JP7452069B2 - Road gradient estimation device, road gradient estimation system, and road gradient estimation method - Google Patents

Description

The present disclosure relates to techniques for estimating road slope used in vehicles.

Acquire the 3D vehicle coordinate axes or 3D distance measurement data of the white line using a camera or lidar, calculate the relative slope of the road, and further collect detection data from a moving body sensor equipped with the vehicle that detects throttle opening, etc. A technique for predicting the absolute road slope using the above-mentioned method is known (for example, Patent Document 1).

JP 2016-200499 Publication

However, the conventional road slope prediction technology does not take into account the prediction of the road slope with high accuracy using detection data from cameras and lidar, that is, captured image data and detection point data. Furthermore, since either a camera or a lidar is used, there is also the problem that the road slope cannot be predicted if a problem occurs in the camera or lidar, or if the detection performance of the camera or lidar is low.

Therefore, there is a need to improve the accuracy of estimating road gradients using captured images and detection points.

The present disclosure can be implemented as the following aspects.

A first aspect provides a road gradient estimation device for use in a vehicle. The road slope estimating device according to the first aspect includes an attention area determination unit that determines an attention area on a road using a captured image or an environmental light image acquired from an imaging device or a lidar, and a lidar control unit. a detection point group acquisition unit that acquires a detection point group by increasing the resolution or signal strength of the detection point group in the region of interest, and is capable of acquiring a detection point group of interest by increasing the resolution or signal strength of the detection point group in the region of interest, and a detection point group acquisition unit that increases the resolution of the target detection point group to obtain a high-resolution detection point group as the target detection point group, and a gradient estimation unit that estimates a road slope using the target detection point group , has a light-receiving element array (22) constituted by a plurality of light-receiving elements (220), and the detection point group acquisition unit acquires the detection point group using a predetermined number of the plurality of light-receiving elements as a light reception unit. The detection point group acquisition unit increases the resolution in the vertical or horizontal direction by decreasing the number of the light receiving elements constituting the light receiving unit and increasing the number of the light receiving units.

According to the road slope estimating device for a vehicle according to the first aspect, it is possible to improve the accuracy of estimating the road slope using the captured image and the detection points.

A second aspect provides a method for estimating road slope in a vehicle. A road gradient estimation method according to a second aspect determines an area of interest on a road using a captured image obtained from an imaging device or an environmental light image obtained from a lidar, and the lidar detects a plurality of light receiving areas. The lidar has a light-receiving element array composed of elements, and by decreasing the predetermined number of the light-receiving elements constituting the light-receiving unit and increasing the number of the light-receiving units, the lidar can be adjusted vertically or horizontally. obtaining a high-resolution detection point group as a group of detection points of interest in which the resolution of the group of detection points in the region of interest is increased by increasing the resolution of the direction , and estimating a road slope using the group of detection points of interest; Equipped with

According to the road gradient estimation method for a vehicle according to the second aspect, it is possible to improve the estimation accuracy of the road gradient using the captured image and the detection points. Note that the present disclosure can also be realized as a road gradient estimation program for a vehicle or a computer-readable recording medium that records the program.

FIG. 1 is an explanatory diagram showing an example of a vehicle equipped with a road slope estimating device according to a first embodiment. FIG. 1 is an explanatory diagram showing a schematic configuration of a rider used in the first embodiment. FIG. 2 is an explanatory diagram schematically showing a light receiving element array used in the first embodiment. FIG. 1 is a block diagram showing the functional configuration of a road slope estimating device according to a first embodiment. 5 is a flowchart showing a process flow of gradient estimation processing executed by the road gradient estimation device according to the first embodiment. FIG. 3 is an explanatory diagram for explaining determination of a region of interest in a captured image. FIG. 3 is an explanatory diagram showing an example of a detection point group during high-resolution processing. An explanatory diagram showing an example of an environmental light image. FIG. 3 is an explanatory diagram showing how a road defined by a group of detection points is extrapolated using an environmental light image. An explanatory diagram showing a front region of interest. FIG. 7 is an explanatory diagram showing how a current region of interest is determined using a previous region of interest. 5 is a flowchart showing a processing flow executed when detecting a current target object using a target object detected in the past. FIG. 2 is an explanatory diagram illustrating the resolution of a conventional lidar in the vertical direction. FIG. 7 is an explanatory diagram illustrating the resolution of the lidar in the vertical direction in the fifth embodiment. FIG. 3 is an explanatory diagram showing light receiving units in a light receiving element array for increasing the resolution of the lidar in the vertical direction as the distance from the vehicle increases.

