Connect public, paid and private patent data with Google Patents Public Datasets

Method for extracting vehicle queue state information in urban road network based on high-resolution remote-sensing image

Download PDF

Info

Publication number
CN104021670A
CN104021670A CN 201410118280 CN201410118280A CN104021670A CN 104021670 A CN104021670 A CN 104021670A CN 201410118280 CN201410118280 CN 201410118280 CN 201410118280 A CN201410118280 A CN 201410118280A CN 104021670 A CN104021670 A CN 104021670A
Authority
CN
Grant status
Application
Patent type
Prior art keywords
road
vehicle
searching
queue
information
Prior art date
Application number
CN 201410118280
Other languages
Chinese (zh)
Other versions
CN104021670B (en )
Inventor
谷正气
张勇
李程
李健
Original Assignee
湖南工业大学
Priority date (The priority date 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 date listed.)
Filing date
Publication date

Links

Abstract

Vehicle queue state information is one of basic parameters for urban road traffic analysis. The invention provides a method for extracting vehicle queue state information in an urban road network based on a high-resolution remote-sensing image. The extraction method in the invention comprises the following steps: 1) obtaining planar road and planar vehicle information data in the high-resolution remote-sensing image; 2) extracting center line and edge lines of a planar road and carrying out cutting according to a certain length threshold to form road searching blocks; 3) calculating road vehicle occupation ratio of each road searching block; 4) taking a certain vehicle occupation ratio as a threshold value under the condition of a congestion queue, and screening out the road searching blocks, of which the road vehicle occupation ratio is larger or equal to the threshold value, at this moment, the total length of the plurality of road searching blocks being the length of the congestion vehicle queue in this road section. Specific position is determined by matching the central position of the road searching blocks with an electronic map or combining GPS data.

Description

一种高分遥感影像提取城市路网车辆队列状态信息方法 One kind of remote sensing images to extract high urban road network of the vehicle queue status information method

技术领域 FIELD

[0001] 本发明属于宏观交通状态判别技术领域,具体涉及交通遥感应用与智能交通技术。 [0001] The present invention belongs to the technical field macroscopic traffic state discrimination, and particularly sensing applications involving transportation intelligent transportation technology.

背景技术 Background technique

[0002] 近年来,我国城市化进程加快,交通拥堵问题日益严重。 [0002] In recent years, China has accelerated the process of urbanization, the growing problem of traffic congestion. 如何全面、快速掌握整个城市路网交通拥堵和交通瓶颈分布节点、了解交通拥堵状态、乃至于预测拥堵扩散范围,为交通管理部门进行交通疏导和路网科学建设提供科学性数据依据。 How comprehensive, quickly grasp the entire city road network traffic congestion and traffic bottlenecks in the distribution node, to understand the state of traffic congestion, jams and even forecast the diffusion range, perform traffic control and road network construction as the scientific traffic management departments to provide scientific data basis. 车辆队列(信息包括位置和长度)是城市拥堵的重要表现形式,是进行城市道路交通分析的基本参数之一。 Vehicle queue (information including the location and length) is an important form of urban congestion, is one of the basic parameters of urban road traffic analysis. 然而,当前地面交通信息数据获取手段(视频、线圈、微波、雷达)多为点源微观交通信息,数据获取中间传输环节过多,存储物理空间大,数据保存和处理困难,宏观可视化程度低,数据覆盖面有限,设备运营维护成本昂贵,无法满足城市乃至城市群路网大范围交通状态分析和宏观规划的智能交通数据需求。 However, the current ground traffic information data acquisition means (video, coils, microwave, radar) multi-point source microscopic traffic information data acquisition intermediate transmission links too large store physical space, difficulties in data storage and processing, a low macroscopic visualization of, data coverage is limited, expensive equipment operation and maintenance costs, can not meet the city and even large-scale urban agglomeration road network traffic analysis and macro planning of intelligent data traffic demand. 随着卫星技术和遥感影像成像质量的不断提升,2012年国家高分卫星“资源3号”上天,高分辨率遥感影像已进入“平民化时代”。 With the continuous improvement of satellite technology and remote sensing image quality, the 2012 national high satellite "Resource No. 3" God, high-resolution remote sensing image has entered the "civilian era." 将高分遥感数据应用于智能交通领域进行交通拥堵下的道路车辆队列判别成为一个重要而热门的研究课题。 The score of remote sensing data used in intelligent transportation field for road vehicles in traffic congestion queue discrimination has become an important and hot research topic.

[0003] 专利对比文献I (201010220935.2),公布了一种十字路口摄像头视频图像进行车俩长度的检测,该发明只能获取一个路口有限长度车辆队列情况,无法从宏观上获得整个路网的车辆队列信息。 [0003] Patent Document Comparison I (201010220935.2), discloses a video camera crossroads detecting both vehicle length, the invention can only get a limited length of the vehicle train crossing, the vehicle can not be obtained from the entire network macroscopically queue information.

