CN108898840A - A kind of intelligent traffic lamp control method based on video monitoring - Google Patents

Abstract

The intelligent traffic lamp control method based on video monitoring that the invention discloses a kind of, belongs to field of locating technology, includes the following steps：Background image is chosen from video monitoring image；Foreground image is extracted using background subtraction；Estimate Vehicle length；Distribute the traffic light time in each lane.The present invention passes through the validity and feasibility of this paper system of the experiment show in DSP analogue system simulation traffic condition, time algorithm is distributed compared to traditional traffic lights, traffic lights vehicle pass-through efficiency is improved, traffic jam environment is effectively relieved, fully meets the requirement of true traffic condition.

Description

A method for controlling intelligent traffic lights based on video surveillance

technical field

The invention relates to a traffic signal lamp control method, in particular to an intelligent traffic signal lamp control method based on video monitoring, which belongs to the field of positioning technology.

Background technique

At present, the number of automobiles in my country exceeds 190 million. Especially in recent years, the total number of automobiles has increased rapidly, resulting in more and more frequent traffic failures and urban traffic congestion. Although the roads are improving day by day, the coordination of the relationship between people, vehicles and roads has not been satisfactorily resolved. The signal light control system can no longer satisfy the current urban traffic conditions. Under this background, the intelligent control traffic signal light system came into being. Its main purpose is to intelligently control traffic signal lights to improve road utilization, improve traffic congestion, and reduce automobile environmental pollution.

Developed countries such as Europe and the United States have already started a lot of research and practice to solve this problem. Technologies in different fields have also been combined and applied to the research of intelligent transportation. For intelligent traffic light timing schemes, there are probabilistic, fuzzy Control, digital image processing, and even neural networks and genetic immune algorithms in biology are applied to timing strategies. my country's big cities are more and more affected by traffic congestion, and the level of research and practice is lower than that of developed countries. There is still a big gap. Many small and medium-sized cities still use the most primitive fixed-length time distribution scheme. The details of mature intelligent transportation systems in developed countries in Europe and the United States have not been made public. Therefore, it is of great significance for relevant researchers in my country to develop intelligent transportation algorithms with independent property rights.

Contents of the invention

The main purpose of the present invention is in order to provide a kind of intelligent traffic signal light control method based on video monitoring, under the condition of existing traffic facilities, utilize image processing and neural network related method to calculate in real time from video monitoring each phase waiting vehicle traffic It is of great practical significance to realize intelligent control of traffic lights, improve road utilization, and reduce traffic congestion.

The purpose of the present invention can be achieved by adopting the following technical solutions:

A method for controlling intelligent traffic lights based on video monitoring, comprising the following steps:

Step S1: selecting a background image from video surveillance images;

Step S2: Extracting the foreground image by using the background subtraction method;

Step S3: Estimate the length of the vehicle;

Step S4: Allocate traffic light time for each lane.

Further, in step S1, selecting a background image from video surveillance images includes:

Assuming that the season is represented by Season, its values are spring, summer, autumn, and winter. The background images under the fixed Season variable are dawn, early morning, morning, a period of time from morning to afternoon, dusk, and those not in the above time periods. A total of 7 background images;

There may be three situations of rain, fog and snow every day, and a background image is taken in each of the above situations;

According to the amount of rainfall, rain is divided into light rain, moderate rain, heavy rain and heavy rain;

According to the degree of visibility, fog is divided into light fog, fog, heavy fog, dense fog and strong dense fog;

According to the amount of snow, snow is divided into light snow, moderate snow and heavy snow.

Further, in the step S2, the background difference method uses the difference between the current image and the background image to detect the moving area, selects a frame of image as the background image, and performs a difference operation between the current image including the background and the foreground and the background image, using the following formula conduct:

Where: C _n (i, j) represents the pixel grayscale of the i-th row and j-th column of the n-th image;

B _n (i, j) represents the pixel grayscale of the i-th row and j-column of the n-th background image;

F _n (i, j) is the differential pixel grayscale of the i-th row and j-column of the n-th image;

T is the threshold.

Further, in the step S3, estimating the vehicle length includes the following steps:

Step S31: Calibration of the surveillance camera

Estimate the length of the waiting vehicle in the real environment from the pixels of the waiting vehicle in the image, ignore the radial and tangential distortion of the camera, ignore the error caused by the vehicle not facing the camera, and establish a mathematical model;

Step S32: Estimate vehicle length

For any line segment HG between M ₁ and M ₂ , the mapping points on the line segment of points A and B are C and D respectively, assuming that the distances from C and D to points A and B are d ₁ and d ₂ respectively, assuming d ₁ < d ₂ , assuming that the coordinates of point A are (x _a , y _a );

Get:

Straight lines SC and SD are thus determined.

Further, in the step S31, the established mathematical model is shown in the following formula:

After calibration, the lines l ₁ , l ₂ and l ₃ can be determined, and then the coordinates of points A and B can be obtained as:

Among them: the variable b is determined according to the distance between A and B, which is the pixel height of the image collected by the camera, and the solution expression is:

Among them: the unit of line l3 is pixel, the unit _of the actual distance between line _l1 and line _l2 is _meter , the projection of any line segment between M1 and _{M2 on} line l3 and _l4 occupies line segment AB and line segment EF is in equal proportion.

Further, in the step S32, the straight line SC and the straight line SD are respectively:

Straight SC:

Straight line SD:

The coordinates of solving points G and H are respectively:

Find the length of GH as:

The length of the GH is equal to the vehicle length.

Further, in the step S4, allocating the traffic light time of each lane includes the following steps:

Step S41: Assuming that the red and green time of a cycle is T, the time required for the vehicles in one direction to complete driving, their phases are l ₁ , l ₂ , l ₅ , l ₆ respectively, and the corresponding time for all vehicles to pass is AT ₁ , AT ₂ , AT ₅ , AT ₆ , sort the passage time of these four phases from the longest to the shortest, if the time is the same, sort according to AT ₁ to AT ₈ , and the results after sorting are assumed to be t ₁ and t ₂ , t ₃ , t ₄ , the phase corresponding to t ₁ not only takes a long time to pass, but also is ahead of AT ₁ to AT ₈ , allowing it to pass first;

Step S42: t ₁ , t ₂ , t ₃ , and t ₄ are the time required for the complete passage of the corresponding phases, the phase corresponding to t ₁ passes first, and the phase corresponding to t ₂ or the phase corresponding to t ₃ can pass at the same time;

When phase t ₁ passes through, select any one of the phases corresponding to t ₂ and t ₃ to pass, and during the passage of phase t ₁ , the phase passing at the same time as it ends;

When the phase corresponding to t ₁ passes at the same time, another phase that allows simultaneous passage can pass when the phase corresponding to t ₁ passes;

When the passage of t ₁ ends, the phase corresponding to t ₄ can pass, and when the phase corresponding to t ₄ and the current passage phase both complete, the passage in this direction ends;

Step S43: Assume that the phases in the same direction are sorted according to the complete transit time from long to short. If there are the same, sort them from AT ₁ to AT _8. The first one is t ₁ , and t ₁ corresponds to the phase that is allowed to pass when the phase passes. The complete passage time is t ₂ and t ₃ , and the remaining one is t ₄ , let t ₁ pass first, and the time for vehicles in this direction to pass completely is TW;

If two phases are allowed to pass first, and then the other two phases are allowed to pass, the complete passage time is TY;

Calculate TY≥TW, and select the TW corresponding method, that is, once a phase is completed, the phases allowed by it can pass without affecting other phases.

Furthermore, the time TW for the vehicle to pass completely is:

If two phases are allowed to pass first, the complete passage time TY is:

When (t ₂ +t ₃ -t ₁ )≤t ₄ , that is, t ₂ +t ₃ ≤t ₁ +t ₄ , if t ₂ ≥t ₄ , t ₁ +t ₂ ≥t ₁ +t ₄ , if t ₂ <t ₄ , t ₁ +t ₄ =t ₁ +t ₄ , therefore, TY≥TW;

When (t ₂ +t ₃ -t ₁ )>t ₄ , that is, t ₂ +t ₃ >t ₁ +t ₄ , if t ₂ ≥t ₄ , t ₁ +t ₂ ≥t ₂ +t ₃ , if t ₂ <t ₄ , t ₁ +t ₄ ≥t ₂ +t ₃ , therefore, TY≥TW.

Further, in the step S43,

Calculate the complete passage time of vehicles in the horizontal direction as TW ₁ , the passage time in the vertical direction as TW ₂ , and the total time of traffic lights as T, and the time can be allocated as follows:

The horizontal allocation time is:

The vertical travel time is:

When TW ₁ +TW ₂ = 0, there is no vehicle on the current road, and the time is set according to the default timing method.

Beneficial technical effects of the present invention: according to the intelligent traffic light control method based on video monitoring of the present invention, the intelligent traffic signal light control method based on video monitoring provided by the present invention verifies the system in this paper by the experimental results of simulating traffic conditions in DSP emulation system Compared with the traditional traffic light allocation time algorithm, it improves the traffic efficiency of traffic light vehicles, effectively alleviates traffic jams, and fully meets the requirements of real traffic conditions.

Description of drawings

Fig. 1 is the flow chart of a preferred embodiment of the intelligent traffic light control method based on video monitoring according to the present invention;

Fig. 2 is according to the monitoring camera calibration method figure of a preferred embodiment of the intelligent traffic signal light control method based on video monitoring of the present invention;

Fig. 3 is the vertical image interface diagram of the center of the camera according to a preferred embodiment of the intelligent traffic light control method based on video surveillance of the present invention;

Fig. 4 is the simplified schematic diagram of the actual traffic crossing according to a preferred embodiment of the intelligent traffic light control method based on video surveillance of the present invention;

FIG. 5 is a diagram of assumed conditions of a preferred embodiment of a method for controlling intelligent traffic lights based on video monitoring according to the present invention.

Detailed ways

In order to make the technical solutions of the present invention clearer and clearer to those skilled in the art, the present invention will be further described in detail below in conjunction with the examples and accompanying drawings, but the embodiments of the present invention are not limited thereto.

As shown in Figure 1, a kind of intelligent traffic light method based on video surveillance provided by the present embodiment comprises the following steps:

Step S1, selecting a background image from video surveillance images;

Step S2, using the background difference method to extract the foreground image (also called target or vehicle);

Step S3, estimating the length of the vehicle;

Step S4, allocating the traffic light time of each lane;

In this embodiment, in step S1, the background image is selected from the video surveillance image, and the background image is selected from the video surveillance image. Due to the particularity of the application environment, the position of the surveillance camera is fixed, and the shooting range is relatively fixed. of. The change of the captured background image is mainly affected by the light intensity, and the light intensity is almost different in different seasons and in the same season every day. Assuming that the season is represented by Season, its value is spring, summer, autumn and winter. Under the fixed Season variable The background images can be selected from dawn, early morning, morning, a period of time from morning to afternoon, dusk, and a total of 7 background images that are not in the above time periods; and there may be rain every day (divided into light rain, moderate rain, heavy rain and heavy rain ), fog (according to the visibility can be divided into light fog, fog, heavy fog, dense fog and strong dense fog) and snow (light snow, moderate snow and heavy snow), each situation has several other situations , are all within the scope of protection, and a background image is taken in each of the above cases;

In this embodiment, in step S2, the background difference method is used to extract the foreground image (also known as the target or vehicle). The background difference method is a common method for detecting moving objects in the visual system. A technique of motion area, the basic idea of the background difference method is to select a frame of image as the background image, that is, step S1 to obtain the background image, and perform a difference operation between the current image (including the background and the foreground) and the background image, if the reference image is selected as Appropriate, the foreground can be more accurately segmented, that is, the target or the vehicle. The formula is:

Where: C _n (i, j) represents the pixel grayscale of the i-th row and j-th column of the n-th image;

B _n (i, j) represents the pixel grayscale of the i-th row and j-column of the n-th background image;

F _n (i, j) is the differential pixel grayscale of the i-th row and j-column of the n-th image;

T is the threshold.

In the present embodiment, in step S3, the vehicle length is estimated, and its steps are:

Step S31: Calibration of the surveillance camera

The purpose of calibrating the camera is to estimate the length of the waiting vehicle in the real environment from the length (pixels) of the waiting vehicle in the image. Since the camera is facing the road, the waiting vehicles are approximately vertical in the image, ignoring the path of the camera. To and tangential distortion, ignoring the error caused by the vehicle not facing the camera directly, and establishing a mathematical model, as shown in Figure 2 and Figure 3, in Figure 3, the line l ₃ is in pixels, and the distance between the line l ₁ and the line l ₂ The actual distance unit is meter (m). It is easy to prove that any line segment between points M ₁ and M ₂ has a linear relationship with the projection on lines l ₃ and l ₄ , that is, between M ₁ and M ₂ The projection of any line segment on line l ₃ and l ₄ occupies the same proportion of line segment AB and line segment EF, so line l ₃ is in pixels, which actually enlarges the actual coordinates of l ₃ , but the proportion of the image in the total image is still the same has not changed, and the size of the corresponding real object has not changed, so the line l ₃ can be in pixels;

After calibration, the lines l ₁ , l ₂ and l ₃ can be determined, assuming that they are:

Then the coordinates of points A and B can be obtained as:

Among them: the variable b can be determined according to the distance between A and B, which is the pixel height of the image collected by the camera.

The solution expression is:

By solving the quadratic equation of the above formula, the unknown variable b can be solved as:

and

Step S32: Estimate vehicle length

For any line segment HG between M ₁ and M ₂ , their mapping points on the line segments of points A and B are C and D respectively, assuming that the distances from C and D to points A and B are d ₁ and d ₂ respectively (assuming d ₁ <d ₂ ), then the coordinates of points C and D can be calculated from the coordinates of point A, the table of line l ₃ and d ₁ and d ₂ , assuming that the coordinates of point A are (x _a , y _a ), according to the geometry Find the relationship and From this, the straight lines SC and SD can be determined, which are:

The coordinates of points G and H can be solved as follows:

The length of GH can be calculated according to the distance formula between two points:

The length of the GH is equal to the vehicle length.

In the present embodiment, in the step S4, the traffic light time of each lane is allocated, and the steps are:

Figure 4 is a simplified schematic diagram of the actual traffic intersection. Assuming that the red and green time of a cycle is T, we can consider the time required for vehicles in one direction to complete driving. For example, in the horizontal direction, the phases are l ₁ , l ₂ , l ₅ , l ₆ , the time corresponding to the completion of all traffic is AT ₁ , AT ₂ , AT ₅ , AT ₆ . Sort the passing time of these four phases from longest to shortest. If the time is the same, sort them according to AT ₁ to AT _8. The results after sorting are assumed to be t ₁ , t ₂ , t ₃ , and t ₄ . The phase corresponding to t ₁ not only takes a long time to pass, but also is ahead of AT ₁ to AT ₈ , allowing it to pass first. Suppose the final result is shown in Figure 5 below;

t ₁ , t ₂ , t ₃ , and t ₄ in Fig. 5 are the time required for the complete passage of the corresponding phase. _The phase corresponding to t ₁ passes _first , then the phase corresponding to _{t 2} or the phase corresponding to t ₃ can pass at the same time. does not affect the total time of the final passage), during the passage of phase t ₁ , the phase passing at the same time as it must end, because the passage time of t ₁ is the longest, at least it will not be shorter than other phases, when the phase corresponding to t ₁ After the phase of simultaneous passage ends, when the phase corresponding to t ₁ passes, another phase that allows simultaneous passage can pass. When the passage of t ₁ ends, the phase corresponding to t ₄ can pass. When the phase corresponding to t ₄ and the current passage phase both pass After the end, the traffic in this direction ends;

Assume that the phases in the same direction are sorted from the longest to the shortest time of complete passage. If there are the same, they are sorted from AT ₁ to AT _8. The first one is t ₁ , and t ₁ corresponds to the complete passage of the passing phase when passing The time is t ₂ and t ₃ , and the remaining one is t ₄ , let t ₁ pass first, then, the time TW for vehicles in this direction to pass completely is:

If two phases are allowed to pass first, and then two phases are allowed to pass, in this way, under the above assumptions, the complete passage time TY is:

When (t ₂ +t ₃ -t ₁ )≤t ₄ , that is, t ₂ +t ₃ ≤t ₁ +t ₄ , if t ₂ ≥t ₄ , t ₁ +t ₂ ≥t ₁ +t ₄ , if t ₂ <t ₄ , t ₁ +t ₄ =t ₁ +t ₄ , so TY≥TW;

When (t ₂ +t ₃ -t ₁ )>t ₄ , that is, t ₂ +t ₃ >t ₁ +t ₄ , if t ₂ ≥t ₄ , t ₁ +t ₂ ≥t ₂ +t ₃ , if t ₂ <t ₄ , t ₁ +t ₄ ≥t ₂ +t ₃ , so TY≥TW;

Therefore, TY≥TW, choose the TW corresponding method, that is, once a phase is completed, the phase allowed by it can pass without affecting other phases;

In this way, the complete passage time of vehicles in the horizontal direction is TW ₁ , and the passage time in the vertical direction is TW ₂ . Since the total time of traffic lights is T, the time can be allocated as follows:

The horizontal allocation time is:

The vertical travel time is:

If TW ₁ +TW ₂ = 0 in the above two formulas, it means that there is no vehicle on the current road, and the time is adjusted according to the default timing method;

The time distribution in one direction is (taking the horizontal direction as an example):

In the formula, m is from 1 to 4, TS _m is the actual transit time of the four phases in the hypothesis, and t _m is the time required for complete passage;

If there is no traffic on the road in this direction, t ₁ +t ₄ =0 or t ₂ +t ₃ =0 may exist, which means that TW ₁ is 0. At this time, T ₁ itself is 0, and there is no allocation of time necessary. The vertical direction is similar to the horizontal direction, T ₁ and T ₂ cannot both be 0, because T is not 0, TW ₁ +TW ₂ ≠0;

If TW ₁ +TW ₂ =0, this situation has already been discussed, and the time is allocated by default.

There is a yellow light after the passing time of any phase, which is necessary and not counted in the total time T of traffic lights.

To sum up, in this embodiment, according to the intelligent traffic signal light control method based on video surveillance in this embodiment, the intelligent traffic signal light control method based on video surveillance provided in this embodiment, simulates traffic conditions in the DSP simulation system. The experimental results verify the effectiveness and feasibility of the system in this paper. Compared with the traditional traffic light allocation time algorithm, it improves the traffic efficiency of traffic light vehicles, effectively alleviates traffic jams, and fully meets the requirements of real traffic conditions.

The above is only a further embodiment of the present invention, but the protection scope of the present invention is not limited thereto, any person familiar with the technical field within the scope disclosed in the present invention, according to the technical scheme of the present invention and its Any equivalent replacement or modification of the concept falls within the protection scope of the present invention.

Claims (9)

1. a kind of intelligent traffic lamp control method based on video monitoring, which is characterized in that include the following steps：

Step S1：Background image is chosen from video monitoring image；

Step S2：Foreground image is extracted using background subtraction；

Step S3：Estimate Vehicle length；

Step S4：Distribute the traffic light time in each lane.

2. a kind of intelligent traffic lamp control method based on video monitoring as described in claim 1, which is characterized in that step In rapid S1, choosing background image from video monitoring image includes：

Assuming that season is indicated with Season, value is spring, summer, autumn, winter, the background image difference under fixed Season variable Take daybreak, early morning, morning, a period of time from morning to afternoon, dusk and not in totally 7 Backgrounds of the above period Picture；

And there may be three kinds of situations of raining, mist and snow daily, are taking a background image in each case above；

According to rainfall size, rains and be divided into light rain, moderate rain, heavy rain and heavy rain；

According to visibility size, mists and be divided into mist, mist, dense fog, thick fog and strong thick fog；

According to snowfall size, snows and be divided into slight snow, moderate snow and heavy snow.

3. a kind of intelligent traffic lamp control method based on video monitoring as described in claim 1, which is characterized in that institute It states in step S2, background subtraction utilizes the Differential Detection moving region of present image and background image, chooses a frame image and makees For background image, the current image comprising background and prospect and background image are done into difference operation, carried out using following formula：

Wherein：C_n(i, j) indicates the pixel grey scale of the i-th row jth column of the n-th width image；

B_n(i, j) indicates the pixel grey scale of the i-th row jth column of the n-th width background image；

F_n(i, j) is the differential pixel gray scale of the i-th row jth column of the n-th width image；

T is threshold value.

4. a kind of intelligent traffic lamp control method based on video monitoring as described in claim 1, which is characterized in that institute It states in step S3, estimates Vehicle length, include the following steps：

Step S31：The calibration of monitoring camera

The length for waiting vehicle in vehicle pixel estimation true environment by waiting in image, ignores the radially and tangentially abnormal of camera Become, ignores the vehicle error that face camera does not generate, founding mathematical models；

Step S32：Estimate Vehicle length

For M₁And M₂Between any line segment HG, the mapping point on A and B point line segment is respectively C and D, it is assumed that C and D are arrived respectively A and B point distance is d₁And d₂, it is assumed that d₁< d₂, it is assumed that A point coordinate is (x_a,y_a)；

It acquires：

Thereby determine that straight line SC and SD.

5. a kind of intelligent traffic lamp control method based on video monitoring as claimed in claim 4, which is characterized in that institute It states in step S31, the mathematical model of foundation is shown below：

k₃=tan α；

It can determine line l after calibration₁、l₂And l₃, and then the coordinate that can find out A and B point is：

Wherein：Variable b determines that the high height of the pixel of as camera collection image solves table according to the distance between A and B It is up to formula：

Wherein：Line l₃As unit of pixel, line l₁With line l₂Between actual range unit be rice, M₁、M₂Between any line segment exist Line l₃And l₄On projection the line is busy section AB and line segment EF be in equal proportions.

6. a kind of intelligent traffic lamp control method based on video monitoring as claimed in claim 4, which is characterized in that institute It states in step S32, straight line SC and straight line SD, respectively：

Straight line SC：

Straight line SD：

The coordinate of solution point G and H is respectively：

The length for finding out GH is：

The length of GH is equal to Vehicle length.

7. a kind of intelligent traffic lamp control method based on video monitoring as described in claim 1, which is characterized in that institute It states in step S4, distributes the traffic light time in each lane, include the following steps：

Step S41：Assuming that the red green equal times of a cycle are T, time of the complete needs of the vehicle driving in a direction is allowed, phase Position is l respectively₁、l₂、l₅、l₆, the corresponding time of all having passed through is AT₁、AT₂、AT₅、AT₆, this four phases are all passed through Time sorts from long to short, if the time is the same, according to AT₁To AT₈Sequence, the result after sequence are assumed to be t₁、t₂、 t₃、t₄, t₁Not only transit time is long for corresponding phase, but also in AT₁To AT₈In it is forward, allow it first to pass through；

Step S42：t₁、t₂、t₃、t₄It is the complete current required time of corresponding phase, t₁Corresponding phase is first passed through, while current It can be t₂Corresponding phase or t₃Corresponding phase；

Work as t₁During phase is current, t₂And t₃Any one passage is selected in corresponding phase, in t₁During phase is current, with it Current phase terminates simultaneously；

When with t₁After corresponding phase phase current simultaneously, t₁Allow the another of passage simultaneously when corresponding phase is current A phase P Passable；

Work as t₁After passage, t₄Corresponding phase P Passable, works as t₄After corresponding phase and current current phase are all passed through, This direction passage terminates；

Step S43：Assuming that unidirectional phase is sorted from long to short according to complete transit time, if there is likewise, pressing According to from AT₁To AT₈Sequence, first is t₁, t₁Corresponding phase allows the complete transit time of current phase to be t when current₂And t₃, It remains next for t₄, allow t₁It first passes through, this side up, and time that vehicle passed through completely is TW；

If first allowing two phases current, then make another two phase current, complete transit time is TY；

TY >=TW is calculated, TW corresponded manner is selected, is i.e. a phase is once covered, its permitted phase is not influencing other phases In the case of position, P Passable.

8. a kind of intelligent traffic lamp control method based on video monitoring as claimed in claim 7, which is characterized in that vehicle The time TW to have passed through completely is：

If first allowing two phases current, complete transit time TY is：

As (t₂+t₃-t₁)≤t₄, i.e. t₂+t₃≤t₁+t₄When, if t₂≥t₄, t₁+t₂≥t₁+t₄If t₂< t₄, t₁+t₄=t₁+ t₄, therefore, TY >=TW；

As (t₂+t₃-t₁) > t₄, i.e. t₂+t₃> t₁+t₄When, if t₂≥t₄, t₁+t₂≥t₂+t₃If t₂< t₄, t₁+t₄≥t₂+ t₃, therefore, TY >=TW.

9. a kind of intelligent traffic lamp control method based on video monitoring as claimed in claim 7, which is characterized in that institute It states in step S43,

The complete transit time of vehicle for acquiring horizontal direction is TW₁, the transit time of vertical direction is TW₂, the total time of traffic lights For T, can distribute as follows the time：

Horizontal direction distribution the time be：

Vertical direction transit time is：

Work as TW₁+TW₂=0, without vehicle on present road, by default timing mode timing.