Information processing device, information processing method, and program

Abstract

To provide a technology capable of improving the accuracy of processing based on a route.SOLUTION: An information processing device is provided, including a calculation part for calculating an approximate formula for route information after rotation by each of a plurality of candidate rotation angles to the route information as a candidate approximate formula, and calculating a determination rotation angle and a determination approximate formula to be used for prescribed processing on the basis of the candidate approximate formula and the route information after the rotation, and a processing part for performing the prescribed processing on the basis of the determination rotation angle and the determination approximate formula.SELECTED DRAWING: Figure 2

Description

[Technical Field]

[0001] The present invention relates to an information processing device, an information processing method, and a program. [Background technology]

[0002] In recent years, techniques for performing processing based on the path of an object have become known. One possible technique for this is to calculate an approximate path by approximating the path to a straight line, and then perform processing based on the approximate path. However, paths are not necessarily straight. For example, paths may be curved, or may be a combination of straight and curved lines. Patent Document 1 discloses a technique for decomposing a path into multiple segments, such as straight line segments, arc segments, and clothoid curve segments, approximating each of the multiple segments using a different approximation method, calculating an approximate path, and then performing processing based on the approximate path. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent No. 5749359 Summary of the Invention [Problem to be solved by the invention]

[0004] However, simply decomposing a route into multiple parts and applying different approximation methods to each part may not improve the accuracy of the approximation.Furthermore, if the accuracy of the approximation does not improve, the accuracy of processing based on the approximate route may not improve either.

[0005] Therefore, the present invention has been made in consideration of the above problems, and an object of the present invention is to provide a technique that can improve the accuracy of processing based on a route. [Means for solving the problem]

[0006] In order to solve the above problem, according to one aspect of the present invention, an approximate expression for route information after rotation by each of a plurality of candidate rotation angles for the route information is calculated as a candidate approximate expression, and based on the candidate approximate expression and the route information after rotation, calculating an approximate value based on the first coordinate of the route information after the rotation and the candidate approximate formula for each candidate rotation angle, and determining the candidate rotation angle and the candidate approximate formula when an error between the approximate value and the second coordinate of the route information after the rotation is smallest; Decision rotation angle and The decision approximation formula is do A calculation unit that calculates The method includes a process of specifying a point cloud to be used for object detection based on the point cloud after rotation by the determined rotation angle for each of the plurality of route information included in the route information and the determined approximation formula. An information processing device is provided, comprising: a processing unit that performs a predetermined process.

[0007] The information processing device may include an object detection unit that detects an object based on a point cloud obtained by a sensor, and the processing unit may perform the specified processing based on the object position based on the detection result of the object, the determined rotation angle, and the determined approximation formula.

[0008] The predetermined processing may include a process of calculating a predicted distance that the object will move from the object position to the reference position based on the object position after rotation by the determined rotation angle, the reference position after rotation by the determined rotation angle, and the determined approximation formula.

[0009] The predetermined process may include a process of calculating, as a predicted arrival time, a time required for the object to travel the predicted distance based on the predicted distance and a moving speed of the object.

[0010] before The calculation unit may calculate the determination rotation angle and the determination approximation formula for each of the plurality of route information, and the processing unit may perform the specified processing based on the object position after rotation by the determination rotation angle for each of the plurality of route information and the determination approximation formula.

[0011] Each of the plurality of route information corresponds to a lane in which the object may be present, and the predetermined processing may include processing to determine the lane in which the object is present based on the object position after rotation by the determined rotation angle for each of the plurality of route information and the determined approximation formula.

[0012] The predetermined processing may include a process of acquiring a reference position in the lane in which the object is determined to exist, and a process of calculating a predicted distance that the object will move from the object position to the reference position based on the object position after rotation by the determined rotation angle in the lane in which the object is determined to exist, the reference position after rotation by the determined rotation angle, and the determination approximation formula.

[0013] before The information processing device may include an object detection unit that detects an object based on the point cloud.

[0014] The plurality of route information may include first route information and second route information, and a second coordinate of the first route information after rotation by the determined rotation angle for the same point is smaller than a second coordinate of the second route information after rotation by the determined rotation angle, and the predetermined processing may include processing for identifying, as a point cloud to be used for the object detection, a point cloud that satisfies the following conditions: the second coordinate of the first route information after rotation by the determined rotation angle is greater than or equal to a first approximation based on the first coordinate of the first route information after rotation by the determined rotation angle and the determined approximation formula of the first route information, or a value obtained by subtracting a first margin width from the first approximation; and the second coordinate of the second route information after rotation by the determined rotation angle is less than or equal to a second approximation based on the first coordinate of the second route information after rotation by the determined rotation angle and the determined approximation formula of the second route information, or a value obtained by adding a second margin width to the second approximation.

[0016] According to another aspect of the present invention, in order to solve the above problem, an approximation formula for route information after rotation by each of a plurality of candidate rotation angles for the route information is calculated as a candidate approximation formula, and based on the candidate approximation formula and the route information after rotation, calculating an approximate value based on the first coordinate of the route information after the rotation and the candidate approximate formula for each candidate rotation angle, and determining the candidate rotation angle and the candidate approximate formula when an error between the approximate value and the second coordinate of the route information after the rotation is smallest; Decision rotation angle and The decision approximation formula is do Calculating and The method includes a process of specifying a point cloud to be used for object detection based on the point cloud after rotation by the determined rotation angle for each of the plurality of route information included in the route information and the determined approximation formula. and performing a predetermined process.

[0017] According to another aspect of the present invention for solving the above-described problems, a computer calculates, as candidate approximate expressions, approximate expressions for route information after rotation by each of a plurality of candidate rotation angles for the route information, and performs the following calculation based on the candidate approximate expressions and the route information after rotation: calculating an approximate value based on the first coordinate of the route information after the rotation and the candidate approximate formula for each candidate rotation angle, and determining the candidate rotation angle and the candidate approximate formula when an error between the approximate value and the second coordinate of the route information after the rotation is smallest; Decision rotation angle and The decision approximation formula is do A calculation unit that calculates The method includes a process of specifying a point cloud to be used for object detection based on the point cloud after rotation by the determined rotation angle for each of the plurality of route information included in the route information and the determined approximation formula. A processing unit that performs predetermined processing and a program that functions as such are provided. [Effects of the Invention]

[0018] As described above, the present invention provides a technique that can improve the accuracy of route-based processing. [Brief explanation of the drawings]

[0019] [Figure 1] FIG. 10 is a diagram for explaining an example of an approximated route obtained by approximating a route to a curve expressed by a 100th-order polynomial. [Figure 2] FIG. 1 is a diagram illustrating a situation in which an information processing system according to a first embodiment of the present invention is applied. [Figure 3] 1 is a block diagram illustrating an example of a functional configuration of an information processing system according to a first embodiment of the present invention. [Figure 4] 10 is a flowchart showing an example of the operation of the information processing device 10-1 in the advance preparation stage. [Figure 5] 10 is a diagram illustrating an example of a correspondence relationship between the order k, the maximum error maxErrorkθ, and the determined rotation angle θk_min. FIG. [Figure 6] FIG. 10 is a diagram for explaining an example of an approximated path obtained by approximating a path after rotation to a curve expressed by a sixth-order polynomial. [Figure 7] 10 is a flowchart showing an example of the operation of the information processing device 10-1 in the operation stage. [Figure 8] 10A and 10B are diagrams illustrating examples of route information before and after rotation at a determined rotation angle. [Figure 9]10 is a diagram showing an example of a point group Pc'(x', y') before being identified by the point group identification unit 230. FIG. [Figure 10] 10 is a diagram showing an example of a point group Pc''(x'', y'') identified by the point group identification unit 230. FIG. [Figure 11] FIG. 4 is a diagram showing processing times of the point cloud identification process according to the first embodiment of the present invention and the point cloud identification process according to a comparative example. [Figure 12] FIG. 10 is a diagram illustrating a situation in which an information processing system according to a second embodiment of the present invention is applied. [Figure 13] FIG. 10 is a block diagram illustrating an example of a functional configuration of an information processing system according to a second embodiment of the present invention. [Figure 14] 10 is a flowchart showing an example of the operation of the information processing device 10-2 in the advance preparation stage. [Figure 15] 10 is a flowchart showing an example of the operation of the information processing device 10-2 in the operation stage. [Figure 16] 10 is a diagram for explaining an example of determining a driving lane n when the number of lanes N is 2. FIG. [Figure 17] 1 is a diagram showing a hardware configuration of an information processing device 900 as an example of the information processing device 10 according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0020] Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. In this specification and drawings, components having substantially the same functional configurations are designated by the same reference numerals, and redundant explanations will be omitted.

[0021] <0.Background> First, the background of the embodiment of the present invention will be described.

[0022] In recent years, a technology for performing processing based on the path of an object has become known. As described above, simply performing approximation using different approximation methods for each of the multiple parts obtained by decomposing the path may not improve the accuracy of the approximation. Furthermore, if the accuracy of the approximation does not improve, the accuracy of processing based on the approximate path may not improve either.

