of the area where the binarized image is located; />m 00 Is zero order geometrical moment, m of the area where the binarized image is located 01 、m 10 Is the first-order geometric moment, m, of the area where the binarized image is located pq The geometric moment of the p+q order of the area where the binary image is located, p is the row order of the central moment of the binary image, and q is the column order of the central moment of the binary image.
Preferably, the target recognition characteristic parameters include: region eccentricity, ratio of short and long axes of the region, region area, region perimeter, and region compactness factor.
Preferably, the radial basis function neural network model is a three-layer neural network model:
the first layer is an input layer, and the feature vector is input into the network;
the second layer is a hidden layer which can be completely connected with the input layer, and the hidden layer node selects a Gaussian radial basis function as a transfer function, and the calculation formula is as follows:
in the formula ,||xp -c i I is the European norm, c i Is the center of the Gaussian function, and sigma is the variance of the Gaussian function;
the third layer is an output layer, 2 output quantities are obtained by calculating weights between the hidden layer and the output layer, and a vehicle target is identified; the output of the network can be obtained from the structure of the radial basis function network as:
in the formula ,for the P-th input sample, p=1, 2, …, P is the total number of samples; c i Omega is the center of hidden layer node of network ij For the connection weight from the hidden layer to the output layer, i=1, 2, …, h is the number of nodes of the hidden layer, y j J=1, 2, …, n for the actual output of the j-th output node of the network for the input sample pair;
wherein ,dj Is the expected output value of the sample.
Preferably, the center of the radial basis function neural network is obtained by a K-means clustering algorithm, and the specific process is as follows:
step one, randomly selecting h training samples as a clustering center c i ,i=1,2,…h;
Step two, calculating the Euclidean distance between each training sample and the clustering center, and according to the Euclidean distanceDistance assigns each training sample to a respective cluster set ψ of input samples p (p=1, 2, …, P);
step three, recalculating each cluster set psi p Average value of training samples to obtain a new clustering center c i ′;
Step four, repeating the step two and the step three until a new clustering center c is obtained i ' the variation is less than a given threshold, c is obtained i ' is the final basis function center of the radial basis function neural network.
Preferably, the basis function variance solving formula is:
wherein ,σi As the basis function variance, c max To select the maximum distance between centers.
Preferably, the connection weight between the hidden layer and the output layer is calculated by a least square method, and the calculation formula is as follows:
the beneficial effects of the invention are that
The invention designs and develops a front target identification method of an automobile safety auxiliary system based on a radial basis function neural network, carries out edge extraction and search on a preprocessed road condition image, obtains a region of interest as a possible existing vehicle region, constructs a radial basis function neural network vehicle identifier by counting edge characteristics and region characteristic parameters of the vehicle, realizes classification of the vehicle and an electric bicycle in a detection region, greatly improves accuracy of vehicle target identification, and reduces false detection rate so as to improve reliability of system identification.
Drawings
Fig. 1 is a road condition image of the front of a vehicle according to the present invention.
Fig. 2 is a schematic diagram of a vehicle searching region of interest according to the present invention.
Fig. 3 is a block diagram of an RBF neural network according to the present invention.
FIG. 4 is a schematic diagram of an error performance curve of an RBF neural network test according to the present invention.
Detailed Description
The present invention is described in further detail below with reference to the drawings to enable those skilled in the art to practice the invention by referring to the description.
As shown in fig. 1, the method for identifying the front mesh of the automobile safety auxiliary system based on the radial basis function neural network provided by the invention comprises the following steps:
acquiring road condition images in front of a vehicle, and carrying out segmentation pretreatment on the road condition images;
extracting edges of the preprocessed road condition images, and searching to obtain an interested region;
as shown in fig. 2, extracting vehicle features from the region of interest to obtain edge features and region features corresponding to the region of interest;
constructing a radial basis neural network model by taking the edge features and the region features as input layer vectors, and analyzing the input layer vector features in a neural network to obtain output quantity related to a target;
and obtaining a corresponding vehicle target according to the output quantity, and outputting the vehicle target as a recognition result.
In order to accurately identify a vehicle target, it is necessary to extract vehicle-specific features from the detected image area. Feature extraction is the process of extracting and selecting information representing the features of a target from the original information of the target. At present, methods for identifying the shape of a target can be mainly classified into two types: one is based on the recognition of the shape of the edge of the object, i.e. the edge features. And secondly, identifying the shape of the area covered by the object, namely the area characteristic.