[0004] 专利对比文献2(200910092506.9),公布了一种浮动车数据提取车辆排队长度,但由于浮动车数量的限制,也无法全面准确获取整个城市路网中的车辆队列信息。 [0004] Patent Document 2 Comparative (200910092506.9), discloses a floating car data extracted vehicle queue length, but since the number of floating cars limits can not be fully and accurately obtain the entire urban road network the vehicle queue information.

[0005] 专利对比文献3 (201210044347.7),公布了一种环形线圈检测器的数据进行车辆队列的提取,属于点源监测,其检测器的安装与维护都对道路有一定损毁,且受成本因素限制,无法获取整个城市路网中车辆队列信息的获取。 [0005] Patent Document 3 Comparative (201210044347.7), published an endless loop detector data is extracted in the vehicle train, belonging to a point source monitoring, installation and maintenance of the detector which have a certain damage to the road, and by cost factors restricted and can not get get the entire road network in the vehicle queue information.

[0006] 以上方法均属于地面交通信息获取手段,存在不足,且与本发明方法不属于同一信息来源。 [0006] The above method belong to ground transportation means access to information, shortcomings, and the method of the invention does not belong to the same sources of information.

[0007]科技论文对比文献 I (Jens Leitloff, Stefan Hinz&UweStilla, etal.Detectionof Vehicle Queues in QuickBird Imagery of City [J] Areas PhotogrammetrieFernerkundungGeoinformation, 4/2006, 315-325),提出了一种通过构建高分遥感影像中车辆队列特征知识库,应用形态学、灰度等信息,直接从遥感影像中提取车辆队列信息的方法,也不同于本发明所述的方法。 [0007] Comparative scientific papers document I (Jens Leitloff, Stefan Hinz & UweStilla, etal.Detectionof Vehicle Queues in QuickBird Imagery of City [J] Areas PhotogrammetrieFernerkundungGeoinformation, 4/2006, 315-325), proposed a high remote sensing images by constructing wherein the vehicle queue repository, morphological, grayscale information, queue information extraction method of the vehicle directly from the remote sensing image, also different from the method of the present invention.

发明内容 SUMMARY

[0008] 本发明目的在于克服现有技术中车辆队列判别方法的不足,提出了基于高分辨率遥感影像的滑块阈值搜索法,提取车辆队列信息。 [0008] The object of the present invention is to overcome the shortcomings of the prior determination method of the vehicle train art, is proposed based on the queue information Resolution Image slider threshold search method to extract vehicle.

[0009] 本发明方法包括以下步骤: [0009] The method of the present invention comprises the steps of:

(I)、应用图像处理相关算法,基于遥感影像提取面状道路和面状车辆矢量信息;(2)、提取面状道路中心线和边线,形成搜索块路; (The I), related to the application of image processing algorithms, remote sensing image extracting vector information on the basis of road surface shape and area of ​​the vehicle; (2), and extracted road centerline planar edges, path search block is formed;

a、获取面状道路,找到面状道路的首尾点:如附图1所示,首先获得构成面状道路边界的点集P = (P1, P2,……Pm, Pm+1,……,PnrtJ,共包含m+n (η≥m≥2)个点,将点集按顺时针方向两两组合成向量P = (P1, P2,……Pm,Pm+1,……,?m+n},共包含m+n(n≥m≥2)个 a, obtaining road planar, end to end to find the point of the road surface shape: as shown in Figure 1, is first obtained set of points form a planar road boundary P = (P1, P2, ...... Pm, Pm + 1, ......, PnrtJ, comprising a total of m + n (η≥m≥2) points, the set point clockwise two groups resultant vector P = (P1, P2, ...... Pm, Pm + 1, ......,? m + n}, comprising a total of m + n (n≥m≥2) a

点,将点集按顺时针方向两两组合成向量 Point, the set point clockwise two groups resultant vector

Figure CN104021670AD00051

共m+n个 Total m + n th

向量。 vector. 依次计算其余弦值,如点q: Sequentially calculated cosine, such as point q:

当q = I时,计算巧巧与H两向量的余弦值; When q = I, the calculation of H Qiaoqiao two cosine vector;

当l〈q〈m+n时,计算以点q为起点的向量和点q后一点即点q_l为终点的向量的余弦值,即计算 When l <q <m + n, is calculated after the starting point of the vector q and q points to the point that the vector end point q_l cosine value, i.e., calculates

Figure CN104021670AD00052

其中Θ为向量弋匕,和向量.的夹 Where Θ is the vector Yi dagger, and vectors. Folders

角; angle;

当q = m+n时,计算愚+„&+„ ^P1P2的余弦。 When q = m + n, is calculated yu + "& +" ^ P1P2 cosine.

