CN112669608A - License plate recognition data re-matching method based on maximum travel time probability - Google Patents
License plate recognition data re-matching method based on maximum travel time probability Download PDFInfo
- Publication number
- CN112669608A CN112669608A CN202011558261.7A CN202011558261A CN112669608A CN 112669608 A CN112669608 A CN 112669608A CN 202011558261 A CN202011558261 A CN 202011558261A CN 112669608 A CN112669608 A CN 112669608A
- Authority
- CN
- China
- Prior art keywords
- travel time
- probability
- upstream
- time
- downstream
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 29
- 238000009826 distribution Methods 0.000 claims abstract description 26
- 238000005096 rolling process Methods 0.000 claims abstract description 14
- 238000007781 pre-processing Methods 0.000 claims abstract description 5
- 238000011144 upstream manufacturing Methods 0.000 claims description 43
- 230000001186 cumulative effect Effects 0.000 claims description 25
- 238000005315 distribution function Methods 0.000 claims description 12
- 239000011159 matrix material Substances 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000009499 grossing Methods 0.000 claims description 3
- 238000007689 inspection Methods 0.000 claims description 3
- 238000010586 diagram Methods 0.000 description 5
- 238000012015 optical character recognition Methods 0.000 description 2
- 230000009897 systematic effect Effects 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003672 processing method Methods 0.000 description 1
Images
Landscapes
- Traffic Control Systems (AREA)
Abstract
The invention discloses a license plate recognition data re-matching method based on the maximum travel time probability. The method provided by the invention comprises the following steps: and extracting the travel time of the road section, preprocessing the travel time, constructing travel time distribution, calculating the probability of the travel time and the confidence interval of the travel time, and performing a re-matching algorithm in a rolling mode. The method can automatically calculate the probability of travel time under different matching conditions aiming at the problem of wrong or unidentified license plate recognition, and execute matching according to the maximum occurrence probability of the travel time.
Description
Technical Field
The invention relates to a license plate recognition data re-matching method based on the maximum travel time probability, which is used for the re-matching of the wrong recognition and the unrecognized data of license plates at the upstream and the downstream of an urban road section and belongs to the field of intelligent transportation.
Background
With continuous investment of various major cities in smart traffic and smart cities, infrastructures represented by cloud computing and camera sensors are widely constructed and applied, for example, electronic police and checkpoint systems are widely installed and used on urban roads, License Plate identification data (LPR) obtained based on the video detectors is well applied to the aspects of urban traffic demand (OD) estimation, intersection queuing length estimation, road travel time estimation and the like, wide coverage and high accuracy of traffic information acquisition are effectively improved, and good economic benefits are obtained in the aspect of urban intelligent traffic management.
The license plate identification data is essentially obtained from Optical Character Recognition (OCR) video image information, and due to factors such as illumination conditions, angles and occlusion, systematic errors such as identification errors and non-identification errors generally exist. These systematic errors present challenges to the accurate collection of traffic information.
Although the problem of wrong license plate recognition occurs, the electronic police system can still accurately capture the information such as the time stamp, the position and the like when the vehicle passes through the stop line. According to the characteristics of the license plate identification data, the invention designs a license plate identification data re-matching method based on the maximum travel time probability, and provides scientific basis for fine management of urban traffic.
Disclosure of Invention
The invention aims to provide a license plate recognition data re-matching method based on the maximum travel time probability. The method can automatically calculate the probability of travel time under different matching conditions aiming at the problem of wrong or unidentified license plate recognition, and execute matching according to the maximum occurrence probability of the travel time. In order to achieve the above object, the method provided by the present invention comprises: and extracting the travel time of the road section, preprocessing the travel time, constructing travel time distribution, calculating the probability of the travel time and the confidence interval of the travel time, and performing a re-matching algorithm in a rolling mode.
The basic steps of the invention are as follows:
s1: extracting road section travel time and preprocessing, including matching of upstream and downstream vehicle license plates of a target road section, extracting travel time and processing travel time outliers;
s2: constructing travel time distribution, and estimating the distribution of the travel time of the road section by adopting a nonparametric estimation method according to the acquired road section travel time data;
s3: calculating a travel time probability and a travel time confidence interval, and calculating the travel time probability and the travel time confidence interval according to estimated travel time Cumulative Distribution (CDF) and Inverse Cumulative Distribution (ICDF);
s4: and (4) executing a re-matching algorithm in a rolling mode, and executing the re-matching algorithm by taking the maximum travel time occurrence probability as a target.