Feature extraction of the object is to take care of the following basic requirements:
(1) The extracted and selected features are insensitive to the variable parameters of the target;
(2) The characteristics are stable and easy to extract;
(3) The dimension of the feature quantity is obviously smaller than the original data of the target;
(4) The feature quantities have the smallest correlation;
(5) To improve classification accuracy, features are added that easily separate confusing categories.
Based on the above feature extraction requirements, the invention extracts the hybrid features of the vehicle (including the electric bicycle) and simultaneously includes edge features and regional features, namely 8 sub-coefficients of discrete cosine transform, edge feature parameters formed by 6 independent invariant moment parameters and 5 regional description feature parameters, and the total number of the edge feature parameters is 19.
Preferably, the discrete cosine transform sub-coefficient operation process is as follows:
the target image is cut, preprocessed and edge extracted to obtain outline data f (x m, y m), the closed edge curve formed by N points is put on complex plane to form one-dimensional complex sequence,
f(m)=x(m)+jy(m);1≤m≤n,
f (m) is a one-dimensional complex sequence; the discrete cosine transform of the formula is as follows:
calculating to obtain discrete cosine transform sub-coefficients:
C(k)=|F(k)|/F(1);
wherein C (k) is a discrete cosine transform sub-coefficient; k is the number of discrete sub-coefficients, k=1, 2 …; j is the imaginary part of the complex plane n=1, 2,3 … N-1; n is the characteristic point variable of the closed edge curve obtained by edge extraction after image segmentation, and N is the number of the characteristic points of the closed edge curve obtained by edge extraction after image segmentation.
Discrete cosine transform coefficients have translational, rotational and proportional invariance to the target and are insensitive to the starting point of the contour data. The low frequency part of the cosine transform coefficient reflects the whole outline of the image, the high frequency part only represents the detail of the outline, and the discrete cosine transform does not need complex operation and data modulo operation, and can obtain higher recognition rate by using fewer characteristic quantities.
If a digital image satisfies the segmentation continuity and there are only a limited number of zeros in the XY plane, it can be demonstrated that the moments of the orders of the digital image exist.
For a binary image, since the values of the pixels are only 0 and 1, assuming that the pixel value of the target area is 1 and the pixel value of the background area is 0, the p+q moment of the binary image is as shown in the formula:
the central moment of this region is:
wherein ,mu, the coordinates of the central point of the region pq The central moment of the area where the binarized image is located; />m 00 Is zero order geometrical moment, m of the area where the binarized image is located 01 、m 10 Is the first-order geometric moment, m, of the area where the binarized image is located pq The geometric moment of the p+q order of the area where the binary image is located, p is the row order of the central moment of the binary image, and q is the column order of the central moment of the binary image.
The target recognition characteristic parameters include: the area eccentricity, which is the ratio of the length of the major axis to the length of the minor axis of the boundary, i.e., the eccentricity of an ellipse having the same second moment as the area, is the ratio of the distance between the focal length major axes of the ellipse. Ratio of short axis to long axis of region, region area S, region perimeter L, and region compactness factor 4πS/L 2 。
Radial basis function neural networks are a novel neural network learning method that expands or preprocesses input vectors into a high-dimensional space. The method has good popularization capability, avoids complex calculation such as BP algorithm, and can realize the rapid learning of the neural network.
As shown in FIG. 3, the invention adopts RBF neural network of self-organizing selection center, and takes the edge characteristic parameters and 5 region description characteristic parameters formed by 8 discrete cosine transform coefficients, 6 independent invariant moment parameters as input vectors of the network. The discrimination result of whether the vehicle is or is not required to be recognized. Thus, the designed RBF neural network has 19-dimensional input neurons and 2 outputs.
The radial basis function neural network model is a three-layer neural network model: the first layer is an input layer, and feature vectors are input into the network. The second layer is a hidden layer, which is completely connected with the input layer (weight=1), and is equivalent to performing one-time conversion on the input mode, and converting the low-dimensional mode input data into a high-dimensional space so as to facilitate classification and identification of the output layer. Here the hidden layer node selects a gaussian radial basis function as the transfer function. The third layer is an output layer, and 2 output quantities are obtained by calculating weights between the hidden layer and the output layer, so that a vehicle target is identified.
The RBF neural network learning method of the self-organizing selection center mainly comprises two stages. Stage one: a self-organizing learning stage, namely a learning process without a teacher, and solving hidden layer basis functions; stage two: there is a teacher learning phase, i.e. solving weights between hidden layers to output layers.