[0010] 其中最小的余弦值所对应的点,即为道路首尾点(因为在图像中面状路的出入口处两端的线段近似平行)。 [0010] wherein the minimum value of the cosine of the corresponding point is the end-point of a road (road shape as shown in the image plane line segment approximately parallel to the inlet ends).

[0011] b、道路中心线:如图2所示,假设道路首尾点为APdPPm。 [0011] b, centerline of the road: 2, inclusive assumed road point APdPPm. 将面状道路在首尾点处打断形成两个道路边线点集,A= {P1; P2,……PJ和B={Pm+1,Pm+2,……,Pm+n}。 The planar road interrupted to form two sets road edge points, A = {P1 at the head and tail points; P2, ...... PJ and B = {Pm + 1, Pm + 2, ......, Pm + n}. 在点P1处采用顺时针方向组合点集A,即依次连接PpP2、……Pm形成道路边线X ;点Pm+1处采用逆时针组合点集B,即Pnrt^Pnrtr1、……、Pm+1、Pm形成道路边线Y。 Employed at the point P1 in a clockwise direction composition set point A, which in turn is connected PpP2, ...... Pm X-formed road edge; point Pm + 1 at the point set using counterclockwise composition B, that Pnrt ^ Pnrtr1, ......, Pm + 1 , Pm form a road sideline Y. 取组成道路边线点集中包含点个数较少的点集中的初始点,假设为点P1,分别计算点P1和点Pm+n,点P1和点Pnrtri中点坐标,依次类推,计算两道路边线点集的中点形成点集C= IC1, C2,……,Cn_1; CJ,共包含η个点,将点集C依次连接成线即为道路中线; Composition take the road edge point set comprising fewer number of points in the set initial point, is assumed as a point P1, and points P1 + n, the points P1 and Pm Pnrtri midpoint coordinate points are calculated, and so on, two roads calculated edge forming a midpoint point set point set C = IC1, C2, ......, Cn_1; CJ, contains η points, the point C are sequentially set in a line that is connected to the road centerline;

C、划分道路搜索块:将点集C形成的道路中线按一定长度L打断,L的选取,根据实际常见城市道路拥堵时队列最小长度选取。 C, dividing the road search block: the set point of the road centerline C is formed by a length L interrupted, L the case, the selected queue according to the minimum length of the actual common urban road congestion. 形成中线定长点集D= {D1; D2,……,Di+Dj,共i个点,最后不足L的部分不再分类;定义道路边线长度与道路中线长度的比值为比例因子Φ,将两条边线道路边线分别按ίΧΦ进行打断,分别形成两条边线点集X={X1; X2,……,Xh,XJ, Y= (Y1, Y2,……,Yh,YJ,此时对点集D、X、Y中的相邻两点进行连接,形成以中线为分界线的左右两边道路搜索块; Forming fixed length middle point set D = {D1; D2, ......, Di + Dj, i total points is less than the final classification is no longer part of L; defined as the ratio of road edge path length of the centerline length of the scale factor [Phi], the road edges two edges respectively ίΧΦ be interrupted, the two edges are formed set of points X = {X1; X2, ......, Xh, XJ, Y = (Y1, Y2, ......, Yh, YJ, at this time set of points D, X, Y connecting adjacent two points, as the boundary line is formed to both sides of the road search block;

d、重复a、b、C,对城市路网中所有面状道路进行道路搜索块划分。 d, repeating a, b, C, all of the urban road network in a planar way road search block division.

[0012] ( 3 )、计算各道路搜索块中车辆占有率。 [0012] (3), the search block is calculated for each road vehicle occupancy.

[0013] 道路车辆面积占有率的定义为:在道路中的车辆面积总和占某搜索块路段面积的百分数。 [0013] is defined as a road vehicle occupancy area: the sum of the vehicle in the road area percentage of the area of ​​a search block segment.

[0014] 如用字母Z表示,则可按(I)式的数学模式计算: [0014] As represented by the letter Z, may be mathematically model formula (I) is calculated:

Figure CN104021670AD00053

式中,Z代表某搜索块路中的道路车辆面积占用率,Sdf表该搜索路块中的单个车辆面积,Σ 则代表在该搜索块路中车辆面积的总和,Sk代表某搜索块路的面积。 Formula, Z represents a road vehicle area occupancy rate of a search block passage of, Sdf Table single vehicle the area of ​​the search path block, Σ sum of the vehicle an area of ​​the search block passage represents, Sk representative of a search block passage area. [0015] (4)阈值法检测道路队列信息;设定拥堵状态下的道路面积占用率为阈值z,若所述道路搜索块的道路面积占用率大于或等于阈值z的,判定为包含拥堵车辆队列的道路搜索块,则此时若干拥堵道路搜索块的总长度,即为是该路段上拥堵车辆队列的长度。 [0015] (4) a threshold detected road queue information; and setting the road area in the congestion state of occupancy rate threshold value z, road area occupancy if the path search block is greater than or equal to the threshold value of z is determined to include the congestion of the vehicle path search block of the queue, the total length of the road congestion case several search block, i.e. the length of the vehicle on the road congestion queue. 拥堵车辆队列搜索块的中心即认为是该路段上某个车辆队列所在的位置,结合电子地图或GPS,即可获得该队列所在的区域道路名称。 Center search block queue congestion vehicle that is believed to be the location of a vehicle on the road where the queue, combined with electronic map or GPS, you can get the name of the queue area of ​​road is located.

