CN110728050A - Urban traffic accident black spot identification machine learning method fusing multi-source features - Google Patents
Urban traffic accident black spot identification machine learning method fusing multi-source features Download PDFInfo
- Publication number
- CN110728050A CN110728050A CN201910962247.4A CN201910962247A CN110728050A CN 110728050 A CN110728050 A CN 110728050A CN 201910962247 A CN201910962247 A CN 201910962247A CN 110728050 A CN110728050 A CN 110728050A
- Authority
- CN
- China
- Prior art keywords
- traffic accident
- data
- neural network
- deep neural
- black spot
- 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
- 206010039203 Road traffic accident Diseases 0.000 title claims abstract description 61
- 206010027146 Melanoderma Diseases 0.000 title claims abstract description 27
- 238000010801 machine learning Methods 0.000 title claims abstract description 16
- 238000013528 artificial neural network Methods 0.000 claims abstract description 39
- 238000000034 method Methods 0.000 claims abstract description 27
- 238000012706 support-vector machine Methods 0.000 claims abstract description 14
- 230000006870 function Effects 0.000 claims description 23
- 238000012549 training Methods 0.000 claims description 17
- 238000004458 analytical method Methods 0.000 claims description 10
- 238000005457 optimization Methods 0.000 claims description 10
- 239000013598 vector Substances 0.000 claims description 8
- 238000001514 detection method Methods 0.000 claims description 7
- 230000007613 environmental effect Effects 0.000 claims description 3
- 238000007781 pre-processing Methods 0.000 claims description 3
- 238000005516 engineering process Methods 0.000 description 6
- 238000011160 research Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 3
- 208000027418 Wounds and injury Diseases 0.000 description 2
- 230000006378 damage Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000018109 developmental process Effects 0.000 description 2
- 208000014674 injury Diseases 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000007789 sealing Methods 0.000 description 2
- 238000012163 sequencing technique Methods 0.000 description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 2
- OYPRJOBELJOOCE-UHFFFAOYSA-N Calcium Chemical compound [Ca] OYPRJOBELJOOCE-UHFFFAOYSA-N 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 230000001186 cumulative effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012732 spatial analysis Methods 0.000 description 1
- 230000033772 system development Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/40—Business processes related to the transportation industry
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Economics (AREA)
- Strategic Management (AREA)
- Bioinformatics & Computational Biology (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Tourism & Hospitality (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- Evolutionary Biology (AREA)
- Entrepreneurship & Innovation (AREA)
- Educational Administration (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Development Economics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Game Theory and Decision Science (AREA)
- Quality & Reliability (AREA)
- Operations Research (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses an urban traffic accident black spot identification machine learning method fusing multi-source characteristics, and relates to the technical field of traffic control systems for detecting road vehicles with traffic movement to be counted or controlled. The method comprises the following steps: s1, identifying black points of urban traffic accidents based on a support vector machine; s2, identifying black spots of the urban traffic accident based on a deep neural network; the invention utilizes the support vector machine method to improve the accuracy of black spot identification. With the rapid growth of traffic accident multi-source data, the precision of the check rate and the recall rate of the support vector machine black spot identification is gradually reduced, the deep neural network based on the deep neural network is adopted to establish the related data category information for the black spot identification, the accuracy of the black spot identification is improved, and meanwhile, the efficiency of the black spot identification is improved.
Description
Technical Field
The invention relates to the technical field of traffic control systems for detecting road vehicles with traffic movement to be counted or controlled, in particular to a machine learning method for identifying black spots of urban traffic accidents by fusing multi-source characteristics.
Background
With the rapid development of economy in China, the urbanization process is increasingly accelerated, the traffic volume of people is greatly increased, and the demand on traffic is increased. In order to meet the travel demand, the holding amount of motor vehicles of people shows a rapid increase trend, so that the urban traffic accident data is also increased sharply. With the continuous development of computer network technology, geographic information processing technology, information mining technology, spatial analysis and expression technology and information system development technology, many enterprises, universities and scientific research institutes are also conducting analysis research on traffic accident information data. There are many random factors of traffic accidents, but if traffic accidents occur frequently in a certain part of a road and the types of accidents are close, the characteristics and the internal laws of the road should be considered, and the road is generally called as an accident multiple point and multiple section, also called as a black point and a black section.
The traffic accident data includes not only time and place, but also data of people, roads, vehicles, surrounding environment, and the like. Most of the current research work of black spot identification is based on the spatial position discrete degree of the traffic accident, and other factors during the traffic accident are ignored. How to fuse the multi-source characteristics of the traffic accident to accurately identify the position of the traffic accident black spot and the corresponding key influence factor provides scientific support for the decision of traffic management departments, and becomes the key point of the research on the identification of the accident black spot.
