Information processing apparatus

Abstract

To provide an information processing apparatus that can accurately acquire a vehicle speed.SOLUTION: A control unit 15 acquires, of a result of detection performed by a lidar 21 that detects a feature around a vehicle C, a result of detection of a white line D1 in a window W that moves together with the vehicle C, continuously at a time interval Δt, and from the plurality of detection results, detects the position of a line in the direction of movement of the vehicle C at one end of the white line D1. Based on a temporal change of the position of the detected line, the control unit calculates the average speed v of the vehicle C.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to an information processing device that performs predetermined processing based on the detection results of a detection unit that detects features around a moving object. [Background technology]

[0002] For example, autonomous driving systems, which have been developed in recent years, grasp the situation by recognizing objects around the vehicle, generate an optimal target trajectory, and control the vehicle to travel along that target trajectory. If the accuracy of the vehicle's self-localization is poor, there is a possibility that the actual driving trajectory will deviate from the target trajectory, reducing the safety of autonomous driving. Accurate self-localization is one of the important factors to ensure the safety of autonomous driving.

[0003] Conventional car navigation systems often use the Global Navigation Satellite System (GNSS) for self-position estimation, but this has led to problems such as poor accuracy in places where signal reception is impossible, such as inside tunnels, or in environments where multipath interference is common, such as between buildings.

[0004] Therefore, a so-called dead reckoning technique is known, which estimates the vehicle position based on the vehicle's running conditions (e.g., vehicle speed and yaw rate). In order to improve the accuracy of the vehicle position estimation using dead reckoning, it is necessary to accurately acquire the vehicle's running conditions, such as the vehicle speed.

[0005] A known technique for obtaining vehicle speed with high accuracy is to correct vehicle speed pulses, as described in Patent Document 1. Patent Document 1 describes correcting an arithmetic formula for determining traveling speed based on the relationship between the count value Cp of output pulses from a wheel speed sensor counted between two features and the distance D between the two features. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] Japanese Patent Application Laid-Open No. 2008-8783 Summary of the Invention [Problem to be solved by the invention]

[0007] In the invention described in Patent Document 1, the calculation formula is corrected based on the distance D between two features, so if the accuracy of the distance D between the two features is low, the accuracy of the correction of the calculation formula may also decrease, which may result in a decrease in the accuracy of the calculated vehicle speed.

[0008] One example of a problem to be solved by the present invention is to accurately acquire the vehicle speed as described above. [Means for solving the problem]

[0009] The invention described in claim 1, which was made to solve the above problem, is characterized by comprising an acquisition unit that continuously acquires, at predetermined time intervals, detection results of features within a predetermined detection area that moves together with a moving body from the detection results of a detection unit that detects features around the moving body; an extraction unit that extracts position information of characteristic parts of the features within the detection area in the direction of movement of the moving body from the acquired multiple detection results; and a calculation unit that calculates the speed of the moving body based on changes in the extracted position information over time.

[0010] The invention described in claim 9 is characterized by comprising a detection unit that detects features around a moving body, an acquisition unit that continuously acquires, at predetermined time intervals, detection results of the features within a predetermined detection area that moves together with the moving body from the detection results of the detection unit, an extraction unit that extracts position information of characteristic parts of the features within the detection area in the direction of movement of the moving body from the acquired multiple detection results, and a calculation unit that calculates the speed of the moving body based on changes in the extracted position information over time.

[0011] The invention described in claim 10 is an information processing method executed by an information processing device that performs predetermined processing based on the detection results of a detection unit that detects features around a moving body, and is characterized by including an acquisition step of continuously acquiring, at predetermined time intervals, detection results of the features within a predetermined detection area that moves together with the moving body from the detection results of the detection unit; an extraction step of extracting, from the acquired multiple detection results, position information of characteristic parts of the features within the detection area in the direction of movement of the moving body; and a calculation step of calculating the speed of the moving body based on changes over time in the extracted position information.

[0012] Furthermore, the invention as set forth in claim 11 is characterized in that the information processing method as set forth in claim 10 is executed by a computer. [Brief explanation of the drawings]

[0013] [Figure 1] 1 shows the configuration of a detection device having an information processing device according to a first embodiment of the present invention. [Figure 2] 2 is an explanatory diagram illustrating detection of a white line by the detection device shown in FIG. 1. FIG. [Figure 3] 2 is an explanatory diagram of coordinate conversion of measurement values ​​in the detection device shown in FIG. 1. FIG. [Figure 4] FIG. 2 is an explanatory diagram of the scan angle in the lidar shown in FIG. 1. [Figure 5] FIG. 2 is an explanatory diagram of the scan interval in the lidar shown in FIG. 1. [Figure 6] FIG. 2 is an explanatory diagram of the vertical angle in the lidar shown in FIG. 1. [Figure 7] 2 is an explanatory diagram of a speed calculation method in the control unit shown in FIG. 1. FIG. [Figure 8] 2 is an explanatory diagram of a speed calculation method in the control unit shown in FIG. 1. FIG. [Figure 9] 8 is a flowchart of the velocity calculation method shown in FIG. 7. [Figure 10] 10 is a flowchart of the velocity calculation process shown in FIG. 9. [Figure 11] 9 is a flowchart of the velocity calculation method shown in FIG. 8. [Figure 12] FIG. 10 is an explanatory diagram of a speed calculation method in a control unit according to a second embodiment of the present invention. [Figure 13] FIG. 10 is an explanatory diagram of a speed calculation method in a control unit according to a second embodiment of the present invention. [Figure 14] 13 is a flowchart of the velocity calculation method shown in FIG. 12. [Figure 15] 14 is a flowchart of the velocity calculation method shown in FIG. 13. [Figure 16] FIG. 10 is an explanatory diagram illustrating detection of a white line according to a third embodiment of the present invention. [Figure 17] 17 is an explanatory diagram of a speed calculation method in the control unit shown in FIG. 16. FIG. [Figure 18] 18 is a flowchart of the velocity calculation method shown in FIG. 17. [Figure 19] 17 is an explanatory diagram of another speed calculation method in the control unit shown in FIG. 16. FIG. [Figure 20] 20 is a flowchart of the velocity calculation method shown in FIG. 19. DETAILED DESCRIPTION OF THE INVENTION

[0014] An information processing device according to one embodiment of the present invention will be described below. In the information processing device according to one embodiment of the present invention, an acquisition unit continuously acquires, at a predetermined time interval, detection results of features within a predetermined detection area in which the mobile object moves, from among detection results of a detection unit that detects features around the mobile object. An extraction unit extracts position information of characteristic parts of the features within the detection area in the direction of movement of the mobile object from the multiple detection results acquired by the acquisition unit. A calculation unit then calculates the speed of the mobile object based on temporal changes in the extracted position information. In this way, it is possible to calculate the speed based on position information of characteristic parts, such as boundary parts of features such as road markings, within the detection area of ​​a detection unit such as a lidar, and the speed of the mobile object can be calculated and acquired with high accuracy.