[0016] 本发明采用高分遥感影像实现车辆队列的检测,具有宏观性好、可视化程度高、检测区域面积大。 [0016] The present invention utilizes remote sensing to achieve high detection vehicle train having a macroscopic, high degree of visibility, a large area of ​​detection. 检测结果可为交通管理和规划部门科学决策提供准确的数据支撑。 Test results can be provided accurate data to support traffic management and planning department of scientific decision-making. 应用本发明的方法,同比与对比文献中的车辆拥堵队列提取方法,具有计算精度的优点。 The method of application of the present invention, compared with the reference year of vehicle congestion queue extraction method has the advantage of accuracy. 同时,由于本发明,采用迭代搜索法,构建迭代公式即可实现循环计算,在计算机硬件不断升级和CPU计算次数不断提升下,计算速度越来越短的优势越来越明显。 Meanwhile, since the present invention uses an iterative search method, constructed iterative formula calculation cycle can be realized, in the computer hardware and CPU escalating rising number of calculations, the calculation speed of the more obvious advantages shorter and shorter.

附图说明 BRIEF DESCRIPTION

[0017] 图1是依据遥感图像提取的面状道路边线示意图,其中集合P={P1; P2,……Pm,Pm+1,……,pm+n}是获得构成面状道路边线的点集; [0017] FIG. 1 is a schematic view of a planar road edge extraction based on remote sensing images, wherein the set P = {P1; P2, ...... Pm, Pm + 1, ......, pm + n} are obtained edge points constituting the road surface shape set;

图2是面状道路提取道路中线的示意图,其中集合C=IC1, C2,……,Cn_1; CJ是构成道路中线的点集; FIG 2 is a schematic view of a road surface shape extraction road centerline, wherein the set C = IC1, C2, ......, Cn_1; CJ is a set of points constituting the road line;

图3是形成道路搜索块示意图,集合X= (X1,X2,……,Xi^XihY=H……,Yi+YJ, D= {D1; D2,……,Dh,DJ为两边道路搜索块边线点集; FIG 3 is a block schematic form a search path, a set of X = (X1, X2, ......, Xi ^ XihY = H ......, Yi + YJ, D = {D1; D2, ......, Dh, DJ for the search block both sides of the road Side Lines set;

图4是基于高分遥感影像判别车辆队列的方法流程图。 FIG 4 is a method of remote sensing image score is determined based on a flowchart of the vehicle train.

具体实施方式 detailed description

[0018] 为了更清楚地说明本发明实施例的技术方案,结合附图对本发明作进一步说明: 本发明突出的方法步骤如图4所示: [0018] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, in conjunction with the accompanying drawings of the present invention is further described: The method of the present invention, the step of projecting shown in Figure 4:

(1)、应用图像处理相关算法,基于遥感影像提取面状道路和面状车辆矢量信息; (1), related to the application of image processing algorithms, remote sensing image extracting vector information on the basis of road shape and the planar surface of the vehicle;

(2)、提取面状道路中心线和边线,形成搜索块路; (2), and extracted road centerline planar edges, path search block is formed;

a、获取面状道路,找到面状道路的首尾点:如附图1所示,首先获得构成面状道路边界的点集P = (P1, P2,……Pm, Pm+1,……,PnrtJ,共包含m+n (η≥m≥2)个点,将点集按顺时 a, obtaining road planar, end to end to find the point of the road surface shape: as shown in Figure 1, is first obtained set of points form a planar road boundary P = (P1, P2, ...... Pm, Pm + 1, ......, PnrtJ, comprising a total of m + n (η≥m≥2) points, when the set point by cis

针方向两两组合成向量 Two pin sets a resultant vector direction

Figure CN104021670AD00061

写共m+n个向量。 Write total m + n vectors. 依次计算其 In order to calculate the

余弦值,如点q: Cosine values, such as point q:

当q = I时,计算P1P2与U議两向量的余弦值; When q = I, the value of the cosine vector P1P2 with two U meeting;

当l〈q〈m+n时,计算以点q为起点的向量丨,,,和点q后一点即点q_l为终点的向量 When the l <q <m + n, is calculated in the vector starting point q and the point q ,,, Shu the point that the end point of a vector q_l

___P.PP ___ P.PP

7UfU 的余弦值,即计算 7UfU cosine value, i.e., calculates

Figure CN104021670AD00062

,其中Θ 为向量ξξ;和向量/^ξ—;的夹 , Where Θ is the vector ξξ; and vector / ^ ξ-; folders

角; angle;

当q = m+n时,计算H与PlPz的余弦。 When q = m + n, the calculation of H and PlPz cosine.