A road gradient estimation device, a road gradient estimation system, and a road gradient estimation method for a vehicle according to the present disclosure will be described below based on several embodiments.

First embodiment:
As shown in FIG. 1, a road gradient estimating device 100 for a vehicle according to the first embodiment is mounted on a vehicle 50 and used. The road gradient estimating device 100 acquires surrounding information of the vehicle 50 from a lidar (Light Detection and Ranging) 20 and a camera 30. The road gradient estimation system 10 includes at least a road gradient estimation device 100, a lidar 20, and a camera 30. In addition, the vehicle 50 is other than this. A wheel speed sensor, a yaw rate sensor, and a driving support control device may be provided to perform driving support. When the purpose is forward detection, the lidar 20 is placed, for example, at the front of the vehicle 50, and the camera 30 is placed, for example, at the top of the windshield of the vehicle 50. The term "road gradient" as used herein more specifically refers to a relative road gradient, ie, a change in road gradient; however, in order to simplify the description, the term "road gradient" will be used below.

As shown in FIG. 2, the lidar 20, which is a distance measuring device, includes a light receiving section 21, a light emitting section 24, a scanning mirror 26, an electric motor 40, and a rotation angle sensor 41. For example, when scanning in the horizontal direction, the lidar 20 has a predetermined scanning angle range NR in the horizontal direction, and the light emitting unit 24 uses a unit scanning angle obtained by dividing the scanning angle range NR into a plurality of angles as a unit. By irradiating the detection light and receiving the reflected light by the light receiving section 21, detection points are acquired over the entire scanning angle range NR, and distance measurement is realized. The unit scan angle defines the resolution of the lidar 20 or the resolution of the distance measurement results obtained by the lidar 20, and as the unit scan angle becomes smaller, that is, as the number of detection points increases, the resolution in the horizontal direction HD and the resolution decrease. becomes higher. Acquisition of detection points in the lidar 20, that is, light emission and light reception processing, is performed when scanning the scan angle range NR forward in one direction or when scanning the scan angle range NR back and forth in both directions.

The light receiving section 21 includes a light receiving element array 22, a light receiving control section 23, and a light receiving lens (not shown), and outputs a detection signal indicating a detection point in response to reception of the reflected light of the detection light emitted from the light emitting section 24. Execute light reception processing. As shown in FIG. 3, the light-receiving element array 22 is a flat optical sensor in which a plurality of light-receiving elements 220 are arranged vertically and horizontally, that is, two-dimensionally. A photodiode constitutes each light receiving element. Note that the term "light-receiving pixel" is sometimes used as the minimum unit of light-receiving processing, that is, the light-receiving unit corresponding to a detection point. It means any of the light-receiving pixels 222 configured by the element. In the light-receiving element array 22, as the number of light-receiving pixels, that is, the number of light-receiving elements constituting a light-receiving unit, is decreased, the number of light-receiving units, that is, the number of detection points increases. In the present embodiment, for example, standard light reception processing is performed using a standard resolution using a light receiving pixel 222 constituted by eight light receiving elements 220 as a light receiving unit. In the present embodiment, when increasing the resolution of the vertical direction VD, the light reception control unit 23 receives an instruction from the road slope estimating device 100, and selects a light receiving pixel 221 composed of one light receiving element 220 as a light receiving unit. The light reception process is executed as follows, and high-resolution light reception processing with higher resolution than the standard light reception process is realized. Note that the number of light-receiving elements 220 included in the light-receiving pixels 221 and 222 is merely an example, and may be two or more and nine or more, respectively. It is sufficient that the number of light receiving elements is smaller than the number of light receiving elements 220 included in the pixel 222. Further, the light receiving element array 22 may be an optical sensor in which a plurality of light receiving elements 220 are arranged one-dimensionally in the vertical or horizontal direction.