[0015] Furthermore, the characteristic portion of a feature may be one end of the feature. In this way, it is possible to detect the speed by detecting the characteristic portion at one location, that is, one end of the feature. Therefore, it is not necessary to consider the accuracy of the distance to other features, and the speed of the moving object can be calculated with high accuracy.

[0016] Furthermore, the change in position information over time may be the elapsed time from the start of detection of the characteristic portion to the end of detection, thereby making it possible to calculate the speed of the moving object based on the time during which the characteristic portion is detected.

[0017] Furthermore, the length of the feature corresponding to the direction of movement of the moving object may be known, and the characteristic portions of the feature may be one end and the other end of the feature. In this way, when the length corresponding to the direction of movement of the feature is known, the speed of the moving object can be detected based on that length.

[0018] Furthermore, the change in position information over time may be the elapsed time from the start of detection of one end to the start of detection of the other end, or the elapsed time from the end of detection of one end to the end of detection of the other end. In this way, the speed of the moving object can be calculated based on the time during which the characteristic portion is detected.

[0019] Furthermore, the detection areas may be two locations, one before the other in the direction of movement of the moving object, separated by a predetermined distance, and the characteristic part of the feature may be one end of the feature. By doing so, since the predetermined distance is known, it is possible to calculate the speed of the moving object even if the length of the feature corresponding to the direction of movement is unknown.

[0020] Furthermore, the change in position information over time may be the elapsed time from the end of detection in the detection area in front of the characteristic part to the start of detection in the detection area behind the characteristic part. In this way, the speed of the moving object can be calculated based on the time during which the characteristic part is detected.

[0021] The detection area may also be a rectangular area set within the detectable range of the detection unit. This allows only the area where there is a possibility of detecting a feature to be set as the detection area. Therefore, it is possible to prevent a decrease in the accuracy of speed calculation due to noise caused by the detection of objects other than the target object, such as features.

[0022] Furthermore, a detection device according to one embodiment of the present invention includes a detection unit that detects features around a moving object. An acquisition unit continuously acquires, at predetermined time intervals, detection results of features within a predetermined detection area in which the moving object moves together with the moving object. An extraction unit extracts position information of characteristic parts of the features within the detection area in the moving direction of the moving object from the multiple detection results acquired by the acquisition unit. A calculation unit calculates the speed of the moving object based on temporal changes in the extracted position information. In this manner, a detection device including a detection unit such as a lidar can calculate the speed by detecting position information of characteristic parts, such as boundaries of features like road markings, within the detection area, thereby enabling the speed of the moving object to be calculated with high accuracy.

[0023] In addition, an information processing method according to one embodiment of the present invention includes, in an acquisition step, continuously acquiring, at a predetermined time interval, detection results of features within a predetermined detection area in which the moving object moves, from among the detection results of a detection unit that detects features around the moving object; and, in an extraction step, extracting position information of characteristic parts of the features within the detection area in the direction of movement of the moving object from the multiple detection results acquired in the acquisition step. Then, in a calculation step, calculating the speed of the moving object based on temporal changes in the extracted position information. In this way, it is possible to calculate the speed by detecting position information of characteristic parts, such as boundary parts of features such as road markings, within the detection area of ​​a detection unit such as a lidar, and the speed of the moving object can be calculated and acquired with high accuracy.

[0024] The above-described information processing method may also be implemented as an information processing program that causes a computer to execute the information processing method. In this way, it becomes possible to calculate the speed by using the computer to detect position information of characteristic parts, such as boundaries of features such as road markings, within the detection area of ​​the detection unit, and to accurately calculate and obtain the speed of the moving object. [Example]

[0025] An information processing device according to a first embodiment of the present invention will be described with reference to Figures 1 to 11. The information processing device according to this embodiment is included in a detection device 1, and moves together with a vehicle as a moving body.

[0026] A schematic block diagram of a detection device 1 according to this embodiment is shown in Fig. 1. The detection device 1 includes a sensor group 11, a storage unit 12, a control unit 15, and an output unit 16.

[0027] The sensor group 11 includes a lidar 21 , a vehicle speed sensor 22 , an acceleration sensor 23 , a gyro sensor 24 , an inclination sensor 25 , a temperature sensor 26 , and a GPS receiver 27 .

[0028] The lidar 21, which serves as a detection unit, measures the distance to an object in the external world in a discrete manner by emitting a pulsed laser beam. The lidar 21 outputs a point cloud of measurement points indicated by a pair of the distance to the object where the laser beam is reflected and the emission angle of the laser beam. In this embodiment, the lidar 21 is used to detect features present around the vehicle. The term "feature" encompasses all natural and artificial objects on the ground. Examples of features include route features located on the vehicle's route (i.e., road) and surrounding features located around the road. Examples of route features include road signs, traffic lights, guardrails, pedestrian bridges, etc., as well as the road itself. In other words, route features also include letters and figures drawn on the road surface and the shape of the road (road width and curvature). Examples of surrounding features include buildings (houses, stores) and signs located along the road.