其中最小的余弦值所对应的点,即为道路首尾点(因为在图像中面状路的出入口处两端的线段近似平行); Wherein the minimum value of the cosine of the corresponding point is the end-point of the road (as in the image of the planar passage across the inlet of a line segment approximately parallel to);

b、道路中心线:如图2所示,假设道路首尾点为点P1和Pm。 B, road centerline: 2, the path is assumed as the point P1 and the end-point Pm. 将面状道路在首尾点处打断形成两个道路边线点集,A = (P1, P2,……PJ和B= {Pm+1,Pm+2,……,Pm+n}。在点P1*采用顺时针方向组合点集A,即PpP2、……Pm为依次连接成线段形成道路边线X ;点Pm+1处采用逆时针组合点集B,即……、Pm+1、Pm形成道路边线Y。取组成道路边线点集中包含点个数较少的点集中的初始点,假设为点P1,分别计算点P1和点Pm+n,点P1和点Pnrtri The planar end-road interrupted at two points forming a road edge point set, A = (P1, P2, ...... PJ and B = {Pm + 1, Pm + 2, ......, Pm + n}. At point P1 * using a combination of clockwise point set A, i.e. ppP2, ...... Pm is in turn connected to X-line segments forming a road edge; point Pm + 1 at the point set using counterclockwise composition B, that ......, Pm + 1, Pm formed Y. road edge path taken consisting edge point set point number comprising fewer points set initial point, is assumed to be points P1, calculates the points P1 and Pm + n, the points P1 and Pnrtri

中点坐标,依次类推,计算两道路边线点集的中点形成点集C = (C1, C2,......, cn_1; cn},共 Midpoint coordinates, and so on, a road edge point midway between the calculated set point set formed C = (C1, C2, ......, cn_1; cn}, total

包含η个点,将点集C依次连接成线即为道路中线; Contains η points, the point C are sequentially set in a line that is connected to the road centerline;

C、划分道路搜索块:将点集C形成的道路中线按距离L = 20米(根据经验和实际调查,选取最短队列长度L = 20米,因中国轿车一般车长为5米,大巴和公交一般在10米,取最常见的3辆轿车拥堵或两辆大巴排队的长度)打断,形成中线定长点集D= {D1; D2,……,Di+ DJ,共i个点,如图3中五角星标志,最后道路中线中长度不足d的部分不再分类。 C, dividing the road search blocks: the waypoint set C formed in the line by L = 20 meters (based on experience and the actual survey, with the shortest queue length L = 20 meters due to Chinese car general automobiles length of 5 meters, the bus and bus typically 10 meters, taking the most common two or three bus cars congestion queue length) interrupted to form a fixed length middle point set D = {D1; D2, ......, Di + DJ, i total points, as shown in 3 Pentastar, the last part of the road is less than the length d of the line is no longer in the classification. 定义道路边线长度与道路中线长度的比值为比例因子Φ,由于X、Y长度不同,故对应的两条边线比例因子也不同,具体大小,以计算结果为准。 Defined as the ratio of road edge path length of the centerline length of the scale factor [Phi], due to the different X, Y length, so the two edges corresponding to the different scale factor, the specific size, subject to the calculation result. 将两条边线道路边线分别按LX Φ进行打断,分别形成两条边线点集X = {X1; X2,……,X^ijXJjY = (Y1, Y2,……U},此时对点集D、X、Y中的相邻两点进行连接,形成以中线为分界线的左右两边道路搜索块;如图3中三角形标志,都包含i个点。将形成的点集依次连接成线段,即分别在点集D、X、Y中依次取两点组合,共生成21-2段搜索块路; The two edges of the road edges respectively LX Φ be interrupted, the two edges are formed set of points X = {X1; X2, ......, X ^ ijXJjY = (Y1, Y2, ...... U}, at this time point set D, X, Y connecting adjacent two points, as the boundary line is formed to both sides of the road search block; triangle mark in FIG. 3, comprising i points are points set to be formed sequentially connected line segments. i.e. the set of points D, X, Y points sequentially taken in combination, a total path search block generating section 21-2;

d、重复a、b、c,直到所有面状道路都形成搜索块道路; d, repeating a, b, c, until all roads are formed planar path search block;

(3)、计算各道路搜索块中车辆占有率。 (3), the search block is calculated for each road vehicle occupancy.