The light reception control unit 23 executes a light reception process that outputs an incident light intensity signal according to the amount or intensity of incident light incident on each light receiving pixel in units of unit scan angles at which light emission is performed by the light emitting unit 24. . Specifically, the light reception control unit 23 uses the light reception pixels 221 and 222 as a unit, and controls the current generated by the light reception elements constituting each of the light reception pixels 221 and 222 according to the amount of incident light, or the current for each unit scanning angle. The converted voltage is taken out and outputted to the road slope estimating device 100 as an incident light intensity signal. Alternatively, it can be said that an incident light intensity signal corresponding to the total number of photons received by the light receiving elements constituting each light receiving pixel is output to the road slope estimating device 100. Therefore, as the number of times of light reception increases due to repeated scanning, the signal strength increases. In general, in SPAD, since the amount of incident light obtained by one light receiving element 220 is small, the incident intensity signals from eight light receiving elements 220, such as the light receiving pixel 222, are added by an adder (not shown) and S/ N can be improved. Therefore, in order to increase the resolution in the vertical direction VD, during high-resolution light reception processing in which the light receiving pixel 221 made up of one light receiving element is used, the reference light receiving process in which the light receiving pixel 222 made up of eight light receiving elements is used. A decrease in signal strength, ie, a decrease in S, occurs compared to during processing. As will be described later, this decrease in signal intensity is caused by an increase in the incident light intensity signal due to an increase in the number of acquisitions of a group of detection points in a specific scanning range in the horizontal HD, that is, an increase in the number of times of light emission and light reception processing per unit scanning angle. compensated or reduced by

The light emitting unit 24 includes a light emission control unit (not shown), a light emitting element, and a collimator lens, and emits detection light in units of scanning angles. The light emitting element is, for example, one or more infrared laser diodes, and emits infrared laser light as detection light. The light emitting section 24 may include a single light emitting element or a plurality of light emitting elements in the vertical direction. When a plurality of light emitting elements are provided, the light emitting elements that emit light can be switched depending on the scanning timing. The light emission control unit drives the light emitting element with a drive signal having a pulse drive waveform and generates infrared laser light in accordance with a light emission control signal that instructs the light emitting element to emit light, which is inputted from the road gradient estimating device 100 for each unit scanning angle. Executes light emission. The light emitting unit 24 repeatedly emits detection light to a specific scanning range when performing high-resolution light reception processing. The infrared laser beam emitted from the light emitting unit 24 is reflected by the scanning mirror 26 and is emitted toward the outside of the lidar 20, that is, the range where detection of the target object is desired.