[0029] The vehicle speed sensor 22 measures pulses (also called "axle rotation pulses") consisting of pulse signals generated in accordance with the rotation of the vehicle's wheels, and detects the vehicle speed. The acceleration sensor 23 detects the acceleration in the vehicle's traveling direction. The gyro sensor 24 detects the angular velocity of the vehicle when the vehicle changes direction. The tilt sensor 25 detects the tilt angle (also called "gradient angle") in the pitch direction of the vehicle relative to the horizontal plane. The temperature sensor 26 detects the temperature around the acceleration sensor 23. The GPS (Global Positioning System) receiver 27 receives multiple The absolute position of the vehicle is detected by receiving radio waves containing positioning data from GPS satellites. The output of each sensor in the sensor group 11 is supplied to the control unit 15.

[0030] The storage unit 12 stores information processing programs executed by the control unit 15, information necessary for the control unit 15 to execute predetermined processes, and the like. In this embodiment, the storage unit 12 stores a map database (DB) 10 including road data and information on features. The map DB 10 may be updated periodically. In this case, for example, the control unit 15 receives partial map information related to the area to which the vehicle position belongs from an external server device that manages map information via a communication unit (not shown), and reflects the information in the map DB 10. Instead of the storage unit 12 storing the map DB 10, a server device that can communicate with the detection device 1 may store the map DB 10. In this case, the control unit 15 acquires necessary feature information and the like from the map DB 10 by communicating with the external server device.

[0031] The output unit 16 outputs, for example, the speed information calculated by the control unit 15 to an automatic driving control device or other in-vehicle devices such as meters.

[0032] The control unit 15 includes a CPU (Central Processing Unit) that executes programs. , controls the entire detection device 1. The control unit 15 includes an acquisition unit 15a, an extraction unit 15b, and a calculation unit 15c. In this embodiment, the control unit 15 calculates the speed of the vehicle based on the features detected by the LIDAR 21.

[0033] The acquisition unit 15a continuously acquires detection results in a window, which will be described later, from among the detection results of features detected by the LIDAR 21 at predetermined time intervals.

[0034] The extraction unit 15b extracts position information of characteristic parts of features within the window in the direction of vehicle movement from the detection results acquired by the acquisition unit 15a.

[0035] The calculation unit 15c calculates the speed of the vehicle based on the temporal change in the position information extracted by the extraction unit 15b.

[0036] The control unit 15 of the detection device 1 having the above-described configuration functions as the information processing device according to this embodiment.

[0037] Next, a speed detection method in the control unit 15 (information processing device) of the detection device 1 configured as described above will be described. In the following explanation, dashed lane boundary lines (so-called white lines) are used as features. Because white lines are coated with retroreflective material, they have high reflective intensity and are easy to detect by LIDAR.

[0038] The detection of white lines in this embodiment will be described with reference to Fig. 2. In Fig. 2, a vehicle C is traveling from left to right in the figure. A lidar 21L is installed on the left front side of the vehicle C, and a lidar 21R is installed on the right front side of the vehicle C.

[0039] If the detection range of the lidars 21L and 21R is A, then a window W, which is a rectangular area, is set within the detection range A. This window W is set at a position within the detection range A where the white lines D1 and D2 are easily detected. This window W becomes the detection area that moves along with the moving object in this embodiment. Note that, although this embodiment will be described using the lidar 21 installed in front of the vehicle C, a lidar installed behind the vehicle C may also be used. Furthermore, only one of the lidars 21L and 21R may be used.

[0040] Next, coordinate conversion of the measurement values ​​by the lidar 21 will be described with reference to FIG. 3. As described above, the lidar 21 outputs the distance to the object and the emission angle of the laser light. That is, it outputs the distance to the measurement target and the horizontal and vertical angles of the measurement target. As shown on the left side of FIG. 3, the longitudinal axis of the vehicle C is defined as xb, the lateral axis of the vehicle C as yb, and the vertical axis of the vehicle C as zb. In this case, as shown on the right side of FIG. 3, if the distance from the center of gravity of the vehicle C to the measurement target is defined as r, the horizontal angle to the measurement target is defined as α, and the vertical angle to the measurement target is defined as β, then the position Z(i) of the measurement target can be expressed by the following equation (1).

number

[0041] Next, the scan interval of the LIDAR 21 will be described. The LIDAR 21 in this embodiment scans an object by sequentially emitting pulsed light from one side to the other in the horizontal direction. Therefore, as shown in the upper part of FIG. 4, the scan trajectory becomes a line when viewed from above. Therefore, the acquisition unit 15a acquires information from the LIDAR 21 at the intervals of this scanned line. In other words, the acquisition unit 15a continuously acquires the detection results of the features at predetermined time intervals.

[0042] Also, typical lidars include those that obtain multiple lines by moving a beam that scans horizontally up and down vertically, and those that obtain multiple lines by arranging multiple optical systems that scan horizontally in the vertical direction. It is known that the scan interval of these types of lidars increases with increasing distance from the vehicle C (see also Figure 5). This is because the angle between the lidar 21 and the feature (road surface) decreases with increasing distance from the vehicle C.

[0043] As shown in Figure 5, the scanning lines within the window W are divided into S1, S2, ..., S 10and the intervals between each line are d1, d2, ..., d9. As shown in Fig. 6, when the vertical angles of each line are β(i), β(i+1), β(i+2), β(i+3) and the distances from the emission point to each line are r(i), r(i+1), r(i+2), r(i+3), the intervals between the lines d(i), d(i+1), d(i+2) are expressed by the following equations (2) to (4).

number

[0044] Then, the speed of vehicle C is calculated based on the detection of one end of one of the broken lines that constitute the white line as a result of the white line passing through the range of window W. The following explanation will be given using the symbols of the lines and the symbols of the intervals between the lines shown in FIG.

[0045] First, as shown in the upper part of FIG. 7, it is assumed that lines S1 and S2 detect one end of white line D1. Then, as shown in the middle part of FIG. 7, it is assumed that lines S1 to S4 detect white line D1 after a time Δt has elapsed. The travel distance x(k) at this time is expressed by the following equation (5), and the speed v(k) of vehicle C at this time is expressed by equation (6). Here, d2 / 2, d4 / 2, etc. in equations (5) and (6) are regarded as 1 / 2 of the scan interval because the exact position is unknown when the position of the end of white line D1 is between scans, as shown in FIG. 7. Therefore, the closer the scan interval, the smaller the error between this 1 / 2 and the actual position.

number

[0046] Next, assume that lines S1 to S7 detect white line D1 after time Δt has elapsed, as shown in the lower part of Figure 7. The travel distance x(k+1) at this time is expressed by the following equation (7), and the speed v(k+1) of vehicle C at this time is expressed by equation (8).

number

[0047] Therefore, the velocity v obtained by averaging the velocity v(k) and the velocity v(k+1) is expressed by the following equation (9).

number

[0048] In Figure 7, the speed was calculated by detecting the end of the white line D1 closer to the vehicle C, but as shown in Figure 8, the speed can also be calculated by detecting the end of the white line D1 farther from the vehicle C.