[0023] It is also possible to calculate an approximate route by approximating the route to a curve expressed by a polynomial with a maximum degree of 2 or more, without decomposing the route, and then execute processing based on the approximate route. As an example, with reference to Fig. 1, an example of an approximate route by approximating the route to a curve expressed by a 100th-order polynomial with a maximum degree of 100 will be described.

[0024] FIG. 1 is a diagram illustrating an example of an approximated route in which a route is approximated to a curve expressed by a 100th-order polynomial. In the example shown in FIG. 1, "route information" is shown on an xy plane with the horizontal axis as the x-axis and the vertical axis as the y-axis. The route information includes multiple points that the route passes through. Hereinafter, each of the multiple points included in the route information will also be referred to as a "route element." In addition, an approximated route in which a route is approximated to a curve expressed by a 100th-order polynomial is shown as an "approximation result using a 100th-order polynomial."

[0025] Referring to FIG. 1, it can be seen that there are some parts of the "approximation result using a 100th-order polynomial" that do not fit the "route information." This is mainly due to the difficulty of approximating route information in which there are multiple outputs y for one input x to a polynomial. In other words, it can be seen that simply approximating a route to a curve represented by a 100th-order polynomial does not improve the accuracy of the approximation. Therefore, this specification mainly proposes a technology that can improve the accuracy of processing based on an approximate route by improving the accuracy of the approximation of the route.

[0026] The background of the embodiments of the present invention has been described above.

[0027] <1. First embodiment> Next, a first embodiment of the present invention will be described.

[0028] [1-1. Application of information processing systems] First, with reference to FIG. 2, a situation in which the information processing system according to the first embodiment of the present invention is applied will be described.

[0029] Fig. 2 is a diagram for explaining a situation in which the information processing system according to the first embodiment of the present invention is applied. Referring to Fig. 2, there is shown an information processing device 10-1 and a laser sensor 200 included in the information processing system according to the first embodiment of the present invention. The laser sensor 200 can also be referred to as a LiDAR (Light Detection and Ranging) sensor.

[0030] Further referring to FIG. 2, a vehicle C1 traveling in lane L1 and a vehicle C2 traveling in lane L2 are shown. The traveling direction of vehicle C1 and the traveling direction of vehicle C2 are the same. In the first embodiment of the present invention, lanes L1 and L2, which are traveling in the same direction, are the main target lane area D1. In the example shown in FIG. 2, the number of lanes in the target lane area D1 is two, lane L1 and lane L2, but the number of lanes in the target lane area D1 is not limited to two, and may be one lane, or three or more lanes.

[0031] On the other hand, the opposite lane area D2 is an area where vehicles travel in the opposite direction to the direction of travel of vehicles C1 and C2. The number of lanes in the opposite lane area D2 is not particularly limited, and it may be one lane or two or more lanes. The target lane area D1 and the opposite lane area D2 correspond to the public road (hereinafter also referred to as the "main road") N1.

[0032] Autonomous vehicle C3 is traveling on private road N2 and is located just before exiting private road N2 onto main road N1. Note that a manually driven vehicle may be traveling on private road N2 instead of autonomous vehicle C3.

[0033] In recent years, the realization of self-driving cars is expected to solve social issues such as reducing traffic accidents and alleviating labor shortages. As self-driving cars become more widespread, there is a demand for infrastructure-cooperative ITS (Intelligent Transport System) services.

[0034] In particular, when an autonomous vehicle (for example, autonomous vehicle C3 shown in Figure 2) makes a right or left turn when exiting private road N2 onto main road N1, it is difficult for the autonomous vehicle to grasp traffic information about main road N1. As a result, there is a risk that the autonomous vehicle will disrupt traffic flow by forcibly merging from private road N2 onto main road N1. There is also a risk that the autonomous vehicle will disrupt traffic flow when turning right or left at an intersection. Due to these circumstances, there is a strong demand for the provision of such a service.

[0035] 2, if laser sensor 200 is installed as an infrastructure sensor on the roadside of main line N1, information processing device 10-1 can extract traffic information for main line N1 (such as information about vehicle C1 traveling in lane L1 and vehicle C2 traveling in lane L2) from data obtained by laser sensor 200. Information processing device 10-1 can then notify automatically driven vehicle C3 of the traffic information for main line N1 via wireless communication or the like. This allows automatically driven vehicle C3 to merge onto main line N1 safely and smoothly.

[0036] Here, depending on the angle of view of the laser sensor 200, data on an area outside the target lane area D1 (for example, the opposite lane area D2) may also be acquired. In such a case, although vehicles should be detected only from the target lane area D1, vehicles may also be detected from the area outside the target lane area D1, which may result in a situation where vehicles are erroneously detected and processing costs are incurred for unnecessary object detection.

[0037] According to a first embodiment of the present invention, which will be described below, even if the main road N1 is a curved road, only the vehicles C1 and C2 traveling in the target lane area D1 can be detected with high accuracy. Furthermore, according to the first embodiment of the present invention, the possibility of incurring processing costs for unnecessary object detection can be reduced. As a result, even an edge PC (Personal Computer) with low processing performance can detect only the vehicles C1 and C2 traveling in the target lane area D1.

[0038] 2, route information R is shown. The route information R includes multiple route elements. However, since assigning a symbol to each of the multiple routes would impair the readability of the diagram, the symbol P is assigned to one of the multiple route elements. In the first embodiment of the present invention, the route information R is located at the boundary between the inside and outside of the target lane area D1.

[0039] The boundaries are located at both ends in the width direction of the target lane area D1. In the following description, of these boundaries, the route information R located at the boundary on the left side of lane L1 in the direction from upstream to downstream of the target lane area D1 (hereinafter also referred to as the "length direction") will be simply referred to as "route information R on the lane L1 side." Also, the route information R located at the boundary on the right side of lane L2 in the length direction will be simply referred to as "route information R on the lane L2 side." The route information R will be described in more detail later.

[0040] Furthermore, vehicles C1 and C2 are examples of objects detected by laser sensor 200. Therefore, instead of vehicles C1 and C2, other objects may be detected by laser sensor 200. For example, the object detected by laser sensor 200 may be a person traveling on a train or an escalator.

[0041] In this case, the term "vehicle movement" that appears in this specification may be appropriately replaced with the term "movement of an object," and the term "lane" that appears in this specification may be appropriately replaced with the term "lane."

[0042] The above has described the situations in which the information processing system according to the first embodiment of the present invention is applied.

[0043] [1-2. Example of information processing system configuration] Next, an example of the functional configuration of the information processing system according to the first embodiment of the present invention will be described. Fig. 3 is a block diagram showing an example of the functional configuration of the information processing system according to the first embodiment of the present invention. As shown in Fig. 3, the information processing system according to the first embodiment of the present invention includes an information processing device 10-1, a laser sensor 200, a notification device 280, and a display device 290. The information processing device 10-1, the laser sensor 200, the notification device 280, and the display device 290 may be connected via a network.

[0044] (Laser Sensor 200) The laser sensor 200 is installed at a position and orientation such that the vehicles C1 and C2 traveling in the target lane area D1 fall within the measurement range of the laser sensor 200. Note that a sensor other than the laser sensor 200 (such as a distance measurement sensor) may be used instead of the laser sensor 200. For example, a radar (such as a millimeter wave radar) that uses electromagnetic waves instead of laser light may be used as the sensor. Alternatively, a sonar that uses sound waves instead of laser light may be used as the sensor.

[0045] (Information processing device 10-1) 3, the information processing device 10-1 includes a route information storage unit 100, a reference coordinate system information storage unit 110, a coordinate system conversion unit 120, an approximate route calculation unit 130, an approximate route information storage unit 140, and a route information output unit 160.

[0046] Furthermore, the information processing device 10-1 includes a point cloud acquisition unit 210, a sensor coordinate system conversion unit 220, a point cloud identification unit 230, an object detection unit 250, a past frame detected object information storage unit 240, an object tracking unit 260, and an object information calculation unit 270. Details of these components included in the information processing device 10-1 will be described later.

[0047] For example, the coordinate system conversion unit 120, the approximate route calculation unit 130, the route information output unit 160, the point cloud acquisition unit 210, the sensor coordinate system conversion unit 220, the point cloud identification unit 230, the past frame detected object information storage unit 240, the object detection unit 250, the object tracking unit 260, and the object information calculation unit 270 may be realized by a control unit (not shown). On the other hand, the route information storage unit 100, the reference coordinate system information storage unit 110, and the approximate route information storage unit 140 may be realized by a storage unit (not shown).

[0048] The control unit (not shown) includes a CPU (Central Processing Unit) and the like, and its functions can be realized by the CPU expanding a program stored in a non-volatile storage device into RAM (Random Access Memory) and executing it. In this case, a computer-readable recording medium on which the program is recorded can also be provided. Alternatively, the control unit (not shown) can be configured with dedicated hardware or a combination of multiple pieces of hardware.