[0019] 道路车辆占有率的定义为:在道路中的车辆面积总和占某搜索块路段面积的百分数。 Definitions [0019] Road vehicle occupancy is: the total area of ​​a road vehicle in a percentage of the area of ​​the search block segment. 如用字母Z表示,则可按(I)式的数学模式计算: As indicated by the letter Z, may be mathematically model formula (I) is calculated:

Figure CN104021670AD00071

式中,Z代表某搜索块路中的道路车辆面积占用率,Sdf表该搜索路块中的单个车辆面积,Σ 则代表在该搜索块路中车辆面积的总和,Sk代表某搜索块路的面积。 Formula, Z represents a road vehicle area occupancy rate of a search block passage of, Sdf Table single vehicle the area of ​​the search path block, Σ sum of the vehicle an area of ​​the search block passage represents, Sk representative of a search block passage area.

[0020] (4)、阈值法检测道路队列信息;设定某队列状态下的道路面积占用率为阈值z为70% (考虑到车辆间距,一条道路不可能全部被车首尾无间隙占满),大于阈值z的,即判定为该程度下的拥堵车辆队列搜索块,则此时若干拥堵道路搜索块的总长度,即为是该路段上拥堵车辆队列的长度。 [0020] (4), the threshold queue information detected by the road; a road area occupancy rate threshold is set at a z queue status of 70% (taking into account the inter-vehicle distance, a road vehicle is not all inclusive filled without gaps) , z is greater than the threshold, i.e., the vehicle is determined that the congestion degree of the queue for the search block, at this time the total length of the road congestion several search block, i.e. the length of the vehicle on the road congestion queue. 拥堵车辆队列搜索块的中心即认为是该路段上某个车辆队列所在的位置,结合电子地图或GPS,即可获得该队列所在的区域道路名称。 Center search block queue congestion vehicle that is believed to be the location of a vehicle on the road where the queue, combined with electronic map or GPS, you can get the name of the queue area of ​​road is located.

Claims (4)