[0049] First, as shown in the upper part of Figure 8, lines S3 to S 10 8, after the time Δt has elapsed, the white line D1 is detected. 10 When the vehicle C detects the white line D1, the travel distance x(k) is expressed by the following equation (10), and the speed v(k) of the vehicle C at this time is expressed by the following equation (11).

number

[0050] Next, as shown in the lower part of FIG. 8, after the time Δt has elapsed, lines S8 to S 10 Suppose that vehicle C detects white line D1. The travel distance x(k+1) at this time is expressed by the following equation (12), and the speed v(k+1) of vehicle C at this time is expressed by equation (13).

number

[0051] Therefore, the velocity v obtained by averaging the velocity v(k) and the velocity v(k+1) is expressed by the following equation (14).

number

[0052] Next, the operation (information processing method) of the control unit 15 (information processing device) configured as described above will be described with reference to the flowcharts of Figures 9 to 11. Furthermore, by configuring the control unit 15 as a computer having a CPU or the like, these flowcharts can be configured as an information processing program executed by the computer.

[0053] First, the speed calculation process described in FIG. 7 will be described with reference to the flowchart in FIG. 9. In step S101, the extraction unit 15b determines whether or not the starting point of the white line D1 has been detected within the window W. If not (No), this step is repeated. If detected (Yes), the speed calculation process is performed in step S102. The detection of the starting point of the white line D1 indicates, for example, the state shown in the upper part of FIG. 7. The speed calculation process will be described later. That is, in the case of FIG. 7, the starting point (one end) of the white line D1 is a characteristic part of the feature. As described above, the starting point of this white line D1 is the start of one line that constitutes the dashed line.

[0054] Next, in step S103, the extraction unit 15b determines whether the start of the white line D1 has been detected within the window W. If the start of the white line D1 has not been detected (No), the process returns to step S102. If the start of the white line D1 has been detected (Yes), the process determines that detection of the end of the white line D1 has been completed and terminates the speed calculation process. The state in which the start of the white line D1 has been detected within the window W refers to, for example, a state in which more time has passed since the lower part of FIG. 7 and one end of the white line D1 can no longer be detected within the window W. In this state, the one end, which is a characteristic part of the white line D1, cannot be detected, and the speed calculation process is terminated. That is, in steps S101 and S103, the extraction unit 15a extracts position information of the characteristic part of the feature within the window W (detection area) in the movement direction of the vehicle C (moving object) from the multiple detection results acquired by the acquisition unit 15a.

[0055] Next, the speed calculation process (step S102) in FIG. 9 will be described with reference to the flowchart in FIG. 10. This flowchart is executed by the calculation unit 15c. First, in step S201, the movement distance x(k) is calculated from the line number of the previous cycle and the current line number. The line number of the previous cycle is, for example, line S2 in the upper part of FIG. 7, and the line number of the current cycle is, for example, line S4 in the middle part of FIG. 7. Therefore, the movement distance from line S2 to S4 is as shown in equation (5). Furthermore, one cycle refers to Δt when the speed is calculated at time intervals of Δt.

[0056] Next, in step S202, the velocity v(k) is calculated from the moving distance x(k) and the scan period Δt.

[0057] Then, in step S203, the k velocities are averaged to obtain a velocity v. As described in the flowchart of FIG. 9, the velocity calculation process shown in FIG. 10 is executed multiple times depending on the determination in step S103, so steps S201 and S202 are essentially executed multiple times, and averaging is performed in step S203 each time they are executed. That is, the flowchart of FIG. 10 (step S102) calculates the velocity of vehicle C (moving body) based on the temporal change in the extracted position information. Furthermore, by repeating the velocity calculation process until a No determination is made in step S103, the velocity v that is finally calculated is calculated based on the temporal change in the position information, that is, the elapsed time from the start to the end of detection of the characteristic portion.

[0058] That is, step S101 functions as an acquisition step and an extraction step, step S102 functions as a calculation step, and step S103 functions as an extraction step.

[0059] Next, the speed calculation process described in FIG. 8 will be described with reference to the flowchart in FIG. 11. In step S301, the extraction unit 15b determines whether or not the end of the white line D1 has been detected within the window W. If not detected (No), this step is repeated. If detected (Yes), speed calculation process is performed in step S302. The end of the white line D1 indicates, for example, the state shown in the upper part of FIG. 8. The speed calculation process is the process shown in the flowchart in FIG. 10. That is, in the case of FIG. 8, the end (one end) of the white line D1 is the characteristic part of the feature.

[0060] Next, in step S303, the extraction unit 15b determines whether the end of the white line D1 has been detected within the window W. If the end of the white line D1 has not been detected (No), the process returns to step S302. If the end of the white line D1 has been detected (Yes), the process determines that the detection of the end of the white line D1 has been completed and ends the speed calculation process. The state in which the end of the white line D1 has been detected within the window W refers to, for example, a state in which the white line D1 is no longer detected within the window W after further time has passed since the lower part of FIG. 8. In this state, one end, which is a characteristic part of the white line D1, cannot be detected, and therefore the speed calculation process ends.

[0061] According to this embodiment, the control unit 15 continuously acquires, at time intervals of Δt, detection results of the white line D1 within a window W that moves together with the vehicle C, from among the detection results of the LIDAR 21 that detects features around the vehicle C. From these detection results, the control unit 15 detects the position of the scan line of the LIDAR 21 in the direction of movement of the vehicle C at one end of the white line D1. Then, the control unit 15 calculates the speed v of the vehicle C based on the change over time in the position of the detected scan line. In this way, it is possible to calculate the speed based on the position of a characteristic feature, such as the boundary of a feature such as the white line D1, within the window W set in a detection unit such as the LIDAR. Furthermore, the speed of the moving object can be calculated based on the time during which the characteristic feature was detected. Therefore, the speed of the moving object can be calculated and acquired with high accuracy.