[0049] The storage unit (not shown) is a storage device capable of storing programs and data for operating the control unit (not shown). The storage unit (not shown) can also temporarily store various data required in the operation of the control unit (not shown). For example, the storage device may be a non-volatile storage device.

[0050] (Notification device 280) In accordance with the control of information processing device 10-1, notification device 280 notifies information output from information processing device 10-1 to notification device 280. For example, notification device 280 may be configured with a display, and notify information by displaying the information output from information processing device 10-1 to notification device 280 on the display.

[0051] There is no particular limitation on the location where the notification device 280 is installed. For example, the notification device 280 may be installed in the automatically driven vehicle C3.

[0052] (Display device 290) Under the control of information processing device 10-1, display device 290 displays information output from information processing device 10-1 to display device 290. For example, display device 290 may be configured with a display, and may display information output from information processing device 10-1 to display device 290 on the display.

[0053] There is no particular limitation on the location where display device 290 is installed. For example, display device 290 may be installed in a monitoring room where an observer who monitors the running conditions of the vehicle is present.

[0054] An example of the functional configuration of the information processing system according to the first embodiment of the present invention has been described above.

[0055] [1-3. Example of operation of information processing device] Next, an example of operation of the information processing device 10-1 according to the first embodiment of the present invention will be described. The example of operation of the information processing device 10-1 is mainly divided into two stages. The first stage is a preparation stage. The second stage is an operation stage. Hereinafter, an example of operation of the information processing device 10-1 relating to the preparation stage will be described with reference to Figs. 4 to 6 (and Figs. 1 to 3 as well, as appropriate).

[0056] (Preparation stage) 4 is a flowchart showing an example of the operation of the information processing device 10-1 in the advance preparation stage. As shown in FIG. 4, in the advance preparation stage, the information processing device 10-1 performs step A12 and step A13 for the left and right ends of the target lane area D1 (i.e., for the route information R on the lane L1 side and the route information R on the lane L2 side, respectively) (step A11).

[0057] Here, route information R is registered in advance in the route information storage unit 100. The route information R includes the positions of a plurality of route elements P.

[0058] For example, a mobile body equipped with an RTK (Real Time Kinematic) mobile station may detect its own position while moving along the boundary between the inside and outside of a target lane area D1 and register it as the position of each of multiple route elements P. For example, if a white line is drawn on the boundary between the inside and outside of the target lane area D1, the mobile body equipped with an RTK mobile station may move along the white line.

[0059] Alternatively, the positions on the boundary between the inside and outside of the target lane area D1 obtained from map information of the Geospatial Information Authority of Japan may be registered as the positions of each of the multiple route elements P.

[0060] The route information R is expressed by positions in a coordinate system before conversion in the laser sensor 200 (hereinafter also referred to as the "original coordinate system"). However, a coordinate system suitable for processing the point cloud obtained by the laser sensor 200 may be a coordinate system (hereinafter also referred to as the "reference coordinate system") different from the original coordinate system. Therefore, the coordinate system conversion unit 120 converts the coordinate system representing the route information R from the original coordinate system to the reference coordinate system (step A12).

[0061] Note that, in cases where processing of the point group can be performed appropriately without converting the coordinate system, step A12 may be omitted.

[0062] For example, the reference coordinate system may be a coordinate system in which the position of the laser sensor 200 is the origin, the direction from upstream to downstream of the target lane area D1 in a predetermined section (i.e., the length direction) is the y-axis, and the direction from the left end of lane L1 to the right end of lane L2 in the length direction of the predetermined section (hereinafter also referred to as the "width direction") is the x-axis. The predetermined section does not need to be particularly limited. As an example, the predetermined section may be a section located close to the position of the laser sensor 200.

[0063] More specifically, the position of the laser sensor 200 in the original coordinate system is (O x ,O y ), and the position of the path element P in the original coordinate system is (P x ,P y ), and the longitudinal direction of the target lane area D1 in a predetermined section is shifted by an angle α from the y-axis in the original coordinate system. In this case, the route information storage unit 100 stores the position of the route element P in the original coordinate system (P x ,P y ) is stored in the reference coordinate system information storage unit 110, and the position (O x ,O y ) and the angle α are stored.

[0064] The coordinate system conversion unit 120 receives the position of the route element P in the original coordinate system (P x ,P y ) and obtains the position (O x ,O y ) and angle α. Then, the coordinate system conversion unit 120 obtains the position (P x ,P y ) and the position of the laser sensor 200 (O x ,O y ) and the angle α, the position P' (P' x ,P' y ) can be calculated as follows:

[0065]

number

[0066] In this way, the position P' (P' x ,P' y The approximate path calculation unit 130 calculates the position P' (P' x ,P' y ) is calculated as an approximate route (step A13).

[0067] More specifically, the approximate path calculation unit 130 calculates the position P' (P' x ,P' y ) is rotated by each of the plurality of candidate rotation angles θ, and a plurality of route elements P'' after rotation for each candidate rotation angle θ are obtained. θ Then, the approximate path calculation unit 130 calculates the multiple post-rotation path elements P''. θ The approximate formula for the candidate approximate formula y=f kθ (x) where k indicates the highest degree of the candidate approximation formula.

[0068] Multiple path elements P'' after rotation θ The approximation to may be a polynomial approximation (hereinafter also referred to as "polyfit"). In this specification, the term "polynomial" is not limited to a polynomial with multiple terms, but may also include a polynomial with one term. In this specification, the polynomial may be an expression with a maximum degree of zero or one or more dimensions. That is, the candidate approximate expression y=f kθ (x) can be expressed using k coefficients (where k is an integer greater than or equal to 0) and one constant. For example, f kθ (x) can be calculated as shown in equation (2) below.

[0069]

number

[0070] Here, the multiple candidate rotation angles θ may be determined in any manner. For example, the multiple candidate rotation angles θ may be angles that vary in increments of Δθ (for example, 1 degree) within the range of −0≦θ<180 degrees.

[0071] The approximate path calculation unit 130 calculates a candidate approximate equation y=f for each candidate rotation angle θ. kθ (x) and multiple path elements P'' after rotation θ Based on the above, a rotation angle to be used in the predetermined process is calculated as a determined rotation angle, and an approximation formula to be used in the predetermined process is calculated as a determined approximation formula.

[0072] In the first embodiment of the present invention, the predetermined processing is processing in which the point cloud identification unit 230 identifies a point cloud to be used for object detection. That is, the point cloud identification unit 230 corresponds to an example of a processing unit that performs the predetermined processing based on the determined rotation angle and the determined approximation formula.

[0073] For example, the approximate path calculation unit 130 calculates the plurality of post-rotation path elements P''. θ P'' is the x-coordinate (first coordinate) of θx and the candidate approximation equation y=f kθ (x) and the approximate value f based on kθ (P'' θx ) may be calculated for each candidate rotation angle θ.

[0074] Then, the approximate path calculation unit 130 calculates the approximate value f kθ (P'' θx ) and multiple path elements P'' after rotation θ P'' is the y-coordinate (second coordinate) of θy The candidate rotation angle θ and candidate approximate equation y=f when the error with kθ (x) is determined by the rotation angle θ k_min and the determined approximation formula y=f kθ_min It may also be calculated as (x).

[0075] At this time, f kθ (P'' θx ) and P'' θy The error between and is the maximum error maxError kθAlso, the maximum error maxError kθ is f kθ (P'' θx ) and P'' θy For example, the approximate path calculation unit 130 may calculate the determined rotation angle θ k_min can be calculated as shown in equation (3) below:

[0076]

number

[0077] In the above formula (3), argmin is the maximum error maxError kθ The maximum error is maxError kθ is a function that outputs the candidate rotation angle θ when the is minimum. In other words, the candidate rotation angle θ output from argmin is the same as the determined rotation angle θ k_min The candidate approximate equation y=f corresponding to the candidate rotation angle θ is calculated as follows: kθ (x) is determined by the approximate equation y=f kθ_min It is calculated as (x).

[0078] The approximate formula for determination is y=f kθ_min The higher the degree k of (x), the more easily the approximate equation y=f kθ_min It is considered that the processing accuracy improves as the processing load in the subsequent stage using (x) increases. Therefore, the approximate path calculation unit 130 does not decide on one order k from the beginning, but calculates the maxError kθ Calculate maxError kθ The order k may be determined based on:

[0079] Figure 5 shows the relationship between the order k and the maximum error maxError kθ and the determined rotation angle θ k_min In the example shown in FIG. 5, even if the degree k is lowered to some extent, the maximum error maxError kθ is kept small.

[0080] For example, the approximate path calculation unit 130 calculates the order k and the maximum error maxError kθ (i.e., taking into consideration the processing accuracy and the processing load). kθ The lowest degree k may be determined from among the degrees k for which k is lower than the threshold.

[0081] The determined rotation angle θ calculated in this way k_min and the determined approximation formula y=f kθ_min (x) may correspond to an approximate route. That is, according to the first embodiment of the present invention, an approximate route based on route information can be calculated with high accuracy. The approximate route calculation unit 130 calculates the determined rotation angle θ k_min and the determined approximation equation y=f kθ_min The approximate route information storage unit 140 stores k coefficients and one constant that express (x).