The electric motor 40 includes an electric motor driver (not shown). A rotation angle sensor 41 for detecting the rotation angle of the electric motor 40 is arranged on the electric motor 40 . The motor driver controls the rotation angle of the motor 40 by changing the voltage applied to the motor 40 in response to the rotation angle signal input from the rotation angle sensor 41 and the rotation angle instruction signal output from the road slope estimating device 100. . The electric motor 40 is, for example, an ultrasonic motor, a brushless motor, a brush motor, or a voice coil motor, and includes a known mechanism for reciprocating in the scanning angle range NR. A scanning mirror 26 is attached to the tip of the output shaft of the electric motor 40. The scanning mirror 26 is a reflector, that is, a mirror body, that scans the detection light emitted from the light emitting unit 24 in the horizontal direction HD, and is reciprocated by the electric motor 40 to scan the scanning angle range NR in the horizontal direction HD. is realized. The scanning mirror 26 realizes scanning of detection light and reception of reflected light in a scanning angle range of 120 degrees and 180 degrees, for example. In the area of interest FA where detection points are to be acquired intensively or intensively by the lidar 20, a group of detection points that are acquired with a higher number of detections than the number of detections in other areas in the scanning angle range NR are other than the group of detection points of interest. This detection point group can be called a high signal strength detection point group which has a higher signal strength than when scanning other areas in the scanning angle range NR. As a result of increasing the signal strength of each detection point forming the target detection point group, it is possible to compensate for the signal strength of the incident light intensity signal, that is, the decrease in S, which decreases due to high-resolution light reception processing. In addition to the horizontal direction HD, the scanning mirror 26 scans in the vertical direction VD in addition to the horizontal direction HD, when a single light emitting element is provided or when the detection light cannot be emitted over the entire vertical direction VD, that is, in the vertical direction VD. It is also possible to change the scanning position in . In order to realize scanning in the horizontal direction HD and the vertical direction VD, the scanning mirror 26 may be a polygon mirror, for example a polygon mirror, or a single surface with a mechanism to be swung in the vertical direction VD. It may also include a mirror body or another single mirror body that is swung in the vertical direction VD. Note that the scanning mirror 26 may be rotationally driven by the electric motor 40 to perform rotational scanning, and in this case, the light emitting unit 24 and the light receiving unit 21 perform light emission and light reception processing in accordance with the scanning angle range NR. It would be good if it were done. By making the scanning speed in the specific scanning angle range CR slower than the scanning speed in the other scanning angle ranges NR, resolution in the horizontal direction HD can be increased. In this embodiment, an example in which the resolution in the vertical direction VD is increased will be described, but instead of the vertical direction VD, only the resolution in the horizontal direction HD may be increased, or both the vertical direction VD and the horizontal direction HD may be increased. The resolution in may be increased. An increase in the resolution in the horizontal direction HD can be achieved by reducing the unit scanning angle or slowing the unit scanning speed of the scanning mirror 26.

The detection light emitted from the light emitting section 24 is reflected by the scanning mirror 26 and scanned over the horizontal scanning angle range NR with unit scanning angle SA as a unit. The reflected light, which is the detection light reflected by the target object, is reflected by the scanning mirror 26 to the light receiving section 21, and is incident on the light receiving section 21 at every unit scanning angle SA. The unit scan angle SA at which the light reception process is performed is sequentially incremented, and as a result, scanning for the light reception process over a desired scan angle range NR becomes possible. Note that if the entire area in the vertical direction VD cannot be irradiated with the detection light, the irradiation position in the vertical direction VD is changed during each scan in the horizontal direction HD, and multiple scans in the horizontal direction HD are performed. The light emitting section 24 and the light receiving section 21 may be rotated by the electric motor 40 together with the scanning mirror 26, or they may be separate from the scanning mirror 26 and do not need to be rotated by the electric motor 40. Furthermore, a configuration is provided in which the scanning mirror 26 is not provided, but a plurality of light emitting elements and light receiving element arrays 22 arranged in an array are provided, and the laser beam is sequentially irradiated directly to the outside world and the reflected light is directly received. It's good to be prepared.

The road slope estimating device 100 includes a central processing unit (CPU) 101 as an arithmetic unit, a memory 102 as a storage unit, an input/output interface 103 as an input/output unit, and a clock generator (not shown). The CPU 101, memory 102, input/output interface 103, and clock generator are connected via an internal bus 104 so that they can communicate in both directions. The memory 102 performs a process of determining an area of interest on a road using a captured image or an ambient light image obtained from the camera 30 or the lidar 20, a process of controlling the lidar 20 to obtain a group of detection points of interest in the area of interest, More specifically, the process of increasing the number of times the LIDAR 20 acquires a group of detection points in the area of interest FA, that is, the number of times of light emission/light reception processing, and increasing the incident light intensity signal of the group of detection points in the area of interest FA, the area of interest. A slope estimation program for executing a process of increasing the resolution of a detection point group in FA to obtain a high-resolution detection point group as a target detection point group, and a slope estimation process of estimating a road slope using a high-resolution detection point group. It includes a memory that stores Pr1 in a nonvolatile and read-only manner, such as a ROM, and a memory that can be read and written by the CPU 101, such as a RAM. In the read/write area 102a of the memory 102, information regarding detected objects, such as coordinate position information and shape information, is stored as object history OR. The CPU 101, that is, the road gradient estimating device 100, expands the gradient estimation program Pr1 stored in the memory 102 into a readable/writable memory and executes the program. functions as Note that the CPU 101 may be a single CPU, a plurality of CPUs that execute each program, or a multitasking type CPU that can execute a plurality of programs simultaneously.