[0062] Furthermore, it is possible to detect the speed by detecting a single characteristic portion, namely, one end of the white line D1, which eliminates the need to consider the accuracy of the distance to other features.

[0063] The detection area is a rectangular window W set within the detection range A of the lidar 21. This allows only the area where there is a possibility of detecting the white line D1 to be set as the detection area. Therefore, it is possible to prevent a decrease in the accuracy of speed calculation due to noise caused by the detection of an object other than the white line D1 that is expected to be detected, for example. [Example]

[0064] Next, a detection device and an information processing device according to a second embodiment of the present invention will be described with reference to Figures 12 to 15. Note that the same parts as those in the first embodiment described above will be given the same reference numerals and descriptions thereof will be omitted.

[0065] This embodiment has the same configuration as that shown in FIG. 1, but is a method for calculating the speed of vehicle C when the length of one of the lines that make up the dashed white line is known. The length of this one line (hereinafter referred to as the white line length) is often determined by law or regulation depending on the type of road. Therefore, the speed can be calculated by using the length of the white line, which is known information. In this embodiment, the length of the white line is assumed to be included in the map DB 10.

[0066] Next, a specific example of speed calculation will be described with reference to FIGS. 12 and 13. FIG. 12 shows a state in which a window W moving together with a vehicle C detects a white line D1. In FIG. 12, it is assumed that, as in the first embodiment, scanning is performed at intervals of, for example, Δt to detect the white line D1. In this case, if the length of the detected portion when the starting point of the white line D1 is first detected is a1, the movement distance between each time (per Δt) is x1 to x8, the length of the undetected portion when the ending point of the white line D1 is first detected is b1, and the length of the white line D is L, then the following equation (15) is established. That is, in this embodiment, the characteristic portions of the feature are one end (the starting point of the white line D1) and the other end (the ending point of the white line D1) of the feature.

number

[0067] Here, if S1 to S2 shown in FIG. 7 detect the white line D1, a1 is d1+d2 / 2. Similarly, b1 is also d1+d2 / 2 if S3 to S2 shown in FIG. 10 If the vehicle detects the white line D1, the result is d1+d2 / 2.

[0068] Transforming equation (15) gives the following equation (16).

number

[0069] Therefore, equation (16) shows the movement distance for eight scans. Therefore, the velocity v can be calculated using the following equation (17), which divides L-a1+b1 by the time required for eight scans. In equation (17), N shows the number of scans. In other words, the change in position information over time is the elapsed time from the start of detection of one end (the beginning of the white line D1) to the start of detection of the other end (the end of the white line D1). Furthermore, by calculating the velocity v in this way, it is not necessary to calculate the movement distance x(k) for each scan, etc.

number

[0070] Note that although a1 is defined as the length immediately after the detection of the starting point of the white line D1, it may also be the detected value a little later. Also, although b1 is defined as the length immediately after the detection of the end point of the white line D1, it may also be the detected value a little later. For example, if the length one time after the detection of the starting point of the white line D1 is detected and used as a1, a1 will be a large value, but the number of scans N will be reduced by one, so the same value will be calculated as a result. Also, for example, if b1 is defined as the length one time after the detection of the end point of the white line D1, b1 will be a large value, but the number of scans will be increased by one, so the same value will be calculated as a result.

[0071] FIG. 13 shows an example in which a white line D1 is disappearing from the window W. In this case, the speed can be calculated using the same concept as in FIG. 12. That is, if the length of the undetected portion of the white line D1 when the starting point of the white line D1 was last detected is a2, the movement distance between each time is x1 to x8, the length of the detected portion when the ending point of the white line D1 was last detected is b2, and the length of the white line D is L, then the following equation (18) is established.

number

[0072] Here, a2 and b2 can be calculated based on what has been explained with reference to FIGS. 7 and 8, in the same way as explained with reference to FIG.

[0073] Transforming equation (18) gives the following equation (19).

number

[0074] Therefore, equation (19) shows the amount of movement for eight scans. Therefore, the speed v can be calculated using the following equation (20), which divides L + a2 - b2 by the time required for eight scans. In equation (20), N shows the number of scans. Also, in the case of Figure 13, the line spacing is denser at the rear of the window (closer to vehicle C) (see Figure 5), so the accuracy of calculating (measuring) a2 and b2 is higher, and as a result, the accuracy of the speed v is improved.

number

[0075] In this case, a2 and b2 do not have to be the last detected values, but can be the values ​​from one time before. Even if a2 and b2 are large values, the number of scans increases or decreases, and the same result is calculated.

[0076] Next, the speed calculation process described in FIG. 12 will be described with reference to the flowchart in FIG. 14. First, in step S401, the extraction unit 15b initializes the number of scans N to "0." Next, in step S402, the extraction unit 15b determines whether or not the starting point of the white line D1 in the window W has been detected. If not detected (No), this step is repeated. If detected (Yes), in step S403, the calculation unit 15c calculates the length a1 of the detected portion when the starting point of the white line D1 was first detected. "Detecting the starting point of the white line D1" refers to a state in which one end of the white line D1 closer to the vehicle has been detected, as shown in the second row from the top in FIG. 12, for example.

[0077] Next, in step S404, the extraction unit 15b counts up the number of scans N, and in step S405, the extraction unit 15b determines whether or not the end of the white line D1 has been detected within the window. If the end of the white line D1 has not been detected, the process returns to step S404. If the end of the white line D1 has been detected, the calculation unit 15c calculates the length b1 of the undetected portion when the end of the white line D1 was first detected in step S406. Detecting the end of the white line D1 refers to a state in which one end of the white line D1 farther from the vehicle has been detected, as shown in the bottom row of Figure 12, for example.

[0078] Then, in step S407, the calculation unit 15c acquires the length L of the white line (the length of the white line D1) from the map DB 10, and in step S408, the calculation unit 15c calculates the speed v by equation (17).

[0079] Next, the speed calculation process described in Fig. 13 will be described with reference to the flowchart in Fig. 15. First, in step S501, the extraction unit 15b initializes the number of scans N to "0." Next, in step S502, the extraction unit 15b determines whether or not the white line D1 has been detected within the window W. If the white line D1 has not been detected (No), the process repeats this step. If the white line D1 has been detected (Yes), the process proceeds to step S503. The detection of the white line D1 refers to a state in which one or more scan lines within the window W detect the white line D1, as shown in the top row of Fig. 13, for example.

[0080] Next, in step S503, the extraction unit 15b determines whether the start of the white line D1 has been detected, and if the start of the white line D1 has not been detected (No), this step is repeated, but if the start of the white line D1 has been detected, the calculation unit 15c calculates the length a2 of the previous undetected portion of the white line D1 in step S504. The start of the white line D1 having been detected means, for example, as shown in the second row from the top in Figure 13, that one end of the white line D1 closer to the vehicle is not detected from within the window W.

[0081] Next, in step S505, the extraction unit 15b counts up the number of scans N, and in step S506, the extraction unit 15b determines whether the end of the white line D1 has been detected within the window W. If the end of the white line D1 has not been detected, the process returns to step S505. If the end of the white line D1 has been detected, the calculation unit 15c calculates the length b2 of the detected portion of the white line D1 immediately before the end of the white line D1 in step S507. The end of the white line D1 has been detected when, for example, further time has passed since the bottom of Figure 13 and the white line D1 is no longer detected within the window W.

[0082] Then, in step S508, the calculation unit 15c acquires the length L of the broken line (the length of the white line D1) from the map DB 10, and in step S509, the calculation unit 15c calculates the speed v using equation (20).

[0083] According to this embodiment, the length of the broken line (white line D1) in the moving direction is known, and the characteristic portions of the feature are one end and the other end of the white line D1. In this way, when the length of the white line D1 in the moving direction is known, the speed of the vehicle C can be detected based on that length.

[0084] The change in position information over time is the elapsed time from the start of detection of one end to the start of detection of the other end, or the elapsed time from the last detection of one end to the last detection of the other end, i.e., the time required for N scans. In this way, the speed of vehicle C can be calculated from the time during which the characteristic part is detected. [Example]

[0085] Next, a detection device and an information processing device according to a third embodiment of the present invention will be described with reference to Figures 16 to 20. Note that the same parts as those in the first and second embodiments described above will be given the same reference numerals and descriptions thereof will be omitted.

[0086] The configuration of this embodiment is the same as that of Fig. 1. In the first and second embodiments described above, the window W is provided only at the front or rear of the vehicle C, but in this embodiment, a front window WF and a rear window WR are provided as shown in Fig. 16. That is, two windows are provided at two locations, one at the front and one at the rear in the direction of movement of the moving object, spaced a predetermined distance apart.

[0087] In this embodiment, a rider 21FL is installed on the left front side of the vehicle C, and similarly, a rider 21FR is installed on the right front side of the vehicle C. Furthermore, a rider 21RL is installed on the left rear side of the vehicle C, and similarly, a rider 21RR is installed on the right rear side of the vehicle C. Note that, although riders are installed on the left and right sides of the vehicle in this embodiment, they may be installed only on the right or left side.

[0088] If the detection range of the lidars 21FL and 21FR is AF, a window WF is set in the detection range AF. If the detection range of the lidars 21RL and 21RR is AR, a window WR is set in the detection range AR. These windows WF and WR are set at positions within the detection range A where the white lines D1 and D2 are easily detected.

[0089] Next, a specific example of speed calculation will be described with reference to Fig. 17. Fig. 17 shows the state in which the windows WF and WR moving together with the vehicle C detect the white line D1. In Fig. 17, it is assumed that the white line D1 is detected by scanning at intervals of Δt, for example, as in the first embodiment. In this case, the length of the last detected end of the white line D1 in the front window WF is a3, and the moving distance between each time (per Δt) is x1 to x 10If the length of the undetected portion when the end of the white line D1 inside the rear window is first detected is b3 and the gap (distance) between the front window WF and the rear window WR is G, the following equation (21) is established. Note that the gap G is a value that can be obtained in advance from the installation position of each lidar, the detection range of the lidar, etc.

number

[0090] Here, a3 and b3 can be calculated based on what has been explained with reference to Figures 7 and 8, as in the case of Figure 12. That is, in this embodiment, the characteristic part of the feature is one end of the feature (the end of the white line).

[0091] Transforming equation (21) gives the following equation (22).

number

[0092] Therefore, equation (22) shows the amount of movement for 10 scans. Therefore, the velocity v can be calculated using the following equation (23), which divides G+a3+b3 by the time required for 10 scans. In equation (23), N shows the number of scans. In other words, the change in position information over time is the elapsed time from the last detection in the detection area (window WF) in front of the characteristic part to the start of detection in the detection area (window WR) behind it. Furthermore, in this embodiment, the velocity v can be calculated even when the length of the dashed line (white line) is unknown.

number

[0093] Note that processing may start from a detection just before the end of the front detection area, or from the beginning to the end of the rear detection area. In this case, a3 and b3 will have larger values, but the scan count number N will also increase, so the same values ​​will be calculated as a result.

[0094] Next, the speed calculation process described with reference to FIG. 17 will be described with reference to the flowchart of FIG. 18. First, in step S601, the extraction unit 15b initializes the number of scans N to "0." Next, in step S602, the extraction unit 15b determines whether or not the white line D1 has been detected in the front window WF. If the white line D1 has not been detected (No), this step is repeated. If the white line D1 has been detected (Yes), in step S603, the extraction unit 15b determines whether or not the end of the white line D1 has been detected in the front window WF. If the end of the white line D1 has not been detected (No), this step is repeated. If the end of the white line D1 has been detected, in step S604, the calculation unit 15c calculates the length a3 of the most recent white line detection portion. Here, detecting the end of the white line D1 in the front window WF refers to a state in which one or more scan lines detect the white line D1 in the front window WF, as shown in the top row of FIG. 17. Furthermore, the terminus of the white line D1 within the front window WF being no longer detected refers to a state in which all scanning lines within the front window WF no longer detect the end of the white line D1 on the side farther from the vehicle, as shown in the second row from the top of Figure 17 onwards.