[0082] Fig. 6 is a diagram for explaining an example of an approximated path obtained by approximating a route after rotation to a curve expressed by a sixth-order polynomial. In the example shown in Fig. 6, an approximated path obtained by approximating a route after 45 degrees rotation to a curve expressed by a sixth-order polynomial is shown as "approximation result using a sixth-order polynomial." Referring to Fig. 6, it can be seen that by rotating the route, an approximated path with a small error is calculated for a curve expressed by a low-order polynomial.

[0083] The route information output unit 160 determines the determined rotation angle θ k_min and the determined approximation equation y=f kθ_min An approximate route based on k coefficients and one constant that express (x) may be output to the display device 290. The display device 290 may then display the approximate route. This allows the observer to visually confirm the approximate route calculated by the approximate route calculation unit 130.

[0084] An example of the operation of the information processing device 10-1 in the advance preparation stage has been described above with reference to FIGS.

[0085] (Operational stage) Next, an example of the operation of the information processing device 10-1 in the operation stage will be described with reference to FIGS.

[0086] 7 is a flowchart showing an example of the operation of the information processing device 10-1 in the operation stage. As shown in Fig. 7, in the operation stage, the point cloud acquisition unit 210 acquires the point cloud obtained by the laser sensor 200 from the laser sensor 200 (step B11).

[0087] Next, the sensor coordinate system conversion unit 220 converts the coordinate system representing the point cloud from the original coordinate system to the reference coordinate system (step B12). Note that, similar to step A12 (FIG. 4), step B12 may be omitted if the processing of the point cloud can be performed appropriately without converting the coordinate system.

[0088] More specifically, the sensor coordinate system conversion unit 220 converts the position (O x ,O y ) and angle α. Then, the sensor coordinate system conversion unit 220 obtains (P x ,P y ) is replaced with the position of the point cloud, the point cloud and the position (O x ,O y ) and the angle α, a point group Pc(x, y, z) in the reference coordinate system can be calculated.

[0089] The point cloud identification unit 230 identifies a point cloud to be used for object detection (step B13). An example will be described below in which the point cloud identification unit 230 identifies the point cloud to be used for object detection in the z-axis direction and in the xy plane (i.e., the direction perpendicular to the z-axis). Note that the process of identifying the point cloud to be used for object detection in the z-axis direction is a process performed to speed up the process of identifying the point cloud to be used for object detection in the xy plane, and therefore may be omitted as appropriate.

[0090] First, the minimum value z_min and maximum value z_max of the z coordinate of the position where the object to be detected (vehicles C1 and C2 in the example shown in FIG. 2) may exist are set in advance.

[0091] Based on the point group Pc(x, y, z) in the reference coordinate system, the point group identification unit 230 identifies a point group Pc' from the point group Pc whose z coordinate is equal to or greater than z_min and equal to or less than z_max. That is, when the z coordinate of the point group Pc in the reference coordinate system is defined as pc_z, the point group identification unit 230 identifies a point group Pc' that satisfies (pc_z≧z_min)&&(pc_z≦z_max). Hereinafter, focusing on the x and y coordinates of the point group Pc', the point group Pc' will be expressed as Pc'(x', y').

[0092] In the following description, the position in the reference coordinate system of the route information R on the lane L1 side will also be referred to as route information R1' in the reference coordinate system. Also, the position in the reference coordinate system of the route information R on the lane L2 side will also be referred to as route information R2' in the reference coordinate system. The route information R1' in the reference coordinate system corresponds to an example of first route information. The route information R2' in the reference coordinate system corresponds to an example of second route information.

[0093] Also, the determined rotation angle θ corresponding to the route information R1′ in the reference coordinate system k_min is also expressed as θ1, and the determined rotation angle θ k_min is also expressed as θ2. The approximate formula y=f corresponding to the route information R1' in the reference coordinate system is kθ_min (x) is also expressed as y=f1(x), and the determined rotation angle y=f kθ_min (x) can also be written as y=f2(x).

[0094] The point cloud identification unit 230 acquires a determined rotation angle θ1 corresponding to the route information R1' in the reference coordinate system, k coefficients expressing the determined approximate equation y = f1(x), and one constant from the approximate route information storage unit 140. Furthermore, the point cloud identification unit 230 acquires a determined rotation angle θ2 corresponding to the route information R2' in the reference coordinate system, k coefficients expressing the determined approximate equation y = f2(x), and one constant from the approximate route information storage unit 140.

[0095] Fig. 8 is a diagram showing an example of route information before and after rotation by a determined rotation angle. The graph shown on the left of Fig. 8 shows route information R1' and route information R2' in the reference coordinate system. The graph on the right of Fig. 8 shows the determined approximate formula y = f1(x) after rotating route information R1' in the reference coordinate system by a determined rotation angle θ1, and the determined approximate formula y = f2(x) after rotating route information R2' in the reference coordinate system by a determined rotation angle θ2.

[0096] Based on the determined rotation angle θ1, the path information R1' in the reference coordinate system, and the determined rotation angle θ2, and the path information R2' in the reference coordinate system, the point cloud identification unit 230 determines the approximate equation that changes the larger y coordinate position between y=f1(x) and y=f2(x) as f_upper(x), and determines the approximate equation that changes the smaller y coordinate position as f_lower(x).

[0097] Various methods can be adopted as a detailed method for determining f_upper(x) and f_lower(x). As one example, the point cloud identification unit 230 may input the x value of a point Pc'(x',y') included in the point cloud Pc' after rotating it by θ1 as x1 into y=f1(x), and obtain the y value that is the output as y1. Similarly, the point cloud identification unit 230 may input the x value of the point Pc'(x',y') after rotating it by θ2 as x2 into y=f2(x), and obtain the y value that is the output as y2.

[0098] Then, when y1 is smaller than y2, the point cloud identification unit 230 determines y=f1(x) as y=f_lower(x), θ1 as θ_lower, y=f2(x) as y=f_upper(x), and θ2 as θ_upper.On the other hand, when y2 is smaller than y1, the point cloud identification unit 230 determines y=f2(x) as y=f_lower(x), θ2 as θ_lower, y=f1(x) as y=f_upper(x), and θ1 as θ_upper.

[0099] The point cloud identification unit 230 calculates the point cloud (x'_rot_lower, y'_rot_lower) after rotating each point (x', y') included in the point cloud Pc' by θ_lower. Furthermore, the point cloud identification unit 230 calculates the point cloud (x'_rot_upper, y'_rot_upper) after rotating each point (x', y') included in the point cloud Pc' by θ_upper. That is, the point cloud after rotation can be calculated as shown in the following formula (4).

[0100]

number

[0101] Furthermore, the point cloud identification unit 230 takes x'_rot_lower as input and calculates the y value (first approximation value) output from y = f_lower(x) as the approximate value y'_est_lower. Also, the point cloud identification unit 230 takes x'_rot_upper as input and calculates the y value (second approximation value) output from y = f_upper(x) as the approximate value y'_est_upper. That is, each approximate value can be calculated as shown in the following formula (5).

[0102]

number

[0103] The point cloud identification unit 230 can identify a point cloud that satisfies the conditions that the approximate value y'_est_lower is greater than or equal to y'_rot_lower and that the approximate value y'_est_upper is less than or equal to y'_rot_upper as a point cloud Pc''(x'',y'') to be used for object detection. That is, the point cloud Pc''(x'',y'') can be identified as shown in the following formula (6).

[0104]

number

[0105] Alternatively, the point cloud identification unit 230 may identify a point cloud that satisfies the conditions that the value obtained by subtracting the first margin width margin1 from the approximate value y'_est_lower is equal to or greater than y'_rot_lower, and the value obtained by adding the second margin width margin2 to the approximate value y'_est_upper is equal to or less than y'_rot_upper, as the point cloud Pc''(x'',y'') to be used for object detection. That is, the point cloud Pc''(x'',y'') may be identified as shown in the following equation (7):

[0106]

number

[0107] The point cloud identification unit 230 may output the identified point cloud Pc''(x'',y'') to the display device 290. Then, the display device 290 may display the point cloud Pc''(x'',y''). This allows the observer to visually confirm the point cloud Pc''(x'',y'') identified by the point cloud identification unit 230.

[0108] FIG. 9 is a diagram showing an example of a point cloud Pc'(x',y') before being identified by the point cloud identification unit 230. FIG. 10 is a diagram showing an example of a point cloud Pc''(x'',y'') identified by the point cloud identification unit 230. The point cloud region V11 corresponds to the target lane area D1. The point cloud region V12 corresponds to the opposite lane area D2. The point cloud region V13 corresponds to another area.