The process of step S1 includes:
s11: and performing matching on the upstream and downstream license plate matching data by taking the identified license plate as information.
S12: and extracting travel time data of the target road section, and rolling and processing travel time outliers.
The process of step S2 includes:
the distribution of travel times is constructed using kernel density estimation. Let n travel-time sample point data { tt1,tt2,...,ttnThe independent co-distributions F are obeyed with the probability density function F. The probability density function of f is estimated as follows:
wherein K (-) is a kernel function-a non-negative function, and h>0 is a smoothing parameter called bandwidth; tt is a road section travel time variable; kernel function K with subscripthCalled the scaling kernel and defined as Kh(tt) ═ 1/h · K (tt/h); using a Gaussian kernel as a kernel function, i.e.
Wherein
Is a standard normal densityA function.
The method of bandwidth selection for kernel density estimation is as follows:
wherein,
and n is the travel time standard deviation, and the number of travel time samples.
The travel time cumulative distribution function f (tt) is calculated as follows:
the process of step S3 includes:
s31: calculating the travel time probability;
travel time ttijCorresponding probability pij(ttij) The calculation is as follows:
pij(ttij)＝F(ttij)-F(ttij-1) (5)
f (-) is the cumulative distribution function of the corresponding travel time.
S32: calculating a travel time confidence interval;
for the travel time cumulative distribution function F, a probability p (p is more than or equal to 0 and less than or equal to 1) and an Inverse Cumulative Distribution Function (ICDF) F are set-1The return journey time threshold tt is such that:
F-1(p)＝inf{tt∈R:F(tt)≥p} (6)
at a Confidence level CL β ═ 1- α, the Confidence Interval (CI) of the journey time is CI ═ tt1,tt2]Then, the confidence interval calculation method of the travel time is as follows:
p(tt1≤tt≤tt2)＝β (7)
wherein, tt1Lower bound of the travel time confidence interval, tt2The upper bound of the travel time confidence interval. F-1For the inverse cumulative distribution, α is the significance level of the travel time confidence interval.
S33: rolling and searching the unmatched license plate data of the upstream and the downstream;
the rolling time window method finds the set of unmatched vehicles downstream, where the time window h is selectedwThen in the qth time window, find n unmatched vehicle sets downstream
Corresponding to a timestamp vector of
Wherein,
denoted as the nth vehicle in the downstream qth group,
time stamp denoted as n-th vehicle in the downstream q-th group passing through the stop line, dL(k)Indicating that the vehicle is in the k-th lane downstream.
Finding m unmatched vehicle sets of the q group within a reasonable time range in the upstream
Corresponding to a timestamp vector of
Wherein,
denoted as the mth vehicle in the upstream qth group,
indicated as the timestamp of the mth vehicle in the upstream qth group passing the stop line.
Since, at the β confidence level, the travel time confidence interval is:
the earliest and latest points in time upstream are therefore as follows:
wherein,
is the timestamp of the j-th vehicle collected upstream,
the timestamp of the downstream ith vehicle being collected, the vehicle is in the kth lane.
The time ranges corresponding to the confidence level β upstream are as follows:
the process of step S4 includes:
s41: performing matching with a maximum probability of occurrence of travel time;
to maximize the product of the matched time-of-flight probabilities, the objective function is as follows:
logarithm of the objective function, then
Let logpij＝aijAnd logz ═ z', the objective function is converted as follows:
the matching matrix is X as follows, if XijAnd 1, the ith downstream license plate number is matched with the jth upstream license plate number.
When m is>n>When 0, namely the number of the downstream license plates is more than that of the upstream license plates, a mx (m-n) virtual license plate array is newly added, the travel time probability corresponding to the virtual license plates is larger than a small positive number of zero, and the logarithm value is a negative large value
New m × m travel time probability log value matrix A'm×mAs follows.
When n is greater than m and greater than 0, namely the number of the downstream license plates is less than that of the upstream license plates, adding (n-m) multiplied by n lines of virtual license plates, wherein a new n multiplied by n travel time probability logarithm value matrix is as follows.