Machine learning is to learn a training set composed of a large number of object and non-object samples, and then use the learned template or classifier for object detection. Machine learning enables machines with processors and computing capabilities to improve the performance of processing problems with increased experience using computer programs. Machine learning based on data is an important aspect in modern intelligent technology, researches are carried out to find rules from observed data, and the rules are utilized to predict future data or data which cannot be observed.
The invention discloses a Chinese invention 'urban road network pedestrian traffic accident black point identification method' with publication number CN104392076B, and provides an urban road network pedestrian traffic accident black point identification method, which comprises the steps of establishing a standardized pedestrian traffic accident database for pedestrian traffic accident data in a specified region of a specified time period, calculating the accident frequency of each unit road section, estimating the parameter value of a pedestrian traffic accident distribution model by using the database, obtaining the probability of each accident frequency and the cumulative probability of the accident frequency, determining the upper limit threshold of the pedestrian traffic accident black point under a given confidence level, identifying the pedestrian traffic accident black point, and realizing the spatial positioning and display of the pedestrian traffic accident black point through reverse decoding. The method overcomes the defects of low precision, large subjectivity, poor visibility and the like of the conventional accident black point identification method, and has important engineering application value in the aspects of reducing the occurrence rate of pedestrian traffic accidents, improving the safety of urban pedestrian traffic systems and the like.
The invention discloses a TOPSIS method-based highway traffic accident black point road section identification method in China with publication number CN108447265A, and provides a TOPSIS method-based highway traffic accident black point road section identification method, which comprises the following steps: 1. equally dividing a road to be researched into n road section units according to the length; 2. collecting road traffic accident data of a plurality of years, and counting data such as total accident amount, total dead number of accidents, total serious injury number of accidents, total light injury number of accidents, total road sealing time caused by traffic accidents and the like of each road section unit; 3. calculating the equivalent casualty total number of the traffic accidents of each road section unit; 4. taking the indexes of the total traffic accident amount, the equivalent casualty total number of the traffic accident, the total lane sealing time caused by the traffic accident and the like of each road section unit as a road traffic safety evaluation index system, and calculating the traffic safety sequencing index of each road section by using a TOPSIS method; 5. and identifying the road section units with the traffic safety sequencing indexes smaller than the safety threshold value as accident black point road sections. The method comprehensively considers various traffic accident indexes, is simple to operate, has strong transportability and is easy to popularize and apply.
The support vector machine proposed by Vapnik et al is a method based on maximizing classification interval, which first extracts support vectors located on class boundaries, and then uses these support vectors to construct an optimal classification hyperplane, which can ensure that the probability of data points being misclassified is minimum. The known conditions and objectives of the support vector machine are as follows:
the known conditions are: given a training set of sample points
S＝{(x1,y1),(x2,y2),L,(xi,yi),L,(xl-1,yl-1)，(xl,yl)} (1)
Wherein x isi∈RnFor the input of a sample, i is 1,2, L, L, yiIs sample x { -1,1 { }iThe corresponding class label, l, is the data of the training sample. The goal of support vector machine training is to classify a given sample into two classes, so that the probability of misclassification is as small as possible. The classification principle of the support vector machine is shown in fig. 3.
As shown in FIG. 3, the triangles and circles represent the two types of samples A and B, l to be classified, respectively+,l-A boundary hyperplane representing two classes (corresponding to a straight line in the case of samples that are two-dimensional); l0To decide the hyperplane. l-,l0,l+The following equations are satisfied, respectively:
l-:wx+b＝-1 (2)
l0:wx+b＝0 (3)
l+:wx+b＝1 (4)
an objective function:
s.t.yi(wxi+b)≥1-ξii＝1,2,L,l(6)
ξi≥0i＝1,2,L,l (7)
wherein w is a normal vector of the decision hyperplane; b is constant, xi ═ xi1,L,ξl)TIs a relaxation coefficient vector, and C is a penalty factor. Hyperplane l at this time0Sample points a and B cannot be separated. For this purpose, it is first constructedLagrange function:
wherein α ═ (α)1,L,αl)Tβ＝(β1,L,βl)TIs the lagrange multiplier vector.