[0109] 9, before the point cloud identification unit 230 performs identification, not only the point cloud region V11 corresponding to the target lane area D1 but also the point cloud region V12 corresponding to the opposite lane area D2 and the point cloud region V13 corresponding to other areas are included in the point cloud Pc'(x',y'). On the other hand, referring to FIG. 10, after the point cloud identification unit 230 performs identification, only the point cloud region V11 corresponding to the target lane area D1 is included in the point cloud Pc''(x'',y'').

[0110] The object detection unit 250 detects an object based on the point cloud Pc''(x'', y'') identified by the point cloud identification unit 230 (step B14). For example, the object detection unit 250 may detect an object by background subtraction processing and clustering processing. Note that the object detected by the object detection unit 250 is obtained as a three-dimensional point cloud of the object. In this specification, the object detected by the object detection unit 250 is mainly a vehicle.

[0111] Object tracking unit 260 associates the detected object in the current frame (point cloud obtained at the current time) with the detected object in the previous frame (point cloud obtained a predetermined time before the current time) (step B15). More specifically, object tracking unit 260 acquires the position of the object in the previous frame from past frame detected object information storage unit 240.

[0112] Furthermore, the object tracking unit 260 detects the position of the object in the current frame based on the object detection result. The position of the object detected by the object tracking unit 260 may be any position of the object. For example, the object tracking unit 260 may detect the position of the center of gravity of the three-dimensional point cloud of the object as the position of the object, or may detect a position estimated based on the position of the three-dimensional point cloud of the object and the velocity of the object as the position of the object. Alternatively, the object tracking unit 260 may detect the leading position or trailing position of the object as the position of the object. Known techniques may be applied to calculate the leading position and trailing position of the object.

[0113] When the distance between the object's position in the previous frame and the object's position in the current frame is equal to or less than a threshold, the object tracking unit 260 associates the object's position in the current frame with the object's position in the previous frame that is closest to the object's position in the current frame and stores the associated object in the previous frame detected object information storage unit 240. For example, the distance between the object's position in the previous frame and the object's position in the current frame may be the Euclidean distance. Note that the previous frame may also be referred to as a frame that is earlier than the current frame (a past frame).

[0114] On the other hand, when the distance between the position of the object in the previous frame and the position of the object in the current frame exceeds a threshold, the object tracking unit 260 stores the position of the object in the current frame in the past frame detected object information storage unit 240 without correlating it with the position of the object in the previous frame.

[0115] The object information calculation unit 270 calculates information about the object detected by the object detection unit 250 as object information (step B16). In the example shown in Fig. 2, the information about the object is information about a vehicle C1 traveling in lane L1 of the target lane area D1 and information about a vehicle C2 traveling in lane L2 of the target lane area D1.

[0116] For example, the object information calculation unit 270 may calculate the speed of an object as an example of object information based on the position of the object in the previous frame and the position of the object in the current frame. Furthermore, the object information calculation unit 270 may calculate the center of gravity of the object as an example of object information. For example, the center of gravity of the object may be the center of gravity of a point cloud corresponding to the object. Furthermore, the object information calculation unit 270 may calculate the leading position or trailing position of the object as an example of object information based on the position of the object in the current frame.

[0117] The object information calculation unit 270 outputs the calculated object information to the notification device 280 (step B17). Then, the notification device 280 may notify the object information. In this way, when the notification device 280 is installed in the autonomously driven vehicle C3 (FIG. 2), the occupants of the autonomously driven vehicle C3 can check the object information calculated by the object information calculation unit 270.

[0118] Furthermore, the object information calculation unit 270 outputs the calculated object information to the display device 290 (step B18). Then, the display device 290 may display the object information. This allows the observer to visually confirm the object information calculated by the object information calculation unit 270.

[0119] An example of the operation of the information processing device 10-1 according to the first embodiment of the present invention has been described above.

[0120] [1-4. Effects of the embodiment] As described above, in the information processing device 10-1 according to the first embodiment of the present invention, the approximate route calculation unit 130 calculates, as candidate approximate formulas, approximate formulas for route information after rotation by each of a plurality of candidate rotation angles for the route information. Then, the approximate route calculation unit 130 calculates a determined rotation angle and determined approximate formula to be used in a predetermined process based on the candidate approximate formulas and the route information after rotation. This improves the accuracy of approximation of the route even when the route has a curved shape.

[0121] Furthermore, the point cloud identification unit 230 performs a predetermined process based on the determined rotation angle and the determined approximation formula. This improves the accuracy of the process. In the first embodiment of the present invention, the predetermined process is a point cloud identification process used for object detection. This improves the accuracy of object detection because only the point cloud corresponding to the route information is used for object detection.

[0122] Furthermore, the point cloud identification process according to the first embodiment of the present invention significantly reduces processing time compared to the point cloud identification process according to the comparative example. For example, the comparative example is an example in which a plurality of three-dimensional regions each having a rectangular parallelepiped shape are arranged, and point clouds within the plurality of three-dimensional regions are identified as point clouds to be used for object detection.

[0123] Fig. 11 is a diagram showing the processing time of the point cloud identification process according to the first embodiment of the present invention and the point cloud identification process according to a comparative example. Referring to Fig. 11, it can be seen that the processing time required for the point cloud identification process according to this embodiment is reduced by 88.7% compared to the processing time required for the point cloud identification process according to the comparative example.

[0124] The effects of the first embodiment of the present invention have been described above.

[0125] 2. Second embodiment Next, a second embodiment of the present invention will be described. In the following description, the description of the configuration of the second embodiment of the present invention that is common to the configuration of the first embodiment of the present invention will be omitted, and the description will focus on the configuration that is different from the configuration of the first embodiment of the present invention.

[0126] [2-1. Application of information processing systems] First, with reference to FIG. 12, a situation in which an information processing system according to the second embodiment of the present invention is applied will be described.

[0127] Fig. 12 is a diagram for explaining a situation in which an information processing system according to the second embodiment of the present invention is applied. Referring to Fig. 12, an information processing device 10-2 and a laser sensor 200 included in the information processing system according to the second embodiment of the present invention are shown.

[0128] As in the first embodiment of the present invention, if laser sensor 200 is installed as an infrastructure sensor on the roadside of main line N1, information processing device 10-2 can extract traffic information for main line N1 from data obtained by laser sensor 200. Then, information processing device 10-2 can notify automatically driven vehicle C3 of the traffic information for main line N1 by wireless or the like. This allows automatically driven vehicle C3 to be expected to merge onto main line N1 safely and smoothly.

[0129] The traffic information for the main road N1 may include the travel lanes of the vehicles C1 and C2 that are present in the target lane area D1 of the main road N1.

[0130] The traffic information for the main lane N1 may also include predicted distances that vehicles C1 and C2 in the target lane area D1 of the main lane N1 will travel from their current positions to the position where autonomous vehicle C3 will make a right or left turn (hereinafter also referred to as the "reference position"). Figure 12 shows the reference position Q1 for lane L1 and the reference position Q2 for lane L2.

[0131] Furthermore, the traffic information for the main road N1 may include the time required for each of the vehicles C1 and C2 in the target lane area D1 of the main road N1 to travel its own predicted distance (hereinafter also referred to as "predicted arrival time").

[0132] If the main road N1 is a straight road, it is possible to easily determine the driving lanes of vehicles C1 and C2 present in the target lane area D1 of the main road N1, calculate the predicted distance that vehicle C1 will travel from its current position to the reference position Q1, and calculate the predicted distance that vehicle C2 will travel from its current position to the reference position Q2.

[0133] For example, assuming that lanes L1 and L2 are each straight lanes, the direction from upstream to downstream of the target lane area D1 (i.e., the length direction) is taken as the y-axis, and the direction from the left end of lane L1 to the right end of lane L2 (i.e., the width direction) facing from upstream to downstream of the target lane area D1 is taken as the x-axis.

[0134] At this time, the driving lanes of the vehicles C1 and C2 can be determined based on a comparison between the x coordinates of the vehicles C1 and C2 and a threshold value of the x component.

[0135] As an example, let's say the x-coordinate of the left edge of lane L1 is 0, and the width of lanes L1 and L2 is 3.0 m. In this case, the lane of a vehicle whose x-coordinate satisfies the condition 0≦x<3.0 m is determined to be lane L1. The x-coordinate of vehicle C1 is considered to satisfy this condition. On the other hand, the lane of a vehicle whose x-coordinate satisfies the condition 3.0 m≦x<6.0 m is determined to be lane L2. The x-coordinate of vehicle C2 is considered to satisfy this condition.

[0136] Furthermore, the predicted distance traveled by vehicle C1 from its current position to reference position Q1 can be calculated by subtracting the y coordinate of the current position of vehicle C1 from the y coordinate of reference position Q1. Similarly, the predicted distance traveled by vehicle C2 from its current position to reference position Q2 can be calculated by subtracting the y coordinate of the current position of vehicle C2 from the y coordinate of reference position Q2.