S42: checking a matching result;
performing inspection, and if the part matched with the virtual license plate in the upstream and the downstream is reset to be unmatched data; if the re-matched travel time is not within the confidence interval, the portion is reset to unmatched data.
The invention has the following beneficial effects:
(1) and the vehicle license plate recognition data is subjected to re-matching according to the maximum travel time probability, the physical meaning is clear, the calculation result is reliable, and the matching result with high accuracy is obtained.
(2) The method provides high-precision traffic basic information for urban traffic demand (OD) estimation, intersection motor vehicle queuing estimation and road section dynamic stock estimation, and provides support for urban fine management.
Drawings
FIG. 1 is an overall flow chart of the inventive scheme.
FIG. 2 is a schematic diagram of a road network and an electronic police device.
FIG. 3 is a schematic diagram of a cumulative time-of-flight distribution and an inverse cumulative distribution.
FIG. 4 is a schematic of the travel time confidence interval.
Detailed Description
The invention is further described below with reference to fig. 1.
The basic steps of the invention are as follows:
s1: and extracting the travel time of the road section and preprocessing, including the matching of the upstream and downstream vehicle license plates of the target road section, the extraction of the travel time and the processing of the travel time outlier.
S2: and constructing travel time distribution, and estimating the distribution of the travel time of the road section by adopting a non-parameter estimation urban road travel time noise data processing method according to the acquired road section travel time data.
S3: and calculating the travel time probability and the travel time confidence interval, and calculating the travel time probability and the travel time confidence interval according to the estimated travel time Cumulative Distribution (CDF) and Inverse Cumulative Distribution (ICDF).
S4: and (4) executing a re-matching algorithm in a rolling mode, and executing the re-matching algorithm by taking the maximum travel time occurrence probability as a target.
FIG. 2 is a schematic diagram of a road network and electronic police equipment layout, and data collected by the license plate recognition video monitors of the upstream and downstream are matched according to the upstream and downstream relations of a target road section.
The process of step S1 includes:
s11: and performing matching on the upstream and downstream license plate matching data by taking the identified license plate as information.
S12: and extracting travel time data of the target road section, and rolling and processing travel time outliers.
Fig. 3 is a schematic diagram of the travel time cumulative distribution and the inverse cumulative distribution, which is a basis for constructing the travel time distribution and calculating the travel time probability, and is used in steps S2 and S3 of the scheme.
The process of step S2 includes:
the distribution of travel times is constructed using kernel density estimation. Let n travel-time sample point data { tt1,tt2,…,ttnThe independent co-distributions F are obeyed with the probability density function F. The probability density function of f is estimated as follows:
wherein K (-) is a kernel function-a non-negative function, and h>0 is a smoothing parameter called bandwidth; tt is a road section travel time variable; kernel function K with subscripthCalled the scaling kernel and defined as Kh(tt) ═ 1/h · K (tt/h); using a Gaussian kernel as a kernel function, i.e.
Wherein
Is a standard normal density function.
The method of bandwidth selection for kernel density estimation is as follows:
wherein,
and n is the travel time standard deviation, and the number of travel time samples.
The travel time cumulative distribution function f (tt) is calculated as follows:
the process of step S3 includes:
s31: calculating the travel time probability;
travel time ttijCorresponding probability pij(ttij) The calculation is as follows:
pij(ttij)＝F(ttij)-F(ttij-1) (5)
f (-) is the cumulative distribution function of the corresponding travel time.
S32: calculating a travel time confidence interval;
for the travel time cumulative distribution function F, a probability p (p is more than or equal to 0 and less than or equal to 1) and an Inverse Cumulative Distribution Function (ICDF) F are set-1The return journey time threshold tt is such that:
F-1(p)＝inf{tt∈R:F(tt)≥p} (6)
at a Confidence level CL β ═ 1- α, the Confidence Interval (CI) of the journey time is CI ═ tt1,tt2]Then, the confidence interval calculation method of the travel time is as follows:
p(tt1≤tt≤tt2)＝β (7)
wherein, tt1Lower bound of the travel time confidence interval, tt2The upper bound of the travel time confidence interval. F-1For the inverse cumulative distribution, α is the significance level of the travel time confidence interval.