The first layer is an input layer, and the feature vector is input into the network;
the second layer is a hidden layer which can be completely connected with the input layer, and the hidden layer node selects a Gaussian radial basis function as a transfer function, and the calculation formula is as follows:
in the formula ,||xp -c i I is the European norm, c i Is the center of Gaussian functionSigma is the variance of the gaussian function;
the third layer is an output layer, 2 output quantities are obtained by calculating weights between the hidden layer and the output layer, and a vehicle target is identified; the output of the network can be obtained from the structure of the radial basis function network as:
in the formula ,for the P-th input sample, p=1, 2, …, P is the total number of samples; c i Omega is the center of hidden layer node of network ij For the connection weight from the hidden layer to the output layer, i=1, 2, …, h is the number of nodes of the hidden layer, y j J=1, 2, …, n for the actual output of the j-th output node of the network for the input sample pair;
wherein ,dj Is the expected output value of the sample.
The center of the radial basis function neural network is obtained by a K-means clustering algorithm, and the specific process is as follows:
step one, randomly selecting h training samples as a clustering center c i ,i=1,2,…h;
Step two, calculating the Euclidean distance between each training sample and the clustering center, and distributing each training sample to each clustering set psi of the input samples according to the Euclidean distance p (p=1, 2, …, P);
step three, recalculating each cluster set psi p Average value of training samples to obtain a new clustering center c i ′;
Step four, repeating the step two and the step three until a new clustering center c is obtained i ' the variation is less than a given threshold, c is obtained i ' is the final basis function center of the radial basis function neural network.
The basis function variance solving formula is:
wherein ,σi As the basis function variance, c max To select the maximum distance between centers.
The connection weight between the hidden layer and the output layer is calculated by a least square method, and the calculation formula is as follows:
as shown in fig. 4, in order to verify the effectiveness of the method for identifying the front mesh of the ADAS system based on the RBF neural network, 850 vehicle samples and 850 electric bicycle samples are respectively established, the RBF neural network is trained, 60% of positive and negative sample images are randomly selected to complete the test, and the identification accuracy of the RBF neural network can reach more than 94%. The error performance curve of the RBF neural network can be used for finding that the designed network meets the requirement of training errors.
The invention designs and develops a front target identification method of an automobile safety auxiliary system based on a radial basis function neural network, carries out edge extraction and search on a preprocessed road condition image, obtains a region of interest as a possible existing vehicle region, constructs a radial basis function neural network vehicle identifier by counting edge characteristics and region characteristic parameters of the vehicle, realizes classification of the vehicle and an electric bicycle in a detection region, greatly improves accuracy of vehicle target identification, and reduces false detection rate so as to improve reliability of system identification.
Although embodiments of the present invention have been disclosed above, it is not limited to the details and embodiments shown and described, it is well suited to various fields of use, and further modifications may be readily made by those skilled in the art without departing from the general concepts defined by the claims and the equivalents thereof, and therefore the invention is not limited to the specific details and examples shown and described herein.
Claims (7)
1. A method for identifying the front destination of an automobile safety auxiliary system based on a radial basis function neural network, which is characterized by comprising the following steps:
acquiring road condition images in front of a vehicle, and carrying out segmentation pretreatment on the road condition images;
extracting edges of the preprocessed road condition images, and searching to obtain an interested region;
extracting features of the region of interest to obtain edge features and region features corresponding to the region of interest;
constructing a radial basis neural network model by taking the edge features and the region features as input layer vectors, and analyzing the input layer vector features in a neural network to obtain output quantity related to a target;
obtaining a corresponding vehicle target according to the output quantity, and outputting the vehicle target as a recognition result;
the region of interest includes electric bicycles and motor vehicles;
the vehicle edge features and the region features corresponding to the region of interest include: edge characteristic parameters and region description characteristic parameters formed by independent invariant moment parameters of sub-coefficients of discrete cosine transform;
the discrete cosine transform sub-coefficient calculation formula is:
C(k)=|F(k)|/F(1);
wherein C (k) is a discrete cosine transform sub-coefficient,k is the number of discrete sub-coefficients, k=1, 2 …; f (k) =x (k) +jy (k);
j is the imaginary part of the complex plane n=1, 2,3 … N-1; n is a characteristic point variable of a closed edge curve obtained by edge extraction after image segmentationN is the number of feature points of a closed edge curve obtained by edge extraction after image segmentation, f (m) =x (m) +jy (m); m is more than or equal to 1 and less than or equal to n, and f (m) is a one-dimensional complex sequence.