[0137] As an example, assume that the y coordinate of the current position of vehicle C1 is 50.0 m and the y coordinate of reference position Q1 is 150.0 m. In this case, the predicted distance from the current position of vehicle C1 to reference position Q1 can be calculated as 150.0 m - 50.0 m = 100.0 m by subtracting the y coordinate of the current position of vehicle C1 from the y coordinate of reference position Q1. The predicted distance from the current position of vehicle C2 to reference position Q2 can be calculated in a similar manner.

[0138] However, as shown in Fig. 12, when the main road N1 is a curved road (i.e., when the main road N1 has a curved shape with a curve), an error occurs in determining the driving lanes of the vehicles C1 and C2. This also causes an error in calculating the predicted distance that the vehicle C1 will travel from its current position to the reference position Q1, and an error in calculating the predicted distance that the vehicle C2 will travel from its current position to the reference position Q2. As a result, an error also occurs in calculating the predicted arrival time that the vehicles C1 and C2 will take to travel their predicted distances.

[0139] According to a second embodiment of the present invention, which will be described below, even when the main road N1 is a curved road, the accuracy of determining the driving lanes of the vehicles C1 and C2 can be improved, and the accuracy of calculating the predicted distance that the vehicle C1 will travel from its current position to the reference position Q1 and the accuracy of calculating the predicted distance that the vehicle C2 will travel from its current position to the reference position Q2 can be improved. As a result, the accuracy of calculating the predicted arrival time required for each of the vehicles C1 and C2 to travel its predicted distance can be improved.

[0140] 12 shows route information R. The route information R includes a plurality of route elements, and similarly to the first embodiment of the present invention, a symbol P is assigned to one of the plurality of route elements.

[0141] In the second embodiment of the present invention, the route information R is located inside each of the lanes L1 and L2 of the target lane area D1. In the following description, the route information R located inside the lane L1 will also be simply referred to as "route information R within lane L1." Also, the route information R located inside the lane L2 will also be simply referred to as "route information R within lane L2." The route information R will be described in more detail later.

[0142] The above has described the situations in which the information processing system according to the second embodiment of the present invention is applied.

[0143] [2-2. Example of information processing system configuration] Next, an example of the functional configuration of an information processing system according to a second embodiment of the present invention will be described. Fig. 13 is a block diagram showing an example of the functional configuration of an information processing system according to the second embodiment of the present invention. As shown in Fig. 13, in the information processing system according to the second embodiment of the present invention, compared to the information processing system according to the first embodiment of the present invention, information processing device 10-1 (Fig. 3) is replaced with information processing device 10-2.

[0144] (Information processing device 10-2) The information processing device 10-2 may be realized by a computer. As shown in Fig. 12, compared to the information processing device 10-1, the information processing device 10-2 does not include the point cloud identification unit 230, but instead includes a reference position information storage unit 150 and an arrival information calculation unit 275. The arrival information calculation unit 275 may be realized by a control unit (not shown). On the other hand, the reference position information storage unit 150 may be realized by a storage unit (not shown).

[0145] An example of the functional configuration of the information processing system according to the second embodiment of the present invention has been described above.

[0146] [2-3. Example of operation of information processing device] Next, an operation example of the information processing device 10-2 according to the second embodiment of the present invention will be described. The operation example of the information processing device 10-2 according to the second embodiment of the present invention can be divided into a preparatory stage and an operation stage, similar to the operation example of the information processing device 10-1 according to the first embodiment of the present invention. Hereinafter, an operation example of the information processing device 10-2 according to the preparatory stage will be described with reference to Fig. 14 (and also with reference to Figs. 12 and 13 as appropriate).

[0147] (Preparation stage) 14 is a flowchart showing an example of the operation of the information processing device 10-2 in the advance preparation stage. As shown in FIG. 14, in the advance preparation stage, the information processing device 10-2 performs step A12 and step A13 for each lane in the target lane area D1 (i.e., for each of the route information R in lane L1 and the route information R in lane L2) (step A21).

[0148] Note that steps A12 and A13 in the second embodiment of the present invention may be executed in the same manner as steps A12 and A13 in the first embodiment of the present invention.

[0149] An example of the operation of the information processing device 10-2 in the advance preparation stage has been described above with reference to FIG.

[0150] (Operational stage) Next, an example of the operation of the information processing device 10-2 in the operation stage will be described with reference to FIGS.

[0151] Fig. 15 is a flowchart showing an example of operation of information processing device 10-2 in the operation phase. As shown in Fig. 15, steps B11 and B12 are executed in the operation phase. Steps B11 and B12 in the second embodiment of the present invention may be executed in the same manner as steps B11 and B12 in the first embodiment of the present invention.

[0152] Subsequently, step B14 is executed. In the first embodiment of the present invention, it is assumed that step B13 (FIG. 7) is executed before step B14, and object detection is performed in step B14 based on the point cloud identified in step B13.

[0153] On the other hand, in the second embodiment of the present invention, it is assumed that object detection is performed based on a point cloud represented by a coordinate system converted from the original coordinate system to a reference coordinate system in step B12. However, in the second embodiment of the present invention, as in the first embodiment of the present invention, step B13 may be performed before step B14, and object detection may be performed in step B14 based on the point cloud identified in step B13.

[0154] Subsequently, steps B15 and B16 are executed. In the following description, the position in the reference coordinate system of the route information R within lane L1 will also be referred to as route information R1' in the reference coordinate system. Furthermore, the position in the reference coordinate system of the route information R within lane L2 will also be referred to as route information R2' in the reference coordinate system. The route information R1' in the reference coordinate system corresponds to an example of first route information. The route information R2' in the reference coordinate system corresponds to an example of second route information.

[0155] Similarly to the first embodiment of the present invention, the determined rotation angle θ corresponding to the route information R1′ in the reference coordinate system k_min is also expressed as θ1, and the determined rotation angle θ k_min is also expressed as θ2. The approximate formula y=f corresponding to the route information R1' in the reference coordinate system is kθ_min (x) is also expressed as y=f1(x), and the determined rotation angle y=f kθ_min (x) can also be written as y=f2(x).

[0156] The arrival information calculation unit 275 acquires a determined rotation angle θ1 corresponding to the route information R1' in the reference coordinate system, k coefficients expressing the determined approximate equation y = f1(x), and one constant from the approximate route information storage unit 140. Furthermore, the arrival information calculation unit 275 acquires a determined rotation angle θ2 corresponding to the route information R2' in the reference coordinate system, k coefficients expressing the determined approximate equation y = f2(x), and one constant from the approximate route information storage unit 140.

[0157] In this specification, it is mainly assumed that the object detected by the object detection unit 250 is a vehicle. In the example shown in Fig. 12, the vehicle C1 and the vehicle C2 are examples of vehicles. Therefore, in the following description, the object position is assumed to be the "vehicle position," and the vehicle position is assumed to be C(Cx, Cy).

[0158] In the following description, the number of lanes is N, the lane number is i (where 1≦i≦N), the determined rotation angle is θi, and the determined approximation equation is y=fi(x). In the example shown in FIG. 12, N=2, but N may be any integer equal to or greater than 2. In the following description, the reference position is Qi(Qix, Qiy). In the example shown in FIG. 12, the reference position Q1 in lane L1 and the reference position Q2 in lane L2 are examples of reference positions.

[0159] The arrival information calculation unit 275 performs predetermined processing based on the vehicle position C(Cx, Cy), the determined rotation angles θ1 to θn, and the determined approximate equations y=f1(x), y=f2(x), . . . , y=fn(x), thereby improving the accuracy of the processing.

[0160] In the second embodiment of the present invention, the arrival information calculation unit 275 executes a process of determining the vehicle's driving lane as an example of the predetermined process (step B21). More specifically, the arrival information calculation unit 275 executes a process of obtaining the rotated vehicle position C'i(C'ix, C'iy) for i=1 to N by rotating the vehicle position C(Cx, Cy) by the determined rotation angle θi as shown in the following equation (8).

[0161]

number

[0162] The arrival information calculation unit 275 inputs C'ix, which is the x value of the vehicle position C'i after rotation, into the determination approximation equation y=fi(x), as shown in the following equation (9), and performs processing to obtain the output yi value as the approximate value yi for i=1 to N.

[0163]

number

[0164] The arrival information calculation unit 275 calculates the error Error between the approximate value yi and the y value C'iy of the vehicle position C'i after rotation, as shown in the following equation (10). iThe process of calculating is performed for i=1 to N.

[0165]

number

[0166] Here, if the lane in which the vehicle is actually traveling is the driving lane n, the error Error1 is subtracted from the error Error n Error corresponding to driving lane n n Therefore, the arrival information calculation unit 275 can calculate the vehicle's driving lane n as shown in the following equation (11).

[0167]

number

[0168] In equation (11), argmin is the Error i With input, Error i is a function that outputs the lane number when is minimum. An example of determining the driving lane n when the number of lanes N is 2 will be described in detail with reference to Fig. 16. Since the number of lanes N is 2, the case where lane number i=1, 2 will be considered.