S33: rolling and searching the unmatched license plate data of the upstream and the downstream;
fig. 4 is a schematic diagram of travel time confidence intervals, which is used for calculating travel time confidence intervals at corresponding confidence levels and searching for unmatched license plate data in upstream and downstream.
The rolling time window method finds the set of unmatched vehicles downstream, where the time window h is selectedwThen in the qth time window, find n unmatched vehicle sets downstream
Corresponding to a timestamp vector of
Wherein,
denoted as the nth vehicle in the downstream qth group,
time stamp denoted as n-th vehicle in the downstream q-th group passing through the stop line, dL(k)Indicating that the vehicle is in the k-th lane downstream.
Finding m unmatched vehicle sets of the q group within a reasonable time range in the upstream
Corresponding to a timestamp vector of
Wherein,
denoted as the mth vehicle in the upstream qth group,
indicated as the timestamp of the mth vehicle in the upstream qth group passing the stop line.
Since, at the β confidence level, the travel time confidence interval is:
the earliest and latest points in time upstream are therefore as follows:
wherein,
is the timestamp of the j-th vehicle collected upstream,
the timestamp of the downstream ith vehicle being collected, the vehicle is in the kth lane.
The time ranges corresponding to the confidence level β upstream are as follows:
the process of step S4 includes:
s41: performing matching with a maximum probability of occurrence of travel time;
to maximize the product of the matched time-of-flight probabilities, the objective function is as follows:
logarithm of the objective function, then
Let logpij＝aijAnd logz ═ z', the objective function is converted as follows:
the matching matrix is X as follows, if XijAnd 1, the ith downstream license plate number is matched with the jth upstream license plate number.
When m is>n>When 0, namely the number of the downstream license plates is more than that of the upstream license plates, a mx (m-n) virtual license plate array is newly added, the travel time probability corresponding to the virtual license plates is larger than a small positive number of zero, and the logarithm value is a negative large value
New m × m travel time probability log value matrix A'm×mAs follows.
When n is greater than m and greater than 0, namely the number of the downstream license plates is less than that of the upstream license plates, adding (n-m) multiplied by n lines of virtual license plates, wherein a new n multiplied by n travel time probability logarithm value matrix is as follows.
S42: checking a matching result;
performing inspection, and if the part matched with the virtual license plate in the upstream and the downstream is reset to be unmatched data; if the re-matched travel time is not within the confidence interval, the portion is reset to unmatched data.
The formulas and descriptions are presented only as a general description to facilitate one of ordinary skill in the art in understanding and practicing the present invention. It is obvious to those skilled in the art that the formula can be easily modified without inventive work, and thus all modifications and improvements made within the scope of the invention should be within the scope of the invention.
Claims (1)
1. The license plate recognition data re-matching method based on the maximum travel time probability is characterized by comprising the following steps of:
s1: extracting road section travel time and preprocessing, including matching of upstream and downstream vehicle license plates of a target road section, extracting travel time and processing travel time outliers;
s11: matching the upstream and downstream license plate matching data by taking the recognized license plate as information;
s12: extracting travel time data of a target road section, and rolling and processing travel time outliers;
s2: constructing travel time distribution, and estimating the distribution of the travel time of the road section by adopting a non-parameter estimation method according to the acquired road section travel time data, wherein the method specifically comprises the following steps:
let n travel-time sample point data { tt1,tt2,...,ttnObey an independent homography F, and the probability density function is F; the probability density function of f is estimated as follows:
where K (-) is a kernel function, h>0 is a smoothing parameter called bandwidth; tt is a road segment travel time variable; kernel function KhTo scale the kernel and define Kh(tt) ═ 1/h · K (tt/h); using a Gaussian kernel as a kernel function, i.e.