A light receiving section 21, a light emitting section 24, a camera 30, an electric motor 40, and a rotation angle sensor 41 are connected to the input/output interface 103 via control signal lines, respectively. A light emission control signal is transmitted to the light emitting unit 24, an incident light intensity signal is received from the light receiving unit 21, a rotation angle instruction signal is transmitted to the electric motor 40, and a rotation angle signal is transmitted from the rotation angle sensor 41. is received. In this embodiment, furthermore, when performing high-resolution light reception processing, the light reception unit 21 transmits a light reception unit change signal to change the light reception unit to the light reception control unit 23.

The camera 30 is an imaging device equipped with one imaging device such as a CCD or an imaging device array, and is a sensor that receives visible light and outputs external shape information or shape information of an object as image data as a detection result. . The camera 30 includes an image processing unit (not shown), and performs object classification processing on the image data using, for example, semantic segmentation, and combines pixels indicating the same object to classify each object. Extract the image region showing the object. In this embodiment, the camera 30 extracts partition lines and road boundary lines generally called white lines, or extracts three-dimensional objects on the road, such as median strips, guardrails, curbs, chatter bars, cat's eyes, etc. Extract the preceding vehicle. The image data output from the camera 30 may be monochrome. In this case, brightness values are used during segmentation. The number of cameras 30 may be one or two or more.

A gradient estimation process executed by the road gradient estimation device 100 according to the first embodiment will be described. The processing routine shown in FIG. 5 is executed at predetermined time intervals, for example, in units of several hundred milliseconds, from the time the vehicle control system is started to the time it is stopped, or from the time the start switch is turned on until the start switch is turned off. is executed repeatedly. The gradient estimation process shown in FIG. 5 is executed by the CPU 101 executing the gradient estimation program Pr1.

The CPU 101 acquires an image captured by the camera 30 via the input/output interface 103 (step S100). The CPU 101 determines whether the area of interest on the road can be determined from the captured image (step S102). On the road means the road surface in front of the host vehicle, or the front area where the planned route of the host vehicle can be defined, such as the edge of the road surface or the road shoulder in front of the host vehicle. The region of interest is a region corresponding to an object on the road or at the edge of the road that is useful for estimating the road gradient, and is a target for intensively acquiring detection points by the lidar 20, that is, increasing the number of times detection points are acquired. It is an area. Target objects include, for example, lot lines commonly called white lines, road boundary lines, median strips, guardrails, curbs, chatter bars, cat's eyes, asphalt, and preceding vehicles. Note that, in this embodiment, the partition lines and road boundary lines are hereinafter referred to as white lines. If these objects can be extracted from the captured image, the CPU 101 determines that the region of interest can be determined, and if the objects cannot be extracted, the CPU 101 determines that the region of interest cannot be determined. Note that the extraction of these objects from the captured image may be executed by the image processing unit included in the camera 30 as described above, and the CPU 101 may be notified of whether or not the object can be extracted. Unprocessed image data may be sent to the CPU 101 and executed there. Further, the attention area FA may be determined using an environmental light image acquired by the lidar 20. The specific processing procedure is the same as for the captured image. Note that the environmental light image is a detection result image that does not depend on the detection light from the light emitting unit 24, and will be described in detail later.