2. The method for identifying the front destination of an automobile safety auxiliary system based on a radial basis function neural network according to claim 1, wherein the independent invariant moment parameter calculation formula is:
wherein ,mu, the coordinates of the central point of the region pq The central moment of the area where the binarized image is located; />m 00 Is zero order geometrical moment, m of the area where the binarized image is located 01 、m 10 Is the first-order geometric moment, m, of the area where the binarized image is located pq The geometric moment of the p+q order of the area where the binary image is located, p is the row order of the central moment of the binary image, and q is the column order of the central moment of the binary image.
3. The method for identifying a front target of an automotive safety assistance system based on a radial basis function network according to claim 2, wherein the target identification characteristic parameters include: region eccentricity, ratio of short axis to long axis of the region, region area, region perimeter, and region compactness factor.
4. A method for identifying a front target of an automotive safety assistance system based on a radial basis function network according to claim 3, wherein the radial basis function network model is a three-layer neural network model:
the first layer is an input layer, and the feature vector is input into the network;
the second layer is a hidden layer which can be completely connected with the input layer, and the hidden layer node selects a Gaussian radial basis function as a transfer function, and the calculation formula is as follows:
in the formula ,||xp -c i I is the European norm, c i Is the center of the Gaussian function, and sigma is the variance of the Gaussian function;
the third layer is an output layer, 2 output quantities are obtained by calculating weights between the hidden layer and the output layer, and a vehicle target is identified; the output of the network can be obtained from the structure of the radial basis function network as:
in the formula ,for the P-th input sample, p=1, 2, …, P is the total number of samples; c i Omega is the center of hidden layer node of network ij For the connection weight from the hidden layer to the output layer, i=1, 2, …, h is the number of nodes of the hidden layer, y j J=1, 2, …, n for the actual output of the j-th output node of the network for the input sample pair;
wherein ,dj Is the expected output value of the sample.
5. The method for identifying the front mesh of the automobile safety auxiliary system based on the radial basis function neural network according to claim 4, wherein the center of the radial basis function neural network is obtained by a K-means clustering algorithm, and the specific process is as follows:
step one, randomly selecting h training aidsTraining samples as cluster center c i ,i=1,2,…h;
Step two, calculating the Euclidean distance between each training sample and the clustering center, and distributing each training sample to each clustering set psi of the input samples according to the Euclidean distance p (p=1, 2, …, P);
step three, recalculating each cluster set psi p Average value of training samples to obtain a new clustering center c i ′;
Step four, repeating the step two and the step three until a new clustering center c is obtained i ' the variation is less than a given threshold, c is obtained i ' is the final basis function center of the radial basis function neural network.
6. The method for identifying the front mesh of an automotive safety auxiliary system based on a radial basis function neural network according to claim 5, wherein the basis function variance solving formula is:
wherein ,σi As the basis function variance, c max To select the maximum distance between centers.
7. The method for identifying the front mesh of the automobile safety auxiliary system based on the radial basis function neural network according to claim 6, wherein the connection weight between the hidden layer and the output layer is calculated by a least square method, and the calculation formula is as follows:
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911263299.9A CN110991377B (en) | 2019-12-11 | 2019-12-11 | Front mesh identification method of automobile safety auxiliary system based on monocular vision neural network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201911263299.9A CN110991377B (en) | 2019-12-11 | 2019-12-11 | Front mesh identification method of automobile safety auxiliary system based on monocular vision neural network |
Publications (2)
Publication Number | Publication Date |
---|---|
CN110991377A CN110991377A (en) | 2020-04-10 |
CN110991377B true CN110991377B (en) | 2023-09-19 |
Family
ID=70092204
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201911263299.9A Active CN110991377B (en) | 2019-12-11 | 2019-12-11 | Front mesh identification method of automobile safety auxiliary system based on monocular vision neural network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN110991377B (en) |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN111862608B (en) * | 2020-07-22 | 2022-06-24 | 湖北文理学院 | Vehicle driving road condition identification method, device, equipment and storage medium |