1.一种利用高分遥感影像提取城市路网拥堵车辆队列信息的方法,其特征在于:选取高分辨率遥感影像中提取的面状道路和面状车辆矢量数据为输入参数,进而计算出车辆队列长度,结合电子地图或GPS数据,可获得车辆队列所在位置,至少包含以下步骤: 步骤一:从高分分辨率遥感影像中提取城市路网中面状道路和面状车辆信息数据; 步骤二:提取面状道路中心线和边线,并按设定的长度阈值进行截断,形成道路搜索块; 步骤三:计算所述道路搜索块的道路车辆占有率; 步骤四:以拥堵条件下的道路车辆占有率作为阈值,将步骤三中计算出的道路车辆占有率与所述阀值作比较,筛选出大于或等于该阈值的道路搜索块,则此时若干道路搜索块的总长度,即为是该路段上拥堵车辆队列的长度; 步骤五:将道路搜索块的中心位置与电子地图匹配或结合GPS数据相 A remote sensing images using a high extraction method of the urban road network congestion queue information of the vehicle, wherein: selecting a planar road vehicle and the planar resolution remote sensing image vector data extracted from the input parameters, and then calculate the vehicle queue length, combined with electronic maps or GPS data, position is obtained where the vehicle train, comprising at least the following steps: step 1: extraction of the urban road network and the road surface planar shape of the vehicle information data from high resolution remote sensing images; step two : extracting planar edges and road centerline, press set truncated length threshold, path search block formation; step three: calculating a road of the road vehicle occupancy search block; step four: a congested road condition under the vehicle occupancy as a threshold value, in step three the calculated road vehicle occupancy compared with the threshold value, the search block is selected path is greater than or equal to the threshold value, then at this time the total length of the road several search block, namely a length of the queue congestion on the road vehicle; step 5: the center position of the search block with an electronic road map matching or in combination with GPS data 合,获得城市路网中拥堵车辆队列位置信息。 Together, get queue position information of the vehicle congestion in urban road network.
2.如权利要求1所述的一种利用高分遥感影像提取城市路网拥堵车辆队列信息的方法,其特征在于:从城市路网中提取面状道路的中心线和边线,获取步骤二中所述道路搜索块,至少包含以下步骤: 步骤一:获取面状道路的首尾点:首先获得构成面状道路边界的点集P= (P1, P2,……Pm, Pm+1,……,Pm+n},共包含m+n (η≥m≥2)个点,将点集按顺时针方向两两组合成向量 As claimed in claim 2. The remote sensing method utilizing high road network congestion queue information extracting the vehicle 1, characterized in that: extracting a center line and road edges from the planar urban road network, the obtaining step II the path search block, comprising at least the following steps: step 1: obtaining a planar end-point path: first obtaining point set form a planar road boundary P = (P1, P2, ...... Pm, Pm + 1, ......, Pm + n}, comprising a total of m + n (η≥m≥2) points, the set point clockwise two groups resultant vector
Figure CN104021670AC00021
共m+n个向量,依次计算其余弦值;其中最小的余弦值所对应的点,即为道路首尾点; 步骤二:获取道路中心线:假设道路首尾点为点PdPPm,将面状道路在首尾点处打断形成两个道路边线点集,A= {P1; P2,……PJ和B={pm+1,Pm+2,……,Pm+n};在点P1处采用顺时针方向组合点集A,即依次连接PpP2、……Pm形成道路边线X ;点Pm+1处采用逆时针组合点集B,即Pm+n、Pnrtri'……、Pm+1、Pm形成道路边线Y ;取组成道路边线点集中包含点个数较少的点集中的初始点,假设为点P1,分别计算点P1和点Pm+n,点P1和点Pnrtri中点坐标,依次类推,计算两道路边线点集的中点形成点集C= IC1,C2,……,Clri,CJ,共包含η个点,将点集C中各点依次连接成线,即为道路中线; 步骤三:获取道路搜索块:将点集C形成的道路中线按一定长度L打断,L的选取,根据实际常见城市道路拥堵时队列最小长度选取 M + n vectors were sequentially calculate cosine; wherein the minimum value of the cosine of the corresponding point is the end-point of a road; Step Two: acquiring road centerline: end-point of a road assumed point PdPPm, the road surface shape two break points are formed at the beginning and end points of road edge set, A = {P1; P2, ...... PJ and B = {pm + 1, Pm + 2, ......, Pm + n}; clockwise using point P1 direction of the combined set of points A, which in turn is connected ppP2, ...... Pm X-formed road edge; point Pm + 1 at the point set using counterclockwise composition B, that Pm + n, Pnrtri '......, Pm + 1, Pm road edge is formed the Y; take the road edge points are concentrated composition contains less of the initial number of points in the set point, the point is assumed to be P1, and points P1 + n, the points P1 and Pm Pnrtri midpoint coordinate points are calculated, and so on, two computing midpoint road edge point set point set form C = IC1, C2, ......, Clri, CJ, contains η points, the set point C of a line connecting sequentially points, that is the road centerline; step three: Get road block Search: the road point set C is formed by a length L of the line break, L is selected, selecting a queue according to the actual length of the minimum common urban road congestion 形成中线定长点集D={D1; D2,……,Di+ DJ,共i个点,最后不足L的部分不再分类;定义道路边线长度与道路中线长度的比值为比例因子Φ,将两条边线道路边线分别按LX Φ进行打断,分别形成两条边线点集X={X1;X2,……,Xh,XJ, Y= (Y1, Y2,……,Yh,YJ,此时对点集D、X、Y中的相邻两点进行连接,形成以中线为分界线的左右两边道路搜索块; 步骤四:重复步骤一至三,对城市路网中所有面状道路进行道路搜索块划分。 Forming fixed length middle point set D = {D1; D2, ......, Di + DJ, i total points is less than the final classification is no longer part of L; defined as the ratio of the length of road edge line length scale factor [Phi] is the road, the two road edge strip edges respectively LX Φ be interrupted, the two edges are formed set of points X = {X1; X2, ......, Xh, XJ, Y = (Y1, Y2, ......, Yh, YJ, at this time set of points D, X, Y connecting adjacent two points, as the boundary line is formed to both sides of the road search block; step four: repeat steps one to three, all the urban road network for road planar path search block division.
3.如权利要求1所述的一种利用高分遥感影像提取城市路网拥堵车辆队列信息的方法,其特征在于:计算所述道路车辆面积占有率采用的数学模式为: According to claim 1. A method of remote sensing images using a high road network congestion queue information extracted vehicle, wherein: calculating the mathematical model used in road vehicles area occupancy is:
Figure CN104021670AC00022
其中,Z代表道路搜索块中的道路车辆面积占用率,S车代表道路搜索块中的单个车辆的面积 Wherein the area occupancy rate of the area of ​​road vehicles and Z represents a search block in the road, S road car represents a single search blocks in a vehicle
Figure CN104021670AC00031
则代表在道路搜索块中车辆面积的总和,S路代表道路搜索块的面积。 Represents the sum of the areas of the vehicle in the road search block, S is the road path search block area.
4.如权利要求1所述的一种利用高分遥感影像提取城市路网拥堵车辆队列信息的方法,其特征在于:设定拥堵状态下的道路面积占用率为阈值z,所述道路搜索块的道路面积占用率大于或等于阈值z的,判定为包含拥堵车辆队列的道路搜索块,则此时若干拥堵道路搜索块的总长度,即为是该路段上拥堵车辆队列的长度。 As claimed in claim image sensing method utilizing high road network congestion queue information extracting the vehicle 1, characterized in that: the road area occupancy rate of the threshold value z set congestion state of the road search block the road area occupancy rate is greater than or equal to the threshold value z determines the congestion road vehicle comprising a search block of the queue, at this time the total length of the road congestion several search block, i.e. the length of the vehicle on the road congestion queue.
CN 201410118280 2014-03-27 2014-03-27 One kind of remote sensing images to extract high urban road network of the vehicle queue status information method CN104021670B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN 201410118280 CN104021670B (en) 2014-03-27 2014-03-27 One kind of remote sensing images to extract high urban road network of the vehicle queue status information method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN 201410118280 CN104021670B (en) 2014-03-27 2014-03-27 One kind of remote sensing images to extract high urban road network of the vehicle queue status information method