If the CPU 101 determines that the region of interest cannot be determined from the captured image (step S102: No), the process proceeds to step S110. This is because if the region of interest cannot be determined, the region to be subjected to high-resolution processing cannot be specified. Cases in which the region of interest cannot be determined include, for example, cases in which the captured image does not include an object, or cases in which the SN of the captured image is so low that the object cannot be determined using pixel values. If the CPU 101 determines that the attention area can be determined from the captured image (step S102: Yes), as shown in FIG. 6, the CPU 101 determines the attention area FA in the captured image CF (step S104). The determination of the attention area FA is the determination of an area as the attention area surrounding the object extracted in the captured image CF, for example, the determination of positional coordinates that can define the area, and the determination of a specific scanning angle range CR. This is a process for More specifically, it may be the position coordinates of multiple points in an area for specifying the contour range of the object, and a rectangular area surrounding the object prepared in advance according to the area of the object on the image. It may also be the position coordinates of the four corners of . In the example of FIG. 6, the white line BL is extracted as the target object, and the attention area FA is defined by a substantially rectangular frame prepared in advance that surrounds the white line BL.

The CPU 101 acquires the detected point group data acquired by the lidar 20, and determines whether high-resolution processing is necessary for the determined region of interest (step S106). Whether or not high-resolution processing is necessary is determined by, for example, whether the detection rate DR of the detection point group data acquired by the lidar 20 is less than or equal to the determination threshold value DRr corresponding to a predetermined reference detection rate. determined by For example, when past detection point cloud data is stored in the memory 102, the detection rate DR is the ratio of the maximum detection distance of the current detection point cloud data to the past detection point cloud data, and the ratio of the maximum detection distance on the road surface. If the past detection point cloud data is not stored in the memory 102, the ratio of the maximum detection distance of the current detection point cloud data to the predetermined standard detection point cloud data is used. The ratio and the ratio of the number of detection points on the road surface are used. Note that the detection rate DR obtained using the standard detection point group data is stored in the memory 102 as past detection point group data. Furthermore, the detection rate may be determined using weather information in addition to or separately from the detection rate that takes past history into account. For example, if the weather is rain or snow, the CPU 101 determines the detection rate DR using weather forecast information as environmental information obtained through communication with the outside, or information on the wiper switch on of the own vehicle. If the detection rate DR is determined to be equal to or less than the threshold value DRr, and the weather is other than rain or snow, the detection rate DR may be determined to be greater than the determination threshold value DRr. Note that as the weather information, information that may affect the detection accuracy of the lidar 20, such as dense fog information or sun altitude information, may be used.

When the CPU 101 determines that the detection rate DR is less than or equal to the determination threshold value DRr, that is, DR≦DRr, the CPU 101 determines that high-resolution processing is necessary (step S106: Yes) and executes the high-resolution processing. (Step S108). In high-resolution processing, the CPU 101 outputs a light emission control signal to the light emission unit 24, that is, the light emission control unit, at a frequency higher than the reference number of times, at an interval shorter than the reference interval, or at a number of times greater than the reference number. . The CPU 101 also instructs the light reception control unit 23 to perform high-level light reception processing using the light reception pixel 221 composed of one light reception element 220 instead of the standard light reception processing using the light reception pixel 222 composed of eight light reception elements 220. Outputs a light reception control signal that executes resolution light reception processing. As a result, for the attention area FA, as shown in the detection frame LF by the lidar 20 in FIG. 7, the number of detection points Dhp in the horizontal direction HD and vertical direction VD is greater than the number of detection points Dnp during normal processing. High-resolution processing that increases the As a result, as shown in FIG. 7, the number of detection points Dhp in the attention area FA is increased, the resolution is increased, and the detection accuracy of the white line BL is improved. As described above, the S level is lowered by using the light receiving pixel 221 composed of one light receiving element 220 in order to increase the number of detection points in the vertical direction VD, but the specific scanning angle range CR By increasing the number of acquisitions of the detection group in , the number of times each light receiving pixel 221 receives light and the number of times it processes light increases while maintaining a constant scanning cycle, so the S level of the incident light intensity signal increases, and as a result, An SN equivalent to the SN during normal processing can be achieved.

If the CPU 101 determines that the detection rate DR is larger than the determination threshold DRr, that is, DR>DRr, the CPU 101 determines that high-resolution processing is unnecessary (step S106: No) and executes normal processing (step S110). In normal processing, the CPU 101 outputs a light emission control signal to the light emitting unit 24 at a standard number of times, and to the light receiving control unit 23, a standard light receiving process using the light receiving pixels 222 composed of eight light receiving elements 220 is performed. Outputs a light reception control signal to execute. As a result, normal processing is realized in which the detection point Dnp shown in FIG. 7 is obtained. Note that the normal processing can also be referred to as reference light reception processing or reference resolution processing.