CN112558510B (en) * | 2020-10-20 | 2022-11-15 | 山东亦贝数据技术有限公司 | Intelligent networking automobile safety early warning system and early warning method |
CN114092701B (en) * | 2021-12-04 | 2022-06-03 | 特斯联科技集团有限公司 | Intelligent symbol identification method based on neural network |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101814160A (en) * | 2010-03-08 | 2010-08-25 | 清华大学 | RBF neural network modeling method based on feature clustering |
CN103020582A (en) * | 2012-09-20 | 2013-04-03 | 苏州两江科技有限公司 | Method for computer to identify vehicle type by video image |
CN106056147A (en) * | 2016-05-27 | 2016-10-26 | 大连楼兰科技股份有限公司 | System and method for establishing target division remote damage assessment of different vehicle types based artificial intelligence radial basis function neural network method |
CN106960075A (en) * | 2017-02-27 | 2017-07-18 | 浙江工业大学 | The Forecasting Methodology of the injector performance of RBF artificial neural network based on linear direct-connected method |
WO2018187953A1 (en) * | 2017-04-12 | 2018-10-18 | 邹霞 | Facial recognition method based on neural network |
CN110261436A (en) * | 2019-06-13 | 2019-09-20 | 暨南大学 | Rail deformation detection method and system based on infrared thermal imaging and computer vision |
-
2019
- 2019-12-11 CN CN201911263299.9A patent/CN110991377B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101814160A (en) * | 2010-03-08 | 2010-08-25 | 清华大学 | RBF neural network modeling method based on feature clustering |
CN103020582A (en) * | 2012-09-20 | 2013-04-03 | 苏州两江科技有限公司 | Method for computer to identify vehicle type by video image |
CN106056147A (en) * | 2016-05-27 | 2016-10-26 | 大连楼兰科技股份有限公司 | System and method for establishing target division remote damage assessment of different vehicle types based artificial intelligence radial basis function neural network method |
CN106960075A (en) * | 2017-02-27 | 2017-07-18 | 浙江工业大学 | The Forecasting Methodology of the injector performance of RBF artificial neural network based on linear direct-connected method |
WO2018187953A1 (en) * | 2017-04-12 | 2018-10-18 | 邹霞 | Facial recognition method based on neural network |
CN110261436A (en) * | 2019-06-13 | 2019-09-20 | 暨南大学 | Rail deformation detection method and system based on infrared thermal imaging and computer vision |
Non-Patent Citations (3)
Title |
---|
张小军 等.基于径向基函数神经网络的车型识别技术.西北大学学报(自然科学网络版).2006,第4卷(第2期),1-6. * |
沈凤龙 ; 毕娟 ; .基于多神经网络分类器的汽车车型识别方法研究.辽东学院学报(自然科学版).(第03期),全文. * |
袁艳 ; 叶俊浩 ; 苏丽娟 ; .基于改进的粒子群径向基神经网络的目标识别.计算机应用.(第S1期),全文. * |
Also Published As
Publication number | Publication date |
---|---|
CN110991377A (en) | 2020-04-10 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN110991377B (en) | Front mesh identification method of automobile safety auxiliary system based on monocular vision neural network | |
CN108875608B (en) | Motor vehicle traffic signal identification method based on deep learning | |
CN110263786B (en) | Road multi-target identification system and method based on feature dimension fusion | |
Cao et al. | A low-cost pedestrian-detection system with a single optical camera | |
CN112749616B (en) | Multi-domain neighborhood embedding and weighting of point cloud data | |
JP2016062610A (en) | Feature model creation method and feature model creation device | |
CN112990065B (en) | Vehicle classification detection method based on optimized YOLOv5 model | |
CN108830254B (en) | Fine-grained vehicle type detection and identification method based on data balance strategy and intensive attention network | |
CN115937655B (en) | Multi-order feature interaction target detection model, construction method, device and application thereof | |
CN112381101B (en) | Infrared road scene segmentation method based on category prototype regression | |
CN103620645A (en) | Object recognition device | |
Seeger et al. | Towards road type classification with occupancy grids | |
CN114913498A (en) | Parallel multi-scale feature aggregation lane line detection method based on key point estimation | |
CN114821519A (en) | Traffic sign identification method and system based on coordinate attention | |
CN112559968B (en) | Driving style representation learning method based on multi-situation data | |
Danapal et al. | Sensor fusion of camera and LiDAR raw data for vehicle detection | |
JP5407723B2 (en) | Recognition device, recognition method, and program | |
CN116630702A (en) | Pavement adhesion coefficient prediction method based on semantic segmentation network | |
CN115129886A (en) | Driving scene recognition method and device and vehicle | |
CN114022705B (en) | Self-adaptive target detection method based on scene complexity pre-classification | |
CN114821508A (en) | Road three-dimensional target detection method based on implicit context learning | |
Rafi et al. | Performance analysis of deep learning YOLO models for South Asian regional vehicle recognition | |
Rostami et al. | An Image Dataset of Vehicles Front Views and Parts for Vehicle Detection, Localization and Alignment Applications | |
Guma et al. | Design a hybrid approach for the classification and recognition of traffic signs using machine learning | |
Yalla et al. | CHASE Algorithm:" Ease of Driving" Classification |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |