JPWO2013038561A1 - Lane judgment device, lane judgment method, and computer program for lane judgment - Google Patents

Abstract

The lane determination device is located at a distance from the radar and a distance from the radar arranged on the detection target lane with the longitudinal direction of the detection range overlapping the detection target lane of the plurality of parallel lanes. An interface unit that receives measurement data including at least one set of a reception level signal that represents the intensity of a radar wave reflected by an object, and a vehicle that travels in one of a plurality of lanes based on the plurality of measurement data Vehicle detection unit that generates tracking information including a distance from the radar to the position and a reception level signal at the position for each of a plurality of positions where the vehicle in the section where the vehicle is detected is detected And the vehicle detected based on the fluctuation in the intensity of the reception level signal in accordance with the change in the distance from the radar to the vehicle detected in the tracking information. And a determining lane determining unit whether running elephants lane.

Description

  The present invention relates to a lane determination apparatus, a lane determination method, and a lane determination computer program for determining a lane in which a vehicle is traveling using, for example, a radar installed so as to include a plurality of lanes in a detection range.

2. Description of the Related Art Conventionally, techniques for determining a lane in which a vehicle is traveling have been studied in order to accurately determine the inter-vehicle distance between the host vehicle and a preceding vehicle (see, for example, Patent Documents 1 to 4).
In the techniques disclosed in Patent Literatures 1 to 4, whether or not the detected object is a preceding vehicle traveling in the same lane as the vehicle is traveling based on a detection signal from a radar mounted on the vehicle. Is determined.

Japanese Patent Laid-Open No. 11-38139 JP 2001-124848 A JP 2004-170269 A JP 2005-69739 A

  Specifying the lane in which the vehicle is traveling can also be used to collect information related to traffic, such as the number of passing vehicles per unit time for each lane. However, in this case, it is preferable to use a radar that is fixedly installed on the road or on the side of the road, for example, in the vicinity of a specific intersection where the traffic volume information is to be acquired. However, any of the techniques disclosed in the above patent documents is based on the assumption that a radar is mounted on the vehicle. Therefore, even if these techniques are applied to determine the lane in which the vehicle is traveling using a fixedly installed radar, there is a possibility that sufficient determination accuracy cannot be obtained.

  Accordingly, an object of the present specification is to provide a lane determination device that can determine a lane in which a vehicle is traveling using a fixedly installed radar.

  According to one embodiment, a lane determination device is provided. This lane determination device is configured so that the longitudinal direction of the detection range overlaps with the detection target lane of a plurality of parallel lanes, and the distance from the radar and the distance from the radar disposed on the detection target lane An interface unit that receives measurement data that includes at least one set of a reception level signal that represents the intensity of a radar wave reflected by an object that travels, and travels in one of a plurality of lanes based on the plurality of measurement data Vehicle detection that detects a vehicle and creates tracking information including a distance from the radar to the position and a reception level signal at the position for each of a plurality of positions where the vehicle in the section where the vehicle is detected is detected. And the vehicle detected based on the fluctuation in the intensity of the reception level signal according to the change in the distance from the radar to the vehicle detected in the tracking information. And a determining lane determining section whether or not the traveling intellectual target lane.

The objects and advantages of the invention will be realized and attained by means of the elements and combinations particularly pointed out in the appended claims.
It should be understood that both the foregoing general description and the following detailed description are exemplary and explanatory and are not restrictive of the invention as claimed.

  The lane determination device disclosed in the present specification can determine the lane in which the vehicle is traveling using a fixedly installed radar.

FIG. 1 is a schematic configuration diagram of a lane determination device according to one embodiment. FIG. 2A is a diagram illustrating an example of a radar detection range on a detection target lane. FIG. 2B is a diagram illustrating an example of a radar detection range on a lane adjacent to the detection target lane. FIG. 2C is a diagram illustrating an example of a radar detection range on the detection target lane when occlusion occurs in the detection target lane. FIG. 3 is a diagram illustrating an example of a relationship between a distance from a radar to a vehicle in a tracking section and a reception level signal for a vehicle traveling in a detection target lane or an adjacent lane. FIG. 4 is a functional block diagram of the control unit. FIG. 5A is a graph showing an example of the relationship between the distance from the radar to the vehicle and the reception level signal included in the tracking information. FIG. 5B is a graph obtained by normalizing the graph shown in FIG. FIG. 6 shows an operation flowchart of the lane determination process.

Hereinafter, a lane determination device according to one embodiment will be described with reference to the drawings.
For example, the lane determination device detects a traveling vehicle on the basis of measurement data from a radar that is installed so as to overlap a part of each of a plurality of lanes having parallel detection ranges. And this lane determination device is based on the relationship between the tracking section in which the vehicle was detected and the fluctuation in the intensity of the reflected wave of the radar according to the change in the distance from the radar to the vehicle in the tracking section. It is determined whether or not the detected vehicle is traveling in a specific lane in the tracking section.

  FIG. 1 is a schematic configuration diagram of a lane determination device according to one embodiment. The lane determination device 1 includes a radar interface unit 11, a storage unit 12, an output unit 13, and a control unit 14. The lane determination device 1 is connected to a radar 2 for detecting a vehicle traveling on a road via a radar interface unit 11.

  In the present embodiment, the radar 2 is a radar device that detects an object reflecting a radar wave by a frequency modulated continuous wave (fmcw) method. The radar 2 is fixed on a specific lane of the plurality of lanes, for example, on a column provided with a traffic signal provided at an intersection or a column installed so as to straddle a plurality of parallel lanes. Installed. For convenience, the specific lane is hereinafter referred to as a detection target lane. Furthermore, the radar 2 is directed so as to face the front surface of the vehicle traveling toward the radar 2 and the longitudinal direction of the detection range of the radar 2 overlaps substantially parallel to the detection target lane.

  The radar 2 transmits a radar wave toward the front surface of the vehicle approaching the intersection on the road, and a radar transmission antenna (for which the direction is adjusted so as to receive the radar wave reflected by the front surface of the vehicle) And a receiving antenna (not shown).

The radar 2 mixes a part of a radar wave transmitted from the transmitting antenna and having a triangular wave-like frequency and a reflected wave detected by the receiving antenna. The radar 2 generates a beat signal that represents the difference between the frequency of the reflected wave and the frequency of the radar wave for each of the upstream section where the frequency is high and the downstream section where the frequency is low. Based on the beat signal, the radar 2 obtains the frequency f _up of the reflected wave in the _up section and the frequency f _down of the reflected wave in the down section, thereby determining the distance to the object reflecting the radar wave and the speed of the object. Ask. The radar 2 outputs measurement data at regular intervals (for example, 100 msec). The measurement data includes, for each of a plurality of positions set at predetermined distance intervals, a reception level signal indicating the intensity of the reflected wave, the distance from the radar 2 to the position, and the moving speed of the object at the position. Includes the measurement values to be paired. The predetermined distance interval corresponds to the resolution of the distance of the radar 2, and is set to an interval of 3m to 10m, for example.

  Radar waves radiated from the radar 2 spread as the distance from the radar 2 increases. Therefore, the reach range of the radar wave, that is, the detection range of the radar 2 includes not only a part of the detection target lane but also a part of the adjacent lane of the detection target lane. As a result, the radar 2 may detect a reflected wave from a vehicle traveling in the adjacent lane.

  The positional relationship between the detection range of the radar 2, the detection target lane, and the adjacent lane will be described with reference to FIGS. FIG. 2A is a diagram illustrating an example of a detection range of the radar 2 on the detection target lane.

  As shown in FIG. 2A, the radar 2 faces the detection target lane 200 above the detection target lane 200 in order to detect a vehicle traveling on the detection target lane 200, and It is installed to emit radar waves diagonally downward. A section where the detection target lane 200 and the detection range 210 of the radar 2 overlap is indicated by an arrow 220. In this example, the distance between the radar 2 and the boundary closer to the radar 2 in the section where the detection target lane 200 and the detection range 210 overlap is Dn, and the distance between the boundary farther from the radar 2 and the radar 2 in that section. Is represented by Df. Dn and Df vary depending on the type of radar 2 and the installation position, but for example, Dn = 20 m and Df = 160 m.

  FIG. 2B is a diagram illustrating an example of a detection range of the radar 2 on the adjacent lane. The range in which the radar waves radiated from the radar 2 reach in the width direction of the detection target lane 200 and the adjacent lane 201 gradually increases as the distance from the radar 2 increases. Therefore, the detection range 210 of the radar 2 also gradually increases as the distance from the radar 2 increases. For this reason, the distance Db between the radar 2 and the boundary farther from the radar 2 in the section where the adjacent lane 201 and the detection range 210 overlap indicated by the arrow 221 is substantially equal to the distance Df. On the other hand, the distance Da between the radar 2 side boundary and the radar 2 in the section where the adjacent lane 201 and the detection range 210 overlap is longer than the distance Dn. Therefore, the lane determination device 1 is detected by comparing the position of the end point on the radar 2 side of the tracking section for the traveling vehicle with the position of the boundary on the radar 2 side of the detection range of each lane. There is a possibility that the lane in which the vehicle is traveling can be determined. Hereinafter, for the sake of convenience, the end point on the radar 2 side of the tracking section is referred to as the near end of the tracking section, and the end point far from the radar 2 in the tracking section is referred to as the far end of the tracking section.

  However, if an object is present closer to the radar 2 than the vehicle, the radar wave radiated from the radar 2 is blocked by the object, so that there is a region where the radar 2 cannot detect the vehicle despite being within the detection range. Sometimes. Such a phenomenon is called occlusion. FIG. 2C is a diagram illustrating an example of a detection range of the radar 2 on the detection target lane when occlusion occurs.

  As shown in FIG. 2C, the radar wave is blocked by the vehicle 231 behind the vehicle 231 existing on the detection target lane 200. Therefore, the radar 2 cannot detect the reflected wave from the vehicle located within the range from the rear end of the vehicle 231 to the distance Do from the radar 2. As a result, the range in which the radar 2 can detect the vehicle in the detection target lane 200 is limited to the range of distances Do to Df from the radar 2 indicated by the arrow 222. The distance Do may be substantially equal to the distance Da between the radar 2 side boundary and the radar 2 in a section where the adjacent lane 201 and the detection range 210 overlap. Thus, when there is a possibility that an occlusion may occur due to an object existing on the detection target lane, the vehicle is traveling based only on the position of the near end of the tracking section for the traveling vehicle. The lane may not be accurately identified.

Next, the relationship between the distance from the radar 2 to the vehicle in the tracking section and the reception level signal for the vehicle traveling in the detection target lane or the adjacent lane will be described with reference to FIG.
In FIG. 3, the horizontal axis represents distance, and the vertical axis represents signal intensity. A graph 301 represents a change in the reception level signal according to a change in distance for a vehicle traveling in the detection target lane when no occlusion occurs. A graph 302 represents a change in the reception level signal according to a change in distance for a vehicle traveling in the detection target lane when occlusion occurs. Further, the graph 303 represents a change in the reception level signal corresponding to a change in distance for a vehicle traveling in the adjacent lane. The threshold Th represents a reception level signal value corresponding to the lower limit value of the intensity of the reflected wave by the vehicle. Therefore, a section in which the reception level signal is equal to or greater than the threshold value Th is a tracking section.

In any case, the reception level signal for the vehicle traveling toward the radar 2 gradually increases as the vehicle approaches the radar 2 and gradually decreases after reaching the peak of the reception level signal, that is, the maximum value. To do. However, the reception level signal for the vehicle traveling in the detection target lane in which the occlusion occurs is received for the vehicle traveling in the adjacent lane as the vehicle approaches the radar 2 from the position Sp at which the signal value reaches the peak. It drops more rapidly than the level signal.
Thus, the tendency of the intensity change of the reception level signal according to the distance change from the radar 2 to the vehicle is different between the vehicle traveling in the detection target lane where the occlusion occurs and the vehicle traveling in the adjacent lane.

Therefore, when the near end of the tracking section is farther than the boundary on the radar 2 side of the detection range of the adjacent lane, the lane determination device 1 changes the intensity of the reception level signal according to the distance from the radar 2 to the vehicle. Based on the degree, it is determined which lane the vehicle is traveling.
Hereinafter, the detail of each part of the lane determination apparatus 1 is demonstrated.

  The radar interface unit 11 has an interface circuit for connecting the lane determination device 1 to the radar 2. This interface circuit can be, for example, a circuit compliant with a serial communication standard such as RS-232C or universal serial bus, or a circuit compliant with Ethernet (registered trademark). The radar interface unit 11 passes the measurement data to the control unit 14 every time it receives measurement data from the radar 2.

  The storage unit 12 includes, for example, a readable / writable semiconductor memory circuit and a read-only semiconductor memory circuit. And the memory | storage part 12 is used in order to determine a lane, and memorize | stores the computer program which operate | moves on the control part 14. FIG. The storage unit 12 stores various data used for determining the lane, for example, measurement data received from the radar 2, vehicle tracking information detected based on the measurement data, and the like. Furthermore, the memory | storage part 12 memorize | stores the reference | standard graph showing the fluctuation | variation of the reception level signal according to the change of the distance about the vehicle which drive | works the lane for every lane. Details of the tracking information and the reference graph will be described later.

  The output unit 13 includes an interface circuit for connecting the lane determination device 1 to another device such as a traffic management system or a traffic light. This interface circuit can be a circuit that complies with, for example, Ethernet (registered trademark). And the output part 13 outputs the information which received the time which the vehicle received from the control part 14 and the lane which the vehicle drive | worked to other apparatuses. The time at which the vehicle is detected can be, for example, the time at which the lane determination device 1 acquires measurement data when the vehicle is first detected.

  The control unit 14 controls the lane determination device 1 as a whole. Further, the control unit 14 detects a vehicle traveling within the detection range of the radar 2 based on the measurement data received from the radar 2, and determines the lane in which the detected vehicle travels. For this purpose, the control unit 14 includes at least one processor, a timer, and a peripheral circuit.

  FIG. 4 is a functional block diagram of the control unit 14. The control unit 14 includes a vehicle detection unit 21 and a lane determination unit 22. Each of these units included in the control unit 14 is a functional module realized by a computer program executed on a processor included in the control unit 14. Alternatively, these units included in the control unit 14 may be mounted on the lane determination device 1 as separate circuits.

Each time the control unit 14 receives measurement data from the radar 2, the vehicle detection unit 21 detects a vehicle that is traveling in the detection range of the radar 2 based on the measurement data. And the vehicle detection part 21 calculates | requires the position and speed of the vehicle in the detection time for every detected vehicle.
For example, for each of a plurality of positions included in the measurement data, the vehicle detection unit 21 compares the reception level signal with a threshold value Th that is a reception level signal value corresponding to the lower limit value of the intensity of the reflected wave from the vehicle. And the vehicle detection part 21 determines with the vehicle existing in the position where the reception level signal becomes more than threshold value Th. And the vehicle detection part 21 makes the speed corresponding to the position where the vehicle was detected the speed of the detected vehicle. Note that the vehicle detection unit 21 may detect the vehicle using any of various other known techniques for detecting the vehicle based on the measurement data received from the radar.

The vehicle detection unit 21 tracks the detected vehicle. For this purpose, the vehicle detection unit 21 estimates the position of the vehicle at the time of obtaining the latest measurement data, for example, from the position of the vehicle detected at the time of acquisition of the past measurement data. For example, the vehicle detection unit 21 obtains the estimated travel distance of the vehicle by multiplying the detected speed of the vehicle by the elapsed time from the acquisition of the past measurement data to the acquisition of the latest measurement data. And the vehicle detection part 21 presumes the position of the vehicle at the time of the latest measurement data acquisition by subtracting the estimated traveling distance from the position of the vehicle at the time of the past measurement data acquisition. Alternatively, the vehicle detection unit 21 may estimate the position and speed of the detected latest vehicle measurement data based on past measurement data according to a vehicle behavior model (for example, a uniform acceleration motion model). Good. In this case, the vehicle detection unit 21 determines that the vehicle detected based on the latest measurement data is the same as the vehicle corresponding to the estimated position and estimated speed that is closest to the position and speed of the vehicle. . In addition, the vehicle detection part 21 calculated | required about the vehicle already detected within the predetermined range (for example, the range of +/- 10m centering on a detection position) from the position of the vehicle detected based on the newest measurement data. If there is no estimated position, it is determined that a new vehicle has been detected. And the vehicle detection part 21 allocates the identification number for distinguishing from the vehicle already detected to the newly detected vehicle.
In addition, the vehicle detection unit 21 determines that the reception level signal is a threshold value in the latest measurement data within a predetermined range including the estimated position of the vehicle when the latest measurement data is acquired (for example, a range of ± 10 m centered on the estimated position). If there is no position equal to or greater than Th, it is determined that the vehicle is out of the detection range.

  The vehicle detection unit 21 stores, in the storage unit 12, a set of the time of measurement data acquisition, the position of the vehicle, and the speed as tracking information together with the assigned identification number for each detected vehicle. While the vehicle is detected, a set of the time of measurement data acquisition, the position of the vehicle, and the speed obtained each time measurement data is acquired is added to the tracking information of the vehicle.

The lane determination unit 22 determines whether the lane in which the vehicle is traveling is the detection target lane or the adjacent lane based on the tracking information of each vehicle stored in the storage unit 12.
Therefore, first, the lane determination unit 22 identifies a vehicle that cannot be tracked, that is, a vehicle that is out of the detection range of the radar 2 as a vehicle of interest, and identifies a tracking section of the vehicle of interest based on tracking information of the vehicle of interest. For this purpose, the lane determination unit 22 includes the detection position of the vehicle farthest from the radar 2, that is, the far end S _max of the tracking section and the detection position of the vehicle closest to the radar 2 included in the tracking information of the vehicle of interest. Then, the near end S _min of the tracking section is obtained. When the vehicle approaches toward the front of the radar 2, the far end _Smax is a position where the reception level signal for the vehicle first becomes equal to or greater than the threshold Th, and the near end _Smin is The position at which the reception level signal for the vehicle has finally reached or exceeded the threshold Th. And the lane determination part 22 makes the area from the near end _Smin to the far end _Smax the tracking area about the vehicle.

Next, the lane determination unit 22 determines whether the near end S _min of the tracking section is closer to the radar 2 side boundary of the detection range in the adjacent lane. Note that the near end S _min of the tracking section varies depending on the size and shape of the vehicle traveling in the adjacent lane. This is because the radar cross section varies depending on the size and shape of the vehicle. For example, when the vehicle traveling in the adjacent lane is a large vehicle such as a truck or a bus, the radar reflection cross-section area of the vehicle is large, so that the near end S _min of the tracking section for the vehicle is relatively close. On the other hand, when the vehicle traveling in the adjacent lane is a small vehicle such as a motorcycle, the portion where the vehicle reflects the radar wave is small. It is farther away from the position outside the detection range. Therefore, the near end S _min of the tracking section for a small vehicle traveling in an adjacent lane is farther than the near end S _min for a large vehicle. Therefore, the distance Xm from the radar 2 corresponding to the boundary on the radar 2 side of the detection range in the adjacent lane shown in FIG. 3 is determined experimentally, for example. The distance Xm is set to, for example, a value obtained by subtracting a predetermined offset value (for example, several m to 20 m) from the average value of the near end S _min of the tracking section for a plurality of large vehicles traveling in adjacent lanes. Thereby, the lane determination unit 22 can suppress erroneous determination that the vehicle traveling in the adjacent lane is traveling in the detection target lane.

If the distance from the radar 2 to the near end S _min of the tracking section is shorter than the distance Xm, the vehicle is detected to a position closer to the radar 2 side boundary of the detection range in the adjacent lane, so the lane determination unit 22 The vehicle is determined to be traveling in the detection target lane.

In some cases, the vehicle cannot be detected at a position farther than the fluctuation range at the near end of the tracking section depending on the size and shape of the vehicle with respect to the boundary on the radar 2 side of the detection range in the adjacent lane. For example, when a preceding vehicle of a vehicle traveling in the detection target lane stops at a position relatively far from the radar 2, the far end of the region in which the radar 2 caused by the preceding vehicle cannot be detected is detected in the adjacent lane. It is farther from the radar 2 than the boundary of the range on the radar 2 side. In such a case, the radar 2 cannot detect a vehicle traveling in the detection target lane at such a distant position. On the other hand, in the adjacent lane, a radar wave hits the vehicle traveling on the adjacent lane from diagonally forward. Therefore, especially near the boundary of the detection range on the radar 2 side, the possibility that the radar wave is blocked by the preceding vehicle of the vehicle traveling in the adjacent lane is low. Therefore, the lane determination unit 22 may determine that the vehicle is traveling in the detection target lane even when the distance from the radar 2 to the near end S _min of the tracking section is longer than a certain distance Ym. In this case, the distance Ym is added by a predetermined offset value (for example, several m to 20 m), for example, than the average value of the near end S _min for a plurality of small vehicles traveling in the adjacent lane obtained experimentally. Set to a value. For example, when the detection range of the radar 2 is set to 20 m to 160 m, Xm = 50 m and Ym = 85 m are set.

On the other hand, when the distance from the radar 2 to the near end S _min of the tracking section is the distance Xm or more and Ym or less, the detected vehicle is traveling in the adjacent lane or on the detection target lane where the occlusion occurs. It is estimated that the vehicle is traveling. In this case, the lane determination unit 22 determines the lane in which the vehicle is traveling based on the fluctuation of the reception level signal according to the change in the distance for the vehicle.

The tracking interval and the reception level signal vary for each detected vehicle due to the external characteristics of the vehicle such as the shape and size of the detected vehicle or the environmental conditions such as the weather. Therefore, in the present embodiment, the lane determination unit 22 normalizes the relationship between the distance and the reception level signal in order to reduce the influence on the relationship between the distance and the reception level signal due to the external feature of the vehicle or the environmental condition.
The lane determination unit 22 has normalized the maximum value and the minimum value of the reception level signal to be a first predetermined value (for example, 1) and a second predetermined value (for example, 0) based on the tracking information, respectively. A signal value is obtained for each of a plurality of sample points that equally divide the tracking section into a plurality of sections. Thereby, the lane determination part 22 produces the normalization graph showing the intensity | strength fluctuation | variation of the reception level signal according to the distance change about the detected vehicle.

ここで、Nはサンプル点の総数である。またP_maxは、検知された車両についての受信レベル信号の最大値を表し、P_minは、検知された車両についての受信レベル信号の最小値を表す。なお、P_minの代わりに、閾値Thが用いられてもよい。またS_r(r=1,2,...,M、ただしMは、追跡情報に含まれる、車両が検知された測定点の総数)は、追跡情報に含まれる、レーダ２から車両までの距離を表し、P_rは、レーダ２から測定点までの距離S_rにおける受信レベル信号値を表す。B_k〜B_k+1は、サンプル点e_kについての正規化信号値を求めるために用いられる、レーダ２からの距離の範囲を表す。そして関数avg(P_r)(B_k≦S_r<B_k+1)は、距離B_k以上かつB_k+1未満の距離S_rについての受信レベル信号値P_rの平均値を算出する関数である。なお、関数avg(P_r)の代わりに、距離B_k以上かつB_k+1未満の距離S_rについての受信レベル信号値P_rの中央値を求める関数が用いられてもよい。For example, the lane determination unit 22 obtains a normalized signal value y _k of each sample point e _k (k = 1, 2,..., N) according to the following equation.

Here, N is the total number of sample points. P _max represents the maximum value of the reception level signal for the detected vehicle, and P _min represents the minimum value of the reception level signal for the detected vehicle. Note that a threshold Th may be used instead of P _min . S _r (r = 1,2, ..., M, where M is the total number of measurement points detected in the vehicle included in the tracking information) is included in the tracking information from the radar 2 to the vehicle. represents the distance, P _r represents the reception level signal value at the distance S _r from the radar 2 to the measurement point. B _{k to} B _{k + 1} represent a range of distance from the radar 2 that is used to obtain a normalized signal value for the sample point e _k . The function avg (P _r ) (B _k ≦ S _r <B _{k + 1} ) is a function for calculating the average value of the reception level signal values P _r for the distance S _r that is greater than or equal to the distance B _k and less than B _{k + 1.} It is. Instead of the function avg (P _r), the distance B _k or more and a function for obtaining a median value of the reception level signal value P _r for B _{k + 1} less than the distance S _r it may be used.

FIG. 5A is a graph showing an example of the relationship between the distance from the radar 2 to the vehicle and the reception level signal included in the tracking information. In FIG. 5A, the horizontal axis represents distance, and the vertical axis represents signal intensity. Each point 501 represents a reception level signal value at the distance S _r . In this example, N = 10. FIG. 5B is a graph obtained by normalizing the graph shown in FIG. In FIG. 5B, the horizontal axis represents the number of sample points, and the vertical axis represents the signal intensity. And each point 502, respectively, represent the normalized signal value at the sample point e _k. Each sample point e _k corresponds to distances B _{k to} B _{k + 1} in FIG. It can be seen that the shape of the normalized graph is similar to the shape of the original graph representing the degree of change in the reception level signal with respect to the distance change.

The lane determination unit 22 is a representative of the fluctuation in the intensity of the reception level signal according to the change in the distance from the radar 2 for the vehicle traveling in the detection target lane, corresponding to the case where the occlusion occurs in the normalization graph. obtaining a correlation coefficient C _o between the reference graph representing the Do relationship. Hereinafter, for convenience, this reference graph is referred to as a detection target lane reference graph. Similarly, the lane determination unit 22 obtains a correlation coefficient C _n between the normalized graph and a reference graph representing a typical relationship between fluctuations in the intensity of the reception level signal with respect to the distance for a vehicle traveling in the adjacent lane. . Hereinafter, for convenience, this reference graph is referred to as an adjacent lane reference graph. Moreover, the correlation coefficient C _o and C _n are each an example of the similarity representing a similarity degree normalized graphs and similar degree and normalization graph between a sense target lane based graph and an adjacent lane based graph.

The detection target lane reference graph, for example, obtains a normalization graph from tracking information about a plurality of vehicles traveling in the detection target lane when occlusion occurs, and normalization signal values in the plurality of normalization graphs Is averaged for each sample point. Similarly, for the adjacent lane reference graph, for example, a normalized graph is obtained from tracking information about a plurality of vehicles traveling in the adjacent lane, and the normalized signal values in the plurality of normalized graphs are averaged for each sample point. Is required. Moreover, the correlation coefficient C _o, in order to simplify the calculation of C _n, the number of sample points of the normalized graph is set to equal to the number of sample points of each reference graph.

ただし、y_avgは、正規化グラフに含まれる各サンプル点の正規化信号値y_k(k=1,2,...,N)の平均値である。またz_kは、検知対象車線基準グラフに含まれるサンプル点e_kの正規化信号値であり、z_avgは、検知対象車線基準グラフに含まれる各サンプル点の正規化信号値の平均値である。そしてw_kは、隣接車線基準グラフに含まれるサンプル点e_kの正規化信号値であり、w_avgは、隣接車線基準グラフに含まれる各サンプル点の正規化信号値の平均値である。Correlation coefficient C _o, C _n, for example, are calculated according to the following equation.

Here, y _avg is an average value of the normalized signal values y _k (k = 1, 2,..., N) of each sample point included in the normalization graph. Z _k is a normalized signal value of the sample points e _k included in the detection target lane reference graph, and z _avg is an average value of the normalized signal values of the respective sample points included in the detection target lane reference graph. . The w _k is the normalized signal value of the sample points e _k included in the adjacent lane based graph, w _avg is the average value of the normalized signal values for each sample point included in the adjacent lane based graph.

Lane determining unit 22, if the correlation coefficient C _n greater than the correlation coefficient C _o, i.e., the normalized graph, if similar to the adjacent lane based graph than the detection target lane based graph, is detected It is determined that the vehicle is traveling in an adjacent lane. On the other hand, if the correlation coefficient C _n is equal to or smaller than the correlation coefficient _Co , that is, if the normalized graph is more similar to the detection target lane reference graph than the adjacent lane reference graph, the lane determination unit 22 detects The determined vehicle is determined to be traveling in the detection target lane.

  The lane determination unit 22 stores, for example, the time at which the vehicle is detected and information related to the lane in which the vehicle travels in the storage unit 12 and outputs the information to other devices via the output unit 13.

FIG. 6 is an operation flowchart of the lane determination process executed by the control unit 14. Each time the control unit 14 acquires measurement data from the radar 2, the control unit 14 executes lane determination processing according to this operation flowchart.
The vehicle detection part 21 of the control part 14 detects a vehicle from measurement data (step S101). And the vehicle detection part 21 updates the tracking information for every vehicle using the position of the vehicle detected based on the newest measurement data, the position of the vehicle detected based on the past measurement data, etc.
Based on the tracking information, the control unit 14 determines whether there is a vehicle that cannot be detected among the vehicles being tracked (step S102).
When there is no vehicle that has become undetectable (step S102-No), the control unit 14 executes the process of step S101 again. On the other hand, when there is a vehicle that has become undetectable (step S102-Yes), the control unit 14 sets the undetectable vehicle as a target vehicle for lane determination. Then, the control unit 14 passes tracking information about the target vehicle for lane determination to the lane determination unit 22 of the control unit 14.

The lane determination unit 22 extracts the near end S _min of the tracking section from the tracking information (step S103). The lane determination unit 22 then determines whether the near end S _min of the tracking section is farther from the radar 2 than the radar 2 side boundary of the detection range of the radar 2 in the adjacent lane (step S104). When the near end _Smin of the tracking section is closer to the radar 2 side boundary of the detection range in the adjacent lane (step S104-No), the lane determination unit 22 determines that the vehicle is traveling in the detection target lane ( Step S108). The lane determination unit 22 outputs information indicating that the vehicle is detected and information indicating that the vehicle is traveling in the detection target lane to other devices via the output unit 13.

On the other hand, when the near end _Smin of the tracking section is farther from the radar 2 than the boundary of the detection range in the adjacent lane on the radar 2 side (step S104-Yes), the vehicle is traveling in the adjacent lane or the occlusion is It is estimated that the vehicle is traveling in the detection target lane. Therefore, the lane determination unit 22 creates a normalized graph representing the intensity fluctuation of the reception level signal according to the change in the distance in the tracking section (step S105). The lane determining section 22, the correlation coefficient normalized graph and detecting the lane based graph C _o, and calculates the correlation coefficient C _n of the normalized graph and an adjacent lane based graph (step S106).

Lane determining unit 22 determines whether the correlation coefficient C _n greater than the correlation coefficient C _o (step S107). If the correlation coefficient C _n is equal to or less than the correlation coefficient C _o (step S107-No), the lane determining unit 22 determines that the vehicle is running on a detection target lane (step S108). The lane determination unit 22 outputs the time when the vehicle is detected and information indicating that the vehicle is traveling in the detection target lane to other devices via the output unit 13.
On the other hand, if the correlation coefficient C _n greater than the correlation coefficient C _o (step S107-Yes), the vehicle is determined to be traveling on the adjacent lane (step S109). The lane determining unit 22 outputs information indicating that the vehicle is detected and information indicating that the vehicle is traveling in the adjacent lane to other devices via the output unit 13.
After step S108 or S109, the control unit 14 ends the lane determination process. In step S104, when the distance from the radar 2 to the near end S _min of the tracking section is longer than the distance Ym, it is estimated that the vehicle cannot be detected because occlusion occurs in the detection target lane. The lane determination unit 22 may determine that the detected vehicle is traveling in the detection target lane.

  As described above, this lane determination device is different in the tendency of the change in the reception level signal according to the distance change and the tracking section, which is different between the vehicle traveling in the detection target lane and the vehicle traveling in the adjacent lane. A lane in which the vehicle is traveling is determined based on the near end. Therefore, the lane determination device can accurately determine the lane in which the vehicle is traveling based on the vehicle detection signal from the fixedly installed radar.

In addition, this invention is not limited to said embodiment. According to the modification, the storage unit may also store a reference graph representing a relationship between a distance and a reception level signal for a vehicle traveling in the detection target lane when no occlusion occurs. In this case, the lane determination unit may also calculate a correlation coefficient C _s between the normalized graph obtained for the detected vehicle and the reference graph when no occlusion occurs in the detection target lane. Then, the lane determining unit may determine the lane corresponding to the reference graph having the highest value among the correlation coefficients C _s , _Co , and C _n as the lane in which the detected vehicle is traveling. In this case, the process of step S103 may be omitted.

In another modification, a lane in which it is known in advance that it is difficult for occlusion to occur may be a detection target lane. In this case, the lane determining unit, when the near-end S _min of the tracking section is far from the radar 2 than the boundary of the radar side of the detection range in the adjacent lane, the correlation coefficient C _o, without obtaining the C _n, is detected It may be determined that the vehicle is traveling in an adjacent lane.

  According to still another modified example, the vehicle detection unit is configured to measure each measurement point included in the tracking information for a certain period (for example, 5 to 10 seconds) even after the vehicle once detected becomes undetectable. The position of the vehicle may be estimated based on the speed of the vehicle and the measurement time. Then, at a certain point in time, the lane determination unit, when the position of the vehicle of interest is closer to the radar than the estimated position of the vehicle of the vehicle immediately before the vehicle of interest, that is, if the vehicle of interest overtakes the vehicle immediately before When estimated, the lane in which the vehicle of interest travels may be determined as a lane different from the lane in which the preceding vehicle travels.

According to still another modification, the lane determining unit obtains one or more feature amounts from the relationship between the detected distance and the reception level signal for the vehicle, and if the feature amount satisfies a predetermined condition, the vehicle May be determined to be traveling in the detection target lane. For example, referring to FIG. 3 again, for a vehicle traveling in the detection target lane when occlusion occurs, the ratio of the distance from the position where the reception level signal peaks to the near end of the tracking section with respect to the total length of the tracking section Is less than that ratio for vehicles traveling in adjacent lanes. Therefore, when the near end S _min of the tracking section is farther from the radar than the radar side boundary of the radar detection range in the adjacent lane, the lane determination unit determines the peak position of the reception level signal and the tracking section with respect to the entire length of the tracking section. The ratio of the distance to the near end S _min may be obtained as a feature amount. The lane determining unit may determine that the detected vehicle is traveling in the detection target lane if the ratio is equal to or less than the predetermined threshold Th2. The predetermined threshold Th2 is set to, for example, the maximum value of the ratio when occlusion occurs in the detection target lane, for example, 0.1 to 0.2.

  According to still another modification, the lane determination unit has an identifier that receives one or more feature values obtained from the relationship between the distance and the reception level signal and outputs a lane in which the vehicle is traveling. May be. In this case, the lane determining unit obtains one or more feature amounts from the relationship between the distance and the reception level signal included in the tracking information about the detected vehicle, and inputs the obtained feature amount to the discriminator. Determine the lane in which the vehicle is traveling. The feature amount can be, for example, the above-mentioned ratio or the maximum value of the absolute value of the slope of the change in the reception level signal with respect to the distance variation within the tracking section.

  The discriminator can be a machine learning system such as a multilayer perceptron or a support vector machine. In order to learn these machine learning systems, each of the feature amount obtained from the tracking information about the plurality of vehicles traveling in the detection target lane and the tracking information about the plurality of vehicles traveling in the adjacent lane, respectively. The obtained feature amount is used as learning sample data. The machine learning system learns to output, for example, the adjacent lane or the detection target lane according to the input feature amount by supervised learning such as back propagation using the learning sample data.

  According to still another modification, the radar may be a radar device according to a pulse compression method or a dual frequency continuous wave (CW) method. In this case, the control unit receives measurement data including a reception level signal for each of a plurality of positions set at predetermined distance intervals from the radar at regular intervals. And a vehicle detection part determines with a vehicle existing in the distance whose reception level signal is more than a predetermined threshold value. In this case, for example, the vehicle detection unit compares the current position of the vehicle detected in the latest measurement data with the past position of the vehicle detected in the measurement data acquired one time before. Then, the vehicle detection unit updates the tracking information for each vehicle by determining that the vehicle at the past position closest to the current position and the vehicle at the current position is far from the current position and the vehicle at the current position is the same vehicle. To do.