Wherein
Is a standard normal density function;
the bandwidth of the kernel density estimate is chosen as follows:
wherein,
is the standard deviation of travel time, and n is the number of travel time samples;
the travel time cumulative distribution function f (tt) is calculated as follows:
s3: calculating a travel time probability and a travel time confidence interval, and calculating the travel time probability and the travel time confidence interval according to the estimated travel time cumulative distribution and the inverse cumulative distribution, wherein the method specifically comprises the following steps:
s31: calculating the travel time probability;
travel time ttijCorresponding probability pij(ttij) The calculation is as follows:
pij(ttij)＝F(ttij)-F(ttij-1) (5)
f (-) is the cumulative distribution function of the corresponding travel time;
s32: calculating a travel time confidence interval;
for the travel time cumulative distribution function F, a probability p (p is more than or equal to 0 and less than or equal to 1) is set, and the inverse cumulative distribution function F-1The return journey time threshold tt is such that:
F-1(p)＝inf{tt∈R:F(tt)≥p} (6)
at a confidence level CL, β ═ 1- α, the confidence interval for the journey time is CI ═ tt1,tt2]Then the confidence interval for the travel time is calculated as follows:
p(tt1≤tt≤tt2)＝β (7)
wherein, tt1Lower bound of the travel time confidence interval, tt2An upper bound for a travel time confidence interval; f-1For inverse cumulative distributions, α is the significance level of the travel time confidence interval;
s33: rolling and searching the unmatched license plate data of the upstream and the downstream;
the rolling time window method finds the set of unmatched vehicles downstream, where the time window h is selectedwThen in the qth time window, find n unmatched vehicle sets downstream
Corresponding to a timestamp vector of
Wherein,
denoted as the nth vehicle in the downstream qth group,
time stamp denoted as n-th vehicle in the downstream q-th group passing through the stop line, dL(k)Indicating that the vehicle is in the k-th lane downstream;
finding m unmatched vehicle sets in the q group within a set time range in the upstream
Corresponding to a timestamp vector of
Wherein,
denoted as the mth vehicle in the upstream qth group,
a time stamp expressed as the mth vehicle in the upstream qth group passing the stop line;
since, at the β confidence level, the travel time confidence interval is:
the earliest and latest points in time upstream are therefore as follows:
wherein,
is the timestamp of the j-th vehicle collected upstream,
a timestamp acquired by the ith vehicle at the downstream, wherein the vehicle is in the kth lane;
the time ranges corresponding to the confidence level β upstream are as follows:
s4: and (3) executing a re-matching algorithm in a rolling way, and executing the re-matching algorithm by taking the maximum probability of occurrence of the travel time as a target, wherein the specific steps are as follows:
s41: performing matching with a maximum probability of occurrence of travel time;
to maximize the product of the matched time-of-flight probabilities, the objective function is as follows:
logarithm of the objective function, then
Let logpij＝aijAnd logz ═ z', the objective function is converted as follows:
the matching matrix is X as follows, if XijIf the number is 1, the ith downstream license plate number is matched with the jth upstream license plate number;
when m is>n>When 0, namely the number of the downstream license plates is more than that of the upstream license plates, a mx (m-n) virtual license plate array is newly added, the travel time probability corresponding to the virtual license plates is larger than a small positive number of zero, and the logarithm value is a negative large value
New m × m travel time probability log value matrix A'm×mThe following;
when n is greater than m and greater than 0, namely the number of the downstream license plates is less than that of the upstream license plates, adding (n-m) multiplied by n lines of virtual license plates, wherein a new n multiplied by n travel time probability logarithm value matrix is as follows;
s42: checking a matching result;
performing inspection, and if the part matched with the virtual license plate in the upstream and the downstream is reset to be unmatched data; if the re-matched travel time is not within the confidence interval, the portion is reset to unmatched data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011558261.7A CN112669608B (en) | 2020-12-25 | 2020-12-25 | License plate recognition data re-matching method based on maximum travel time probability |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011558261.7A CN112669608B (en) | 2020-12-25 | 2020-12-25 | License plate recognition data re-matching method based on maximum travel time probability |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112669608A true CN112669608A (en) | 2021-04-16 |