[0095] Next, in step S605, the extraction unit 15b counts up the number of scans N, and in step S606, the extraction unit 15b determines whether or not the end of the white line D1 has been detected within the rear window WR. If the end of the white line D1 has not been detected, the process returns to step S605. If the end of the white line D1 has been detected, the calculation unit 15c calculates the length b3 of the undetected portion of the white line D1 in step S607. Detecting the end of the white line D1 refers to a state in which one end of the white line D1 closer to the vehicle has been detected, as shown in the bottom row of Figure 17, for example.

[0096] Then, in step S608, the calculation unit 15c obtains the gap G between the front window WF and the rear window WR, and in step S609, the calculation unit 15c calculates the velocity v using equation (22).

[0097] Next, another specific example of speed detection in this embodiment will be described with reference to Fig. 19. Fig. 19, like Fig. 17, shows the state in which the windows WF and WR moving together with the vehicle C detect the white line D1. Also, in Fig. 19, as in the first embodiment, it is assumed that the white line D1 is detected by scanning at intervals of, for example, Δt. In this case, the length of the undetected portion of the white line D1 in the front window WF when the starting end of the white line D1 was last detected is a4, and the moving distance between each time (per Δt) is x1 to x 10 If the length of the detected portion when the starting point of the white line D1 inside the rear window is first detected is b4 and the gap (distance) between the front window WF and the rear window WR is G, the following equation (24) is established. Note that the gap G is a value that can be obtained in advance from the installation position of each lidar, the detection range of the lidar, etc.

number

[0098] Here, a4 and b4 can be calculated based on what has been explained with reference to Figures 7 and 8, as in the case of Figure 12. That is, in this embodiment, the characteristic part of the feature is one end of the feature (the start of the white line).

[0099] Transforming equation (24) gives the following equation (25).

number

[0100] Therefore, equation (25) shows the movement distance for 10 scans. Therefore, the velocity v can be calculated using the following equation (26), which divides G+a4+b4 by the time required for 10 scans. In equation (26), N shows the number of scans. In other words, the change in position information over time is the elapsed time from the last detection in the detection area (window WF) in front of the characteristic part to the start of detection in the detection area (window WR) behind it. Also, in this example, the velocity v can be calculated even when the length of the dashed line (white line) is unknown.

number

[0101] Note that processing may start from the detection just before the end of the front detection area, or from the beginning to the end of the rear detection area. In this case, a4 and b4 will have larger values, but the scan count number N will also increase, so the same values ​​will be calculated as a result.

[0102] Next, the speed calculation process described with reference to FIG. 19 will be described with reference to the flowchart of FIG. 20. First, in step S701, the extraction unit 15b initializes the number of scans N to "0." Next, in step S702, the extraction unit 15b determines whether or not the white line D1 has been detected in the front window WF. If the white line D1 has not been detected (No), this step is repeated. If the white line D1 has been detected (Yes), in step S703, the extraction unit 15b determines whether or not the starting edge of the white line D1 has been detected in the front window WF. If the starting edge has not been detected (No), this step is repeated. If the starting edge has been detected, in step S704, the calculation unit 15c calculates the length a4 of the previous undetected white line portion. Here, detecting the starting edge of the white line D1 in the front window WF refers to a state in which one or more scan lines in the front window WF detect the white line D1, as shown in the top row of FIG. 19. Furthermore, when the starting end of the white line D1 is no longer detected within the front window WF, this refers to a state in which all scan lines within the front window WF no longer detect the end of the white line D1 closest to the vehicle, as shown in the second row from the top in Figure 19 onwards.

[0103] Next, in step S705, the extraction unit 15b counts up the number of scans N, and in step S706, the extraction unit 15b determines whether or not the starting end of the white line D1 has been detected within the rear window WR. If the starting end of the white line D1 has not been detected, the process returns to step S705. If the starting end of the white line D1 has been detected, the calculation unit 15c calculates the length b4 of the detected portion of the white line D1 in step S707. Detecting the starting end of the white line D1 refers to a state in which one end of the white line D1 closer to the vehicle has been detected, as shown in the bottom row of Figure 19, for example.

[0104] Then, in step S708, the calculation unit 15c obtains the gap G between the front window WF and the rear window WR, and in step S709, the calculation unit 15c calculates the velocity v using equation (26).

[0105] According to this embodiment, there are two detection areas, one in front of the vehicle C and one in behind the vehicle C in the direction of travel, separated by a gap G, and the characteristic part of the feature is one end of the white line D1. In this way, since the gap G is known, it is possible to calculate the speed of the vehicle C even if the feature has an unknown length corresponding to the travel direction.

[0106] Furthermore, the change in position information over time is the elapsed time from the last detection in the detection area in front of the characteristic part to the start of detection in the detection area behind it, i.e., the time for N scans. By doing this, it is possible to detect parts with close line spacing and high accuracy in both the front and rear windows, and the speed of vehicle C can be calculated from the time during which the characteristic part was detected.

[0107] The above-described embodiments may be combined. Since a vehicle experiences vertical vibration, pitching, and rolling depending on the road surface conditions, errors may be introduced using only one method of the embodiment. Therefore, for example, the speeds calculated using the methods of the embodiments may be averaged to determine the final speed. Furthermore, since the accuracy of the calculated speed increases when the line spacing is denser, the speed calculated using the dense line spacing may be weighted more heavily.

[0108] In the above description, dashed lines (white lines) of lane markings are used, but in the first and third embodiments, other road markings, signs, and other features may be used. Furthermore, in the above description, a lidar is used as the detection unit, but an in-vehicle camera may also be used.