  A computer program having instructions for causing a computer to realize the functions of the control unit according to the above-described embodiment or modification is provided in a form recorded on a recording medium such as a magnetic recording medium, an optical recording medium, or a nonvolatile semiconductor memory. May be.

  All examples and specific terms listed herein are intended for instructional purposes to help the reader understand the concepts contributed by the inventor to the present invention and the promotion of the technology. It should be construed that it is not limited to the construction of any example herein, such specific examples and conditions, with respect to showing the superiority and inferiority of the present invention. Although embodiments of the present invention have been described in detail, it should be understood that various changes, substitutions and modifications can be made thereto without departing from the spirit and scope of the present invention.

DESCRIPTION OF SYMBOLS 1 Lane determination apparatus 2 Radar 11 Radar interface part 12 Memory | storage part 13 Output part 14 Control part 21 Vehicle detection part 22 Lane determination part

Claims (8)

The longitudinal direction of the detection range overlaps the detection target lane of the plurality of parallel lanes, and is reflected from the radar arranged on the detection target lane and the object located at the distance. An interface unit for receiving measurement data including at least one set of reception level signals representing the intensity of the radar wave;
Based on a plurality of the measurement data, a vehicle traveling in one of the plurality of lanes is detected, and the radar is detected for each of a plurality of positions where the vehicle is detected in a section where the vehicle is detected. A vehicle detection unit for creating tracking information including a distance from the position to the position and the reception level signal at the distance;
Whether or not the detected vehicle is traveling in the detection target lane based on a change in the intensity of the reception level signal corresponding to a change in the distance from the radar to the vehicle represented in the tracking information. A lane determination unit for determining;
A lane determination device having A first reference graph representing a change in intensity of the reception level signal in accordance with a change in a distance from the radar to a second vehicle traveling in the detection target lane, and the radar among the plurality of lanes A storage unit that stores a second reference graph representing a variation in the intensity of the reception level signal according to a change in the distance to the third vehicle traveling in the lane adjacent to the detection target lane;
The lane determining unit is configured to determine a similarity between the detected vehicle and the first reference graph indicating a variation in the intensity of the reception level signal according to a change in the distance from the radar to the vehicle. A first similarity degree to be expressed, and a second similarity degree representing a similarity degree between the graph and the second reference graph, and the first similarity degree is equal to or higher than the second similarity degree, The lane determination device according to claim 1, wherein the detected vehicle is determined to be traveling in the detection target lane.   The first reference graph is obtained when the radar cannot detect a part of the detection range on the detection target lane due to an object on the detection target lane located closer to the radar than the second vehicle. The lane determination device according to claim 2, wherein the lane determination device represents a variation in intensity of the reception level signal in accordance with a change in a distance from the radar to the second vehicle.   The lane determining unit is configured to detect the detected vehicle when a position closest to the radar in the section is closer to the radar than a boundary on the radar side of the detection range in a lane adjacent to the detection target lane. The lane determination device according to claim 1, wherein the lane determination device determines that the vehicle is traveling in the detection target lane.   The lane determining unit obtains at least one feature amount representing a variation in intensity of the reception level signal in accordance with a change in the distance from the radar to the detected vehicle in the section, and the feature amount is 2. The lane determination device according to claim 1, wherein, when a condition for a vehicle traveling in the detection target lane from a radar is satisfied, the detected vehicle is determined to be traveling in the detection target lane. The feature amount is a ratio of a distance from a position corresponding to the maximum value of the reception level signal to a position closest to the radar in the section with respect to the total length of the section,
The condition is when the radar cannot detect a part of the detection range on the detection target lane by an object on the detection target lane located closer to the radar than the detected vehicle. The lane determination device according to claim 5, wherein the lane determination device is equal to or less than a maximum value of the ratio. The longitudinal direction of the detection range overlaps the detection target lane of the plurality of parallel lanes, and is reflected from the radar arranged on the detection target lane and the object located at the distance. Receiving measurement data including at least one pair with a reception level signal representing the intensity of the radar wave,
Based on a plurality of the measurement data, a vehicle traveling in one of the plurality of lanes is detected, and the radar is detected for each of a plurality of positions where the vehicle is detected in a section where the vehicle is detected. Tracking information including the distance from the position to the position and the reception level signal at the distance,
Whether or not the detected vehicle is traveling in the detection target lane based on a change in the intensity of the reception level signal corresponding to a change in the distance from the radar to the vehicle represented in the tracking information. judge,
A lane determination method including the above. The longitudinal direction of the detection range overlaps the detection target lane of the plurality of parallel lanes, and is reflected from the radar arranged on the detection target lane and the object located at the distance. Detecting a vehicle traveling in one of the plurality of lanes based on a plurality of measurement data including at least one pair with a reception level signal representing the intensity of the radar wave,
For each of a plurality of positions where the vehicle is detected in the section where the vehicle is detected, create tracking information including the distance from the radar to the position and the reception level signal at the distance,
Whether or not the detected vehicle is traveling in the detection target lane based on a change in the intensity of the reception level signal corresponding to a change in the distance from the radar to the vehicle represented in the tracking information. judge,
A computer program for lane determination including instructions for causing a computer to execute the above.

Images