Respectively combining L with w and xiiAnd b, calculating the partial derivatives, including:
respectively, making the partial derivatives as 0, we can obtain:
βi＝C-αi(13)
substituting equations (9) to (11) into (8) can yield:
coupled formula (14), so that formula (15) can be continuously transformed into
The dual questions of the original questions (5) to (7) can then be obtained as:
βi＝C-αi,i＝1,2,L,l (19)
βi≥0,i＝1,2,L,l (20)
αi≥0,i＝1,2,L,l (21)
substituting equation (19) into equation (20) yields:
α≤C (22)
the inequalities (21) and (22) are combined to obtain:
0≤α≤C (23)
the optimization problems (17) to (21) can then be written as:
0≤α≤C,i＝1,2,L,l (26)
to convert the maximization problem into the minimization problem equivalent thereto, there are
0≤α≤C,i＝1,2,L,l (29)
Disclosure of Invention
The invention aims to provide a machine learning method for identifying black spots of urban traffic accidents by fusing multi-source characteristics.
In order to solve the problems, the technical scheme of the invention is as follows:
a multi-source feature fused urban traffic accident black spot identification machine learning method comprises the following steps:
s1, identifying urban traffic accident black spots based on a support vector machine, comprising the following steps:
s11, carrying out digitization and standardized preprocessing on traffic accident data, wherein the traffic accident data comprises site, time, accident, people, vehicles, roads and environment characteristic data;
s12, dividing the traffic accident data into training data and detection data;
s13, constructing SVM parameters according to the training data;
s14, solving interface parameters according to an SMO algorithm;
s15, inputting detection data, detecting the identification accuracy of the interface parameters obtained by the training data, judging whether the identification accuracy is met, if so, carrying out the next step, and if not, reconstructing an SVM function;
s16, inputting accident data needing black point identification into interface parameters, and judging whether the accident data are black points;
s17, outputting a judgment result, and finishing the black spot identification;
s2, identifying the black spots of the urban traffic accident based on the deep neural network, comprising the following steps of:
s21, establishing an initial deep neural network;
s22, inputting a training sample vector set, adding data category information in an initial linear analysis function stored locally, and generating a linear category analysis function, wherein the category information is black dots or non-black dots;
s23, obtaining an optimization function of the initial deep neural network according to a locally stored unsupervised coding model optimization function and the linear class analysis function;
s24, acquiring parameters of the initial deep neural network according to the optimization function of the initial deep neural network;
s25, establishing a deep neural network according to a locally stored classification neural network, an initial deep neural network and parameters of the initial deep neural network;
s26, receiving input data to be identified;
s27, inputting data to be identified into an input layer of a deep neural network, and acquiring category information to which the data to be identified output by an output layer of the deep neural network belongs;
further, the accident characteristic data includes simplicity, generality and severity.
Further, the person characteristic data includes gender, age, and driving age.
Further, the vehicle characteristic data includes small-sized vehicles, medium-sized vehicles, and large-sized vehicles.
Further, the road characteristic data comprise the number of lanes, the wet and slippery condition of the road surface and the flow rate, the number of lanes comprises one, two, three and four, and the wet and slippery condition of the road surface comprises dryness, wetness and accumulated water.
Further, the environmental characteristic data includes sunny, snow, rain, and cloudy.
Compared with the prior art, the invention has the following beneficial effects:
1. according to the structured correlation characteristics of the urban traffic accident big data, on the basis of forming a label by identifying the nuclear density black points, 10 items of characteristic data such as accidents, people, vehicles, roads, environments and the like are combined, and the training of a complex model and optimal learning is carried out on the accident black points in the researched area by using a support vector machine method based on the maximized classification interval, so that the accuracy of identifying the black points is improved.
2. Along with the rapid growth of traffic accident multi-source data, the precision of precision and recall rate of support vector machine black spot identification is lower and lower, and the identification efficiency is not high. Therefore, aiming at the rapid growth of the multi-source data of the traffic accident, a black spot identification algorithm based on a deep neural network is provided, and the deep neural network of related data category information is established. Precision, recall rate precision and efficiency of black spot identification are improved.
Drawings
The following describes embodiments of the present invention in further detail with reference to the accompanying drawings.
FIG. 1 is a schematic diagram of an algorithm flow based on support vector machine black spot identification;
FIG. 2 is a schematic diagram of a specific algorithm flow of dark spot identification based on deep neural network;
FIG. 3 is a diagram illustrating the situation of the inseparable classification problem of the prior art SVM processing line.
Detailed Description
In order to make the technical means, the original characteristics, the achieved purpose and the efficacy of the invention easy to understand, the invention is further described with reference to the specific drawings.
Example (b):
as shown in fig. 1-3, a machine learning method for identifying black spots of urban traffic accidents by fusing multi-source features includes the following steps:
s1, identifying urban traffic accident black spots based on a support vector machine, comprising the following steps:
s11, carrying out digitization and standardized preprocessing on traffic accident data, wherein the traffic accident data comprises site, time, accident, people, vehicles, roads and environment characteristic data;
further, the accident characteristic data includes simplicity, generality and severity.