[0109] Furthermore, the present invention is not limited to the above-described embodiments. In other words, a person skilled in the art can implement various modifications in accordance with conventionally known knowledge without departing from the gist of the present invention. As long as such modifications still comprise the configuration of the information processing device of the present invention, they are of course included in the scope of the present invention. [Explanation of symbols]

[0110] 1. Detection device 15 Control unit (information processing device) 15a Acquisition part 15b Extraction part 15c Calculation part 21 Lidar (detection unit) S101 Detect the start of the white line within the window (acquisition process, extraction process) S102 Speed ​​calculation process (calculation process) S103 All white lines in the window (extraction process)

Claims

1. an acquisition unit that continuously acquires, at predetermined time intervals, detection results of features within a predetermined detection area detected by a detection unit that moves together with the mobile object; an extraction unit that extracts position information of characteristic parts of the features within the detection area from the plurality of detection results obtained; a calculation unit that calculates the speed of the moving object based on a change over time in the extracted position information, the characteristic portion of the feature is one end of the feature, The change in the position information over time is the elapsed time from the start of detection of the characteristic part to the end of detection.

1. An information processing device comprising:

2. an acquisition unit that continuously acquires, at predetermined time intervals, detection results of features within a predetermined detection area detected by a detection unit that moves together with the mobile object; an extraction unit that extracts position information of characteristic parts of the features within the detection area from the plurality of detection results obtained; a calculation unit that calculates the speed of the moving object based on a change over time in the extracted position information, a length of the feature corresponding to the direction of movement of the moving object is known, and the feature's characteristic portions are one end and the other end of the feature; The change in the position information over time is at least one of the time elapsed from the start of detection of the one end to the start of detection of the other end and the time elapsed from the end of detection of the one end to the end of detection of the other end.

1. An information processing device comprising:

3. an acquisition unit that continuously acquires, at predetermined time intervals, detection results of features within a predetermined detection area detected by a detection unit that moves together with the mobile object; an extraction unit that extracts position information of characteristic parts of the features within the detection area from the plurality of detection results obtained; a calculation unit that calculates the speed of the moving object based on a change over time in the extracted position information, The detection areas are located at two locations, one before the other and one after the other in the direction of movement of the moving object, with a predetermined distance between them, and the characteristic portion of the feature is one end of the feature.

1. An information processing device comprising:

4. 4. The information processing apparatus according to claim 3, wherein the change in position information over time is the elapsed time from the end of detection in the detection area in front of the characteristic part to the start of detection in the detection area in the rear of the characteristic part.

5. 5. The information processing apparatus according to claim 1, wherein the detection area is a rectangular area set within a detectable range of the detection unit.

6. a detection unit that detects features around the moving object; an acquisition unit that continuously acquires, at predetermined time intervals, detection results of the features within a predetermined detection area detected by the detection unit that moves together with the mobile object; an extraction unit that extracts position information of characteristic parts of the features within the detection area from the plurality of detection results obtained; a calculation unit that calculates the speed of the moving object based on a change over time in the extracted position information, the characteristic portion of the feature is one end of the feature, The change in the position information over time is the elapsed time from the start of detection of the characteristic part to the end of detection. A detection device characterized by:

7. a detection unit that detects features around the moving object; an acquisition unit that continuously acquires, at predetermined time intervals, detection results of the features within a predetermined detection area detected by the detection unit that moves together with the mobile object; an extraction unit that extracts position information of characteristic parts of the features within the detection area from the plurality of detection results obtained; a calculation unit that calculates the speed of the moving object based on a change over time in the extracted position information, a length of the feature corresponding to the direction of movement of the moving object is known, and the feature's characteristic portions are one end and the other end of the feature; The change in the position information over time is at least one of the time elapsed from the start of detection of the one end to the start of detection of the other end and the time elapsed from the end of detection of the one end to the end of detection of the other end. A detection device characterized by:

8. a detection unit that detects features around the moving object; an acquisition unit that continuously acquires, at predetermined time intervals, detection results of the features within a predetermined detection area detected by the detection unit that moves together with the mobile object; an extraction unit that extracts position information of characteristic parts of the features within the detection area from the plurality of detection results obtained; a calculation unit that calculates the speed of the moving object based on a change over time in the extracted position information, The detection areas are located at two locations, one before the other and one after the other in the direction of movement of the moving object, with a predetermined distance between them, and the characteristic portion of the feature is one end of the feature. A detection device characterized by:

9. An information processing method executed by an information processing device that performs predetermined processing based on a detection result of a detection unit that detects features around a moving object, an acquisition step of continuously acquiring, at predetermined time intervals, detection results of the features within a predetermined detection area detected by the detection unit moving together with the mobile object; an extraction step of extracting position information of characteristic portions of the features within the detection area from the plurality of detection results obtained; a calculation step of calculating a speed of the moving object based on a change over time in the extracted position information, the characteristic portion of the feature is one end of the feature, The change in the position information over time is the elapsed time from the start of detection of the characteristic part to the end of detection. An information processing method comprising:

10. An information processing method executed by an information processing device that performs predetermined processing based on a detection result of a detection unit that detects features around a moving object, an acquisition step of continuously acquiring, at predetermined time intervals, detection results of the features within a predetermined detection area detected by the detection unit moving together with the mobile object; an extraction step of extracting position information of characteristic portions of the features within the detection area from the plurality of detection results obtained; a calculation step of calculating a speed of the moving object based on a change over time in the extracted position information, a length of the feature corresponding to the direction of movement of the moving object is known, and the feature's characteristic portions are one end and the other end of the feature; The change in the position information over time is at least one of the time elapsed from the start of detection of the one end to the start of detection of the other end and the time elapsed from the end of detection of the one end to the end of detection of the other end. An information processing method comprising:

11. An information processing method executed by an information processing device that performs predetermined processing based on a detection result of a detection unit that detects features around a moving object, an acquisition step of continuously acquiring, at predetermined time intervals, detection results of the features within a predetermined detection area detected by the detection unit moving together with the mobile object; an extraction step of extracting position information of characteristic portions of the features within the detection area from the plurality of detection results obtained; a calculation step of calculating a speed of the moving object based on a change over time in the extracted position information, The detection areas are located at two locations, one before the other and one after the other in the direction of movement of the moving object, with a predetermined distance between them, and the characteristic portion of the feature is one end of the feature. An information processing method comprising:

12. An information processing program for causing a computer to execute the information processing method according to any one of claims 9 to 11.

Images