| CN112669608B CN112669608B (en) | 2022-04-22 |
Family
ID=75408841
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011558261.7A Active CN112669608B (en) | 2020-12-25 | 2020-12-25 | License plate recognition data re-matching method based on maximum travel time probability |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112669608B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101430828A (en) * | 2008-12-22 | 2009-05-13 | 交通部公路科学研究院 | License plate data processing method based on traffic information extraction computation |
| JP2014153727A (en) * | 2013-02-04 | 2014-08-25 | Mitsubishi Heavy Ind Ltd | Vehicle power consumption simulation device, vehicle power consumption simulation method and program |
| CN110164132A (en) * | 2019-05-29 | 2019-08-23 | 浙江警察学院 | A kind of detection method and system of road traffic exception |
| CN111968369A (en) * | 2020-08-14 | 2020-11-20 | 山东师范大学 | Traffic route guidance method and system |
-
2020
- 2020-12-25 CN CN202011558261.7A patent/CN112669608B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101430828A (en) * | 2008-12-22 | 2009-05-13 | 交通部公路科学研究院 | License plate data processing method based on traffic information extraction computation |
| JP2014153727A (en) * | 2013-02-04 | 2014-08-25 | Mitsubishi Heavy Ind Ltd | Vehicle power consumption simulation device, vehicle power consumption simulation method and program |
| CN110164132A (en) * | 2019-05-29 | 2019-08-23 | 浙江警察学院 | A kind of detection method and system of road traffic exception |
| CN111968369A (en) * | 2020-08-14 | 2020-11-20 | 山东师范大学 | Traffic route guidance method and system |
Non-Patent Citations (7)
| Title |
|---|
| 付凤杰: "基于自动车牌识别数据的城市道路行程时间估计", 《中国优秀博硕士学位论文全文数据库(博士)》 * |
| 何胜学: "自动代客泊车背景下的共享停车匹配模型及对应禁忌搜索算法", 《计算机应用研究》 * |
| 商明菊 等: "基于Ｒ－FPOP 变点检测的城市路段", 《贵州大学学报( 自然科学版)》 * |
| 李雪艳: "基于数字图像处理的车牌识别系统的研究", 《北京工业职业技术学院学报》 * |
| 江勇东等: "探测车占有率对宏观基本图估计精度的影响", 《交通科学与工程》 * |
| 潘义勇 等: "随机交通网络最小条件风险路径问题", 《重庆交通大学学报(自然科学版)》 * |
| 龚越 等: "基于梯度提升回归树的城市道路行程时间预测", 《浙江大学学报(工学版)》 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112669608B (en) | 2022-04-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110796168B (en) | Vehicle detection method based on improved YOLOv3 | |
| CN108986465B (en) | Method, system and terminal equipment for detecting traffic flow | |
| CN110751828B (en) | Road congestion measuring method and device, computer equipment and storage medium | |
| CN108389421B (en) | Parking lot accurate induction system and method based on image re-identification | |
| CN109887283B (en) | Road congestion prediction method, system and device based on checkpoint data | |
| CN110083802B (en) | Highway toll station vehicle flow big data prediction method based on multi-target regression | |
| CN110798805B (en) | Data processing method and device based on GPS track and storage medium | |
| CN114879695A (en) | Track matching method, device, equipment and medium | |
| CN114067095B (en) | Water level identification method based on water gauge character detection and identification | |
| CN113763707B (en) | Road section overflow and lane blocking state identification method based on number plate data | |
| CN113743260B (en) | Pedestrian tracking method under condition of dense pedestrian flow of subway platform | |
| CN104778699B (en) | A kind of tracking of self adaptation characteristics of objects | |
| CN115657002A (en) | Vehicle motion state estimation method based on traffic millimeter wave radar | |
| CN115620518A (en) | Intersection traffic conflict discrimination method based on deep learning | |
| CN111724593A (en) | Bus travel time prediction method considering traffic flow transfer rate based on RFID | |
| CN112669608B (en) | License plate recognition data re-matching method based on maximum travel time probability | |
| CN111259833A (en) | Vehicle counting method based on traffic images | |
| CN116704750B (en) | Traffic state identification method based on clustering algorithm, electronic equipment and medium | |
| CN108053646A (en) | Traffic characteristic acquisition methods, Forecasting Methodology and system based on time-sensitive feature | |
| CN111178573A (en) | Flood dynamic prediction method based on grey correlation analysis method | |
| CN115909717A (en) | Expressway short-term traffic flow prediction method based on deep learning | |
| CN115796590A (en) | Method and device for predicting road fog accident risk, electronic equipment and medium | |
| CN115830555A (en) | Target identification method based on radar point cloud, storage medium and equipment | |
| CN115330841A (en) | Method, apparatus, device and medium for detecting projectile based on radar map | |
| CN113470376B (en) | Real-time regional in-transit vehicle counting method and system based on bayonet vehicle passing data |
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 |