Further, the person characteristic data includes gender, age, and driving age.
Further, the vehicle characteristic data includes small-sized vehicles, medium-sized vehicles, and large-sized vehicles.
Further, the road characteristic data comprise the number of lanes, the wet and slippery condition of the road surface and the flow rate, the number of lanes comprises one, two, three and four, and the wet and slippery condition of the road surface comprises dryness, wetness and accumulated water.
Further, the environmental characteristic data includes sunny, snow, rain, and cloudy.
S12, dividing the traffic accident data into training data and detection data;
s13, constructing SVM parameters according to the training data;
s14, solving interface parameters according to an SMO algorithm;
s15, inputting detection data, detecting the identification accuracy of the interface parameters obtained by the training data, judging whether the identification accuracy is met, if so, carrying out the next step, and if not, reconstructing an SVM function;
s16, inputting accident data needing black point identification into interface parameters, and judging whether the accident data are black points;
s17, outputting a judgment result, and finishing the black spot identification;
s2, with the gradual reduction of precision of the precision and recall ratio of the black spot identification of the support vector machine, the black spot identification of the urban traffic accident is carried out by adopting a deep neural network, and the method comprises the following steps:
s21, establishing an initial deep neural network;
s22, inputting a training sample vector set, adding data category information in an initial linear analysis function stored locally, and generating a linear category analysis function, wherein the category information is black dots or non-black dots;
s23, obtaining an optimization function of the initial deep neural network according to a locally stored unsupervised coding model optimization function and the linear class analysis function;
s24, acquiring parameters of the initial deep neural network according to the optimization function of the initial deep neural network;
s25, establishing a deep neural network according to a locally stored classification neural network, an initial deep neural network and parameters of the initial deep neural network;
s26, receiving input data to be identified;
s27, inputting data to be identified into an input layer of a deep neural network, and acquiring category information to which the data to be identified output by an output layer of the deep neural network belongs;
it will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, and that various changes and modifications may be made without departing from the spirit and scope of the invention as defined in the appended claims. The scope of the invention is defined by the appended claims and equivalents thereof.
Claims (6)
1. A multi-source feature fused urban traffic accident black spot identification machine learning method is characterized by comprising the following steps: the method comprises the following steps:
s1, identifying urban traffic accident black spots based on a support vector machine, comprising the following steps:
s11, carrying out digitization and standardized preprocessing on traffic accident data, wherein the traffic accident data comprises site, time, accident, people, vehicles, roads and environment characteristic data;
s12, dividing the traffic accident data into training data and detection data;
s13, constructing SVM parameters according to the training data;
s14, solving interface parameters according to an SMO algorithm;
s15, inputting detection data, detecting the identification accuracy of the interface parameters obtained by the training data, judging whether the identification accuracy is met, if so, carrying out the next step, and if not, reconstructing an SVM function;
s16, inputting accident data needing black point identification into interface parameters, and judging whether the accident data are black points;
s17, outputting a judgment result, and finishing the black spot identification;
s2, identifying the black spots of the urban traffic accident based on the deep neural network, comprising the following steps of:
s21, establishing an initial deep neural network;
s22, inputting a training sample vector set, adding data category information in an initial linear analysis function stored locally, and generating a linear category analysis function, wherein the category information is black dots or non-black dots;
s23, obtaining an optimization function of the initial deep neural network according to a locally stored unsupervised coding model optimization function and the linear class analysis function;
s24, acquiring parameters of the initial deep neural network according to the optimization function of the initial deep neural network;
s25, establishing a deep neural network according to a locally stored classification neural network, an initial deep neural network and parameters of the initial deep neural network;
s26, receiving input data to be identified;
and S27, inputting the data to be identified into an input layer of the deep neural network, and acquiring the class information of the data to be identified output by the output layer of the deep neural network.
2. The method for machine learning of urban traffic accident black spot identification fused with multi-source features according to claim 1, wherein the method comprises the following steps: the accident characteristic data includes simple, general and severe.
3. The method for machine learning of urban traffic accident black spot identification fused with multi-source features according to claim 1, wherein the method comprises the following steps: the person characteristic data includes gender, age, and driving age.
4. The method for machine learning of urban traffic accident black spot identification fused with multi-source features according to claim 1, wherein the method comprises the following steps: the vehicle characteristic data includes small-sized vehicles, medium-sized vehicles, and large-sized vehicles.