Publications (2)

Publication Number Publication Date
CN104021670A true true CN104021670A (en) 2014-09-03
CN104021670B CN104021670B (en) 2016-06-15

Family

ID=51438400

Family Applications (1)

Application Number Title Priority Date Filing Date
CN 201410118280 CN104021670B (en) 2014-03-27 2014-03-27 One kind of remote sensing images to extract high urban road network of the vehicle queue status information method

Country Status (1)

Country Link
CN (1) CN104021670B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105157710B (en) * 2015-06-06 2016-09-07 烟台惠通网络技术有限公司 Based on Link congestion level sensing detection system in communication

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5862244A (en) * 1995-07-13 1999-01-19 Motorola, Inc. Satellite traffic reporting system and methods
CN101980317A (en) * 2010-09-03 2011-02-23 浙江大学 Method for predicting traffic flow extracted by improved C-V model-based remote sensing image road network
CN102567963A (en) * 2011-11-10 2012-07-11 西安电子科技大学 Quantum multi-target clustering-based remote sensing image segmentation method
CN102609706A (en) * 2011-01-25 2012-07-25 龚建华 System for extracting road information from high-resolution remote-sensing image
US20140074566A1 (en) * 2012-09-11 2014-03-13 James K. McCoy Remote Sensing of Vehicle Occupancy

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5862244A (en) * 1995-07-13 1999-01-19 Motorola, Inc. Satellite traffic reporting system and methods
CN101980317A (en) * 2010-09-03 2011-02-23 浙江大学 Method for predicting traffic flow extracted by improved C-V model-based remote sensing image road network
CN102609706A (en) * 2011-01-25 2012-07-25 龚建华 System for extracting road information from high-resolution remote-sensing image
CN102567963A (en) * 2011-11-10 2012-07-11 西安电子科技大学 Quantum multi-target clustering-based remote sensing image segmentation method
US20140074566A1 (en) * 2012-09-11 2014-03-13 James K. McCoy Remote Sensing of Vehicle Occupancy

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
张勇 等: "高分遥感影像在交通限行决策中的应用", 《公路与汽运》, no. 157, 31 July 2013 (2013-07-31), pages 69 - 71 *
张素兰: "基于卫星遥感影像的交通状态判别研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》, no. 10, 15 October 2010 (2010-10-15), pages 15 - 53 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105157710B (en) * 2015-06-06 2016-09-07 烟台惠通网络技术有限公司 Based on Link congestion level sensing detection system in communication

Also Published As

Publication number Publication date Type
CN104021670B (en) 2016-06-15 grant

Similar Documents

Publication Publication Date Title
Ismail et al. Automated analysis of pedestrian-vehicle: conflicts context for before-and-after studies
Ismail et al. Automated analysis of pedestrian-vehicle conflicts using video data
Biagioni et al. Easytracker: automatic transit tracking, mapping, and arrival time prediction using smartphones
Angel et al. Methods of analyzing traffic imagery collected from aerial platforms
Karagiorgou et al. On vehicle tracking data-based road network generation
Sayed et al. Automated safety diagnosis of vehicle–bicycle interactions using computer vision analysis
Chen et al. Probabilistic modeling of traffic lanes from GPS traces
Serna et al. Urban accessibility diagnosis from mobile laser scanning data
Kong et al. Efficient traffic state estimation for large-scale urban road networks
Yu et al. Object-based spatial cluster analysis of urban landscape pattern using nighttime light satellite images: A case study of China
Hinz et al. Car detection in aerial thermal images by local and global evidence accumulation
Yao et al. Extraction and motion estimation of vehicles in single-pass airborne LiDAR data towards urban traffic analysis
CN101819717A (en) Road network performance judgment method based on traffic state space-time model
CN103093249A (en) Taxi identifying method and system based on high-definition video
Yan et al. A multi-parameter approach to automated building grouping and generalization
Khattak et al. Highway safety assessment through geographic information system-based data visualization
CN103020342A (en) Method for extracting contour and corner of building from ground LiDAR data
Schneider et al. A new urban landscape in East–Southeast Asia, 2000–2010
Ismail et al. Automated collection of pedestrian data using computer vision techniques
CN103150559A (en) Kinect three-dimensional depth image-based head identification and tracking method
Liu et al. Road recognition using coarse-grained vehicular traces
Jianming et al. Traffic congestion identification based on image processing
Konrad et al. Localization in digital maps for road course estimation using grid maps
Oliveira-Neto et al. Tracking large trucks in real time with license plate recognition and text-mining techniques
CN103500338A (en) Road zebra crossing automatic extraction method based on vehicle-mounted laser scanning point cloud

Legal Events

Date Code Title Description
C06 Publication
C10 Entry into substantive examination
C14 Grant of patent or utility model