[0169] Fig. 16 is a diagram illustrating an example of determining the driving lane n when the number of lanes N is 2. In the example shown in Fig. 16, for ease of explanation, it is assumed that the determined rotation angle θ1 and the determined rotation angle θ2 are the same, and the post-rotation vehicle position C'1 (C'1x, C'1y) and the post-rotation vehicle position C'2 (C'2x, C'2y) are the same, and the post-rotation vehicle position is shown as C' (C'x, C'y). It is also assumed that the vehicle is traveling in a lane with lane number i = 2.

[0170] Also shown is the determined approximation formula y=f1(x), and the approximation value y1 of C'x to the determined approximation formula y=f1(x). Also shown is the determined approximation formula y=f2(x), and the approximation value y2 of C'x to the determined approximation formula y=f2(x). Also shown are the error Error1 between the approximation value y1 and C'y, and the error Error2 between the approximation value y2 and C'y.

[0171] 16, it can be seen that the error Error2 corresponding to lane number i = 2 is smaller than the error Error1 corresponding to lane number i = 1. Therefore, the arrival information calculation unit 275 can determine that the vehicle's traveling lane n is 2.

[0172] In the second embodiment of the present invention, the arrival information calculation unit 275 executes, as an example of a predetermined process, a process of calculating a predicted distance that the vehicle will travel from the vehicle position C(Cx, Cy) to a reference position Qn in the vehicle's driving lane n. In the following description, the predicted distance is assumed to be D.

[0173] Furthermore, the arrival information calculation unit 275 executes, as an example of predetermined processing, a process of calculating the time required for the vehicle to travel the predicted distance as a predicted arrival time (step B22). In the following description, the predicted arrival time is represented as T.

[0174] More specifically, the determined rotation angle corresponding to the vehicle's driving lane n is θn. The determined approximation equation corresponding to the vehicle's driving lane n is y=fn(x). The reference position corresponding to the vehicle's driving lane n is Qn(Qnx, Qny). The arrival information calculation unit 275 acquires the reference position Qn(Qnx, Qny) corresponding to lane n from the reference position information storage unit 150.

[0175] Furthermore, the vehicle position for calculating the predicted distance and predicted arrival time is defined as H(Hx, Hy). This vehicle position H(Hx, Hy) may be the same as the vehicle position C(Cx, Cy) already used. However, considering that it is more important from a safety standpoint to know the timing when the head of the vehicle will reach the reference position Qn, it is preferable that the vehicle position H(Hx, Hy) be the head position of the vehicle.

[0176] The arrival information calculation unit 275 rotates the vehicle position H(Hx, Hy) and the reference position Qn(Qnx, Qny) corresponding to the lane n by the determined rotation angle θn corresponding to the vehicle's driving lane n, as shown in the following equation (12). As a result, the arrival information calculation unit 275 obtains the vehicle position H'(H'x, H'y) after the rotation and the reference position Q'n(Q'nx, Q'ny) after the rotation.

[0177]

number

[0178] The arrival information calculation unit 275 calculates the predicted distance the vehicle will travel from the vehicle position H'(H'x, H'y) after rotation to the reference position Q'n(Q'nx, Q'ny) after rotation based on the vehicle position H'(H'x, H'y) after rotation, the reference position Q'n(Q'nx, Q'ny) after rotation, and the determination approximation formula y=fn(x).

[0179] Here, various methods can be applied to the calculation method of the predicted distance D. As an example, the arrival information calculation unit 275 calculates the difference between H'x, which is the x value of the vehicle position H' after rotation, and Q'nx, which is the x value of the reference position Q'n after rotation, as shown in the following equation (13). x It is calculated as follows.

[0180]

number

[0181] Next, the arrival information calculation unit 275 divides range_x by a preset division number M to calculate the x-coordinate interval step_x per division, as shown in the following equation (14). The larger the division number M, the greater the processing load, but it is expected that the calculation accuracy of the predicted distance D will improve.

[0182]

number

[0183] Next, the arrival information calculation unit 275 divides H'x to Q'nx into x-coordinate intervals step_x, as shown in the following formula (15), and calculates the x-coordinates s of M+1 sample positions. xj Calculate the x-coordinates s of the M+1 sample positions xj is input to the decision approximation formula y=fn(x) corresponding to the driving lane n, and the output is the y coordinate x of the M+1 sample positions. yj It is calculated as follows.

[0184]

number

[0185] Next, the arrival information calculation unit 275 calculates the sum of the lengths of M line segments connecting adjacent sample positions s as a predicted distance D, as shown in the following equation (16).

[0186]

number

[0187] Next, the arrival information calculation unit 275 calculates the time required for the vehicle to travel the predicted distance D as the predicted arrival time T based on the predicted distance D and the vehicle's moving speed, as shown in the following equation (17). For example, the object speed calculated by the object information calculation unit 270 may be used as the vehicle's moving speed.

[0188]

number

[0189] The arrival information calculation unit 275 outputs the vehicle's driving lane n, the predicted distance D that the vehicle will travel from the current position to the reference position Qn, and the predicted arrival time T that the vehicle will need to travel the predicted distance D as traffic information to the notification device 280 (step B23). The notification device 280 may then notify the traffic information. As a result, if the notification device 280 is installed in the automatically driven vehicle C3 (FIG. 12), the occupants of the automatically driven vehicle C3 can check the traffic information.

[0190] Furthermore, the object information calculation unit 270 outputs the traffic information to the display device 290 (step B24). Then, the display device 290 may display the traffic information, which allows the monitor to visually check the traffic information.

[0191] The arrival information calculation unit 275 may output the object information calculated by the object information calculation unit 270 together with the traffic information to the notification device 280, or may output the object information calculated by the object information calculation unit 270 together with the traffic information to the display device 290.

[0192] An example of the operation of the information processing device 10-2 according to the second embodiment of the present invention has been described above.

[0193] [2-4. Effects of the embodiment] As described above, according to the second embodiment of the present invention, similar to the first embodiment of the present invention, the accuracy of approximation of the route is improved, even when the route has a curved shape, and the accuracy of predetermined processing based on the approximation is improved.

[0194] In the second embodiment of the present invention, the predetermined processing may include processing for determining the lane in which the vehicle is traveling, processing for calculating a predicted distance that the vehicle will travel from the current position to a reference position, and processing for calculating a predicted arrival time required for the vehicle to travel the predicted distance, etc. Therefore, according to the second embodiment of the present invention, traffic information such as the lane in which the vehicle is traveling, the predicted distance, and the predicted arrival time can be obtained.

[0195] The effects of the second embodiment of the present invention have been described above.

[0196] <3. Hardware configuration example> Next, an example of the hardware configuration of the information processing device 10 according to the embodiment of the present invention will be described. Below, an example of the hardware configuration of the information processing device 900 will be described as an example of the hardware configuration of the information processing device 10 according to the embodiment of the present invention. Note that the example of the hardware configuration of the information processing device 900 described below is merely one example of the hardware configuration of the information processing device 10. Therefore, the hardware configuration of the information processing device 10 may be such that unnecessary components are deleted from the hardware configuration of the information processing device 900 described below, or new components are added. Note that the hardware of the information processing device 10 can also be realized in a similar manner.

[0197] 17 is a diagram showing a hardware configuration of an information processing device 900 as an example of the information processing device 10 according to an embodiment of the present invention. The information processing device 900 includes a CPU (Central Processing Unit) 901, a ROM (Read Only Memory) 902, a RAM (Random Access Memory) 903, a host bus 904, a bridge 905, an external bus 906, an interface 907, an input device 908, an output device 909, a storage device 910, and a communication device 911.

[0198] The CPU 901 functions as an arithmetic processing unit and control unit, and controls the overall operation of the information processing device 900 in accordance with various programs. The CPU 901 may also be a microprocessor. The ROM 902 stores programs used by the CPU 901, calculation parameters, etc. The RAM 903 temporarily stores programs used in the execution of the CPU 901, parameters that change as appropriate during the execution, etc. These are interconnected by a host bus 904 that is composed of a CPU bus, etc.

[0199] The host bus 904 is connected to an external bus 906, such as a PCI (Peripheral Component Interconnect / Interface) bus, via a bridge 905. It is not necessary to configure the host bus 904, bridge 905, and external bus 906 separately, and these functions may be implemented on a single bus.

[0200] The input device 908 is composed of input means such as a mouse, keyboard, touch panel, buttons, microphone, switches, and levers that allow the user to input information, and an input control circuit that generates an input signal based on the user's input and outputs it to the CPU 901. By operating this input device 908, the user operating the information processing device 900 can input various data to the information processing device 900 and instruct the information processing device 900 to perform processing operations.

[0201] The output device 909 includes, for example, a display device such as a CRT (Cathode Ray Tube) display device, a liquid crystal display (LCD) device, an OLED (Organic Light Emitting Diode) device, or a lamp, and an audio output device such as a speaker.

[0202] The storage device 910 is a device for storing data. The storage device 910 may include a storage medium, a recording device for recording data on the storage medium, a reading device for reading data from the storage medium, and a deletion device for deleting data recorded on the storage medium. The storage device 910 is configured, for example, with an HDD (Hard Disk Drive). This storage device 910 drives a hard disk and stores programs executed by the CPU 901 and various data.