5. The method for machine learning of urban traffic accident black spot identification fused with multi-source features according to claim 1, wherein the method comprises the following steps: the road characteristic data comprises lane number, road surface slippery condition and flow, the lane number comprises one, two, three and four, and the road surface slippery condition comprises dryness, wetness and ponding.
6. The method for machine learning of urban traffic accident black spot identification fused with multi-source features according to claim 1, wherein the method comprises the following steps: the environmental characteristic data comprises sunny, snow, rain and cloudy.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910962247.4A CN110728050A (en) | 2019-10-11 | 2019-10-11 | Urban traffic accident black spot identification machine learning method fusing multi-source features |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910962247.4A CN110728050A (en) | 2019-10-11 | 2019-10-11 | Urban traffic accident black spot identification machine learning method fusing multi-source features |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110728050A true CN110728050A (en) | 2020-01-24 |
Family
ID=69220920
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910962247.4A Pending CN110728050A (en) | 2019-10-11 | 2019-10-11 | Urban traffic accident black spot identification machine learning method fusing multi-source features |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110728050A (en) |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102737508A (en) * | 2012-06-19 | 2012-10-17 | 银江股份有限公司 | Urban road traffic state detection method combined with support vector machine (SVM) and back propagation (BP) neural network |
-
2019
- 2019-10-11 CN CN201910962247.4A patent/CN110728050A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102737508A (en) * | 2012-06-19 | 2012-10-17 | 银江股份有限公司 | Urban road traffic state detection method combined with support vector machine (SVM) and back propagation (BP) neural network |
Non-Patent Citations (1)
| Title |
|---|
| 李巧茹 等: "基于GA-BP神经网络算法和粗糙集理论的交通事故黑点模型" * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| He et al. | Obstacle detection of rail transit based on deep learning | |
| CN102708356B (en) | Automatic license plate positioning and recognition method based on complex background | |
| WO2017156772A1 (en) | Method of computing passenger crowdedness and system applying same | |
| CN111460919B (en) | Monocular vision road target detection and distance estimation method based on improved YOLOv3 | |
| CN109544914B (en) | Shared bicycle reverse behavior identification method based on historical GPS track | |
| CN114359876B (en) | Vehicle target identification method and storage medium | |
| CN104978567A (en) | Vehicle detection method based on scenario classification | |
| Pawar et al. | Classification of gaps at uncontrolled intersections and midblock crossings using support vector machines | |
| Zhang et al. | A proactive crash risk prediction framework for lane-changing behavior incorporating individual driving styles | |
| Sarmiento | Pavement distress detection and segmentation using YOLOv4 and DeepLabv3 on pavements in the Philippines | |
| CN102708384A (en) | Bootstrapping weak learning method based on random fern and classifier thereof | |
| Tran et al. | Hyperparameter optimization for improving recognition efficiency of an adaptive learning system | |
| CN117392853A (en) | Big data intelligent lane control system based on high in clouds | |
| CN106529391A (en) | Robust speed-limit traffic sign detection and recognition method | |
| CN113468994A (en) | Three-dimensional target detection method based on weighted sampling and multi-resolution feature extraction | |
| CN104966064A (en) | Pedestrian ahead distance measurement method based on visual sense | |
| Xie et al. | Weakly supervised object localization with soft guidance and channel erasing for auto labelling in autonomous driving systems | |
| Valiente et al. | Robust perception and visual understanding of traffic signs in the wild | |
| CN110728050A (en) | Urban traffic accident black spot identification machine learning method fusing multi-source features | |
| Yu et al. | Research on detection and recognition algorithm of road traffic signs | |
| Satti et al. | R‐ICTS: Recognize the Indian cautionary traffic signs in real‐time using an optimized adaptive boosting cascade classifier and a convolutional neural network | |
| Abid et al. | Vehicle detection for intelligent traffic surveillance system | |
| CN114022705B (en) | Self-adaptive target detection method based on scene complexity pre-classification | |
| CN114429621A (en) | UFSA algorithm-based improved lane line intelligent detection method | |
| Pan et al. | Fake license plate recognition in surveillance videos |
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 | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: 215000 surveying and mapping geographic information building, No. 101, Suhong Middle Road, Suzhou Industrial Park, Suzhou City, Jiangsu Province Applicant after: Yuance Information Technology Co.,Ltd. Address before: 215000 surveying and mapping geographic information building, No. 101, Suhong Middle Road, Suzhou Industrial Park, Suzhou City, Jiangsu Province Applicant before: SUZHOU INDUSTRIAL PARK SURVEYING MAPPING AND GEOINFORMATION Co.,Ltd. |