[0203] The communication device 911 is, for example, a communication interface configured with a communication device for connecting to a network, etc. The communication device 911 may be compatible with either wireless communication or wired communication.

[0204] An example of the hardware configuration of the information processing device 10 according to the embodiment of the present invention has been described above.

[0205] <4. Various Modifications> Although the preferred embodiments of the present invention have been described in detail above with reference to the accompanying drawings, the present invention is not limited to these examples. It is clear that a person skilled in the art to which the present invention pertains can conceive of various modifications and alterations within the scope of the technical ideas set forth in the claims, and it is understood that these also naturally fall within the technical scope of the present invention.

[0206] (Route information settings) In the first embodiment and the second embodiment of the present invention, examples in which route information is set for roads have been mainly described. However, the location where route information is set does not have to be limited to roads. For example, route information may be set for a curved railway station platform. As an example, in the first embodiment of the present invention, if an object is detected in the area between the station platform and the train, route information may be set for the edge of the station platform and the edge of the train, and a point cloud used for object detection may be identified based on this route information.

[0207] (Coordinate conversion timing) In the first and second embodiments of the present invention, examples have been described in which the rotation angle and the approximation formula are determined based on the route information after the coordinate transformation from the original coordinate system to the reference coordinate system has been performed. However, the coordinate transformation may be performed on the approximate route after the determined rotation angle and the determined approximation formula have been applied to the route information.

[0208] (Error variation) In the first embodiment of the present invention and the second embodiment of the present invention, the plurality of path elements P'' after rotation θ P'' is the x-coordinate of θx and the candidate approximation equation y=f kθ (x) and the approximate value f based on kθ (P'' θx ) and multiple path elements P'' after rotation θ P'' is the y coordinate of θy The maximum error is maxError kθ However, the approximate value f kθ (P'' θx ) and P'' θy As the error between the rotation angle θ and the approximation f kθ (P'' θx ) and P'' θy Alternatively, a cumulative error obtained by adding up the errors may be used.

[0209] (Trend curve variations) In the first and second embodiments of the present invention, an example has been described in which one approximation formula (for example, an approximation formula expressed by one curve) is determined for route information. However, the route information may be divided into multiple sections, and different approximation formulas (for example, a linear function expressed by a straight line and a quadratic function expressed by a curve) may be determined for each section.

[0210] (Area where object detection occurs) In the first embodiment of the present invention, an example has been described in which a point cloud used for object detection is identified based on route information, and an object is detected based on the identified point cloud. As a result, objects existing inside the area between the route information can be mainly detected. However, objects may also be detected based on the remaining point cloud after excluding the identified point cloud. As a result, objects existing outside the area between the route information can be mainly detected.

[0211] (Setting the reference position) In the second embodiment of the present invention, an example has been described in which the reference position Q1 of lane L1 is determined in advance. However, when the waiting position of an autonomous vehicle waiting to turn right or left is detected based on the stopping position of the autonomous vehicle, the route element P closest to that waiting position may be determined as the reference position Q1 for lane L1. Similarly, the reference position Q2 of lane L2 may also be determined as the reference position Q2, where the route element P closest to the waiting position of an autonomous vehicle waiting to turn right or left is determined as the reference position Q2.

[0212] (Lane crossing detection) In the second embodiment of the present invention, an example of determining the lane in which a vehicle is traveling has been described. However, a state in which a vehicle is crossing from one lane to another (hereinafter also referred to as "crossing lanes") may also be determined. For example, if the difference between the error Error1 and the error Error2 is smaller than a threshold, the arrival information calculation unit 275 may determine that the vehicle is crossing lanes because it is considered that the vehicle is located near the boundary between the lane L1 and the lane L2.

[0213] Various modifications according to the embodiments of the present invention have been described above. [Explanation of symbols]

[0214] 10. Information processing equipment 100 Route information storage unit 110 Reference coordinate system information storage unit 120 Coordinate system conversion section 130 Approximate route calculation unit 140 Approximate route information storage unit 150 Reference position information storage unit 160 Route information output unit 200 laser sensor 210 Point cloud acquisition part 220 Sensor coordinate system conversion unit 230 Point cloud identification part 240 Past frame detected object information storage unit 250 Object detection unit 260 Object Tracking Unit 270 Object information calculation unit 275 Arrival information calculation unit 280 Notification device 290 Display device

Claims

1. a calculation unit that calculates, as candidate approximate formulas, approximate formulas for route information after rotation by each of a plurality of candidate rotation angles for the route information, calculates, for each candidate rotation angle, an approximate value based on a first coordinate of the rotated route information and the candidate approximate formula, based on the candidate approximate formulas and the rotated route information, and calculates, as a determined rotation angle and a determined approximate formula, the candidate rotation angle and the candidate approximate formula when an error between the approximate value and a second coordinate of the rotated route information is smallest; a processing unit that performs predetermined processing including processing for identifying a point cloud to be used for object detection based on the determined approximation formula and a point cloud after rotation by the determined rotation angle for each of a plurality of pieces of route information included in the route information; An information processing device comprising:

2. The information processing device includes: an object detection unit that detects an object based on the point cloud obtained by the sensor; the processing unit performs the predetermined processing based on the object position based on the detection result of the object, the determined rotation angle, and the determined approximation formula. The information processing device according to claim 1 .

3. the predetermined processing includes a process of calculating a predicted distance that the object will move from the object position to the reference position based on the object position after rotation by the determined rotation angle, the reference position after rotation by the determined rotation angle, and the determination approximation formula. The information processing device according to claim 2 .

4. the predetermined processing includes a process of calculating, as a predicted arrival time, a time required for the object to travel the predicted distance based on the predicted distance and a moving speed of the object. The information processing device according to claim 3 .

5. The calculation unit calculates the determined rotation angle and the determined approximation formula for each of the plurality of route information, the processing unit performs the predetermined processing based on the object positions after rotation by the determined rotation angles of each of the plurality of pieces of route information and the determined approximation equation. The information processing device according to claim 2 .

6. each of the plurality of pieces of route information corresponds to a lane in which the object may be present; the predetermined processing includes processing for determining a lane in which the object exists based on an object position after rotation by the determined rotation angle for each of the plurality of pieces of route information and the determination approximation formula; The information processing device according to claim 5 .

7. The predetermined processing is A process of acquiring a reference position in the lane in which it is determined that the object exists; and calculating a predicted distance that the object will move from the object position to the reference position based on the object position after rotation by the determined rotation angle in the lane in which the object is determined to exist, the reference position after rotation by the determined rotation angle, and the determination approximation formula. The information processing device according to claim 6 .

8. The information processing device an object detection unit that detects an object based on the point cloud; The information processing device according to claim 1 .

9. the plurality of pieces of route information include first route information and second route information; a second coordinate of the same point after rotation by the determined rotation angle of the first path information is smaller than a second coordinate of the same point after rotation by the determined rotation angle of the second path information; The predetermined processing is The second coordinates after rotation of the first route information by the determined rotation angle are a condition that the first coordinate after rotation by the determined rotation angle of the first route information and the determined approximate expression of the first route information are equal to or greater than a first approximate value or a value obtained by subtracting a first margin width from the first approximate value; and The second coordinates of the second route information after rotation by the determined rotation angle are a point cloud that satisfies a condition that the first coordinates after rotation by the determined rotation angle of the second path information and the determined approximation formula of the second path information are equal to or less than a value obtained by adding a second margin width to the second approximation value, a process of identifying the point cloud to be used for the object detection; The information processing device according to claim 8 .

10. calculating, as candidate approximate formulas, approximate formulas for route information after rotation by each of a plurality of candidate rotation angles for the route information; calculating, for each candidate rotation angle, approximate values based on first coordinates of the rotated route information and the candidate approximate formulas, based on the candidate approximate formulas and the rotated route information; and calculating, as a determined rotation angle and determined approximate formula, the candidate rotation angle and the candidate approximate formula when an error between the approximate value and second coordinates of the rotated route information is smallest; performing a predetermined process including a process of identifying a point cloud to be used for object detection based on the determined approximation formula and a point cloud after rotation by the determined rotation angle for each of a plurality of pieces of route information included in the route information; 2. A computer-implemented information processing method, comprising:

11. Computer, a calculation unit that calculates, as candidate approximate formulas, approximate formulas for route information after rotation by each of a plurality of candidate rotation angles for the route information, calculates, for each candidate rotation angle, an approximate value based on a first coordinate of the rotated route information and the candidate approximate formula, based on the candidate approximate formulas and the rotated route information, and calculates, as a determined rotation angle and a determined approximate formula, the candidate rotation angle and the candidate approximate formula when an error between the approximate value and a second coordinate of the rotated route information is smallest; a processing unit that performs predetermined processing including processing for identifying a point cloud to be used for object detection based on the determined approximation formula and a point cloud after rotation by the determined rotation angle for each of a plurality of pieces of route information included in the route information; A program that functions as a

Images