The CPU 101 executes gradient estimation processing using the detection point group obtained by high-resolution processing or normal processing (step S112), and ends this processing routine. In the example of FIG. 7, the attention area FA is shown taking the white line on the right as an example, but the attention area FA may be similarly determined for the white line on the left, and high-resolution processing may be performed. The slope estimation process is a process for estimating the slope of the road in the traveling direction of the own vehicle. Therefore, on the assumption that the road edges or white lines located on the left and right sides of the vehicle's traveling direction are approximately parallel and the road width is approximately constant, at least one line segment indicating the road edge or road boundary is identified. If the road area is specified, the road area is defined, and if two or more line segments are specified, the accuracy of the road area definition is improved. Specifically, a plane is defined using a group of detection points included in the specified road area, and the gradient is estimated using the defined plane. Gradient estimation using a group of detection points is performed by using a known method, for example, the RANSAC (Random Sample Consensus) method or the least squares method. The plane is defined by ax+by+cz+d=0 (Equation 1). Under this condition, the coefficients a, b, c, and d of Equation 2 are calculated using RANSAC to define the road surface ahead on the three-dimensional coordinates. The normal to the plane defined by the obtained coefficient is determined, and the angle formed by the calculated normal with respect to the reference normal of the predefined reference plane is determined. This allows the slope of the road ahead to be estimated. The reference plane is a horizontal plane on which the vehicle is currently riding, and the gradient of the current road surface on which the vehicle is currently riding is the change in gradient from the state where the gradient of the reference plane and the current road surface are the same, that is, the difference in gradient. This can be corrected by storing it in the memory 102 as a history. Furthermore, the slope estimation accuracy is further improved by determining the slopes of two line segments corresponding to the road edges or white lines on the left and right of the own vehicle. Furthermore, by using the coordinate position information of each detection point forming the left and right line segments, it is also possible to estimate the inclination of the road in the left and right direction.

According to the road slope estimating device 100 according to the first embodiment described above, the attention area FA on the road is determined using the camera 30, and the resolution of the detection point group in the determined attention area FA is increased. A high resolution detection point cloud is acquired. Therefore, the resolution of the detection point obtained from the identified object can be determined by identifying the object used to estimate the slope of the road in front of the vehicle, regardless of the presence or absence of a white line or the clarity of the white line. It is possible to improve the estimation accuracy of the road surface slope. More specifically, by increasing the number of light receiving pixels in the light receiving section 21, that is, increasing the light receiving unit, the number of detection point groups in the vertical direction VD is increased, and the resolution of the vertical direction VD in the attention area FA is increased. . This increase in resolution makes it possible to obtain a group of detection points of interest, particularly a group of detection points with high signal strength. Furthermore, by determining the area of interest FA on the road, that is, in the area in front of the host vehicle using the captured image of the camera 30, it becomes possible to specify a scanning area useful for defining the road surface in front of the host vehicle. , it is possible to increase the number of times a group of detection points is acquired in the specific scanning angle range CR corresponding to the area of interest FA. By increasing the number of acquisitions of the detection point group in the specific scanning angle range CR, the signal strength of the incident light intensity signal of the detection point group is increased without extending the predetermined scanning interval, and the signal strength is increased in the vertical direction VD. It is possible to compensate for the decrease in the intensity of the incident light intensity signal at each detection point due to the increase in resolution, and obtain a specific detection point group with high resolution and sufficient signal intensity. As a result, it is possible to improve the accuracy of extracting a plane, that is, the road surface, by using a high-density detection point group in the vertical direction VD, and it is possible to improve the accuracy of estimating the slope of the road. Furthermore, it is possible to improve the accuracy of estimating the gradient when the road gradient changes. Note that the resolution in the vertical direction VD may be increased by increasing the number of detection points by performing scanning in the vertical direction VD, in addition to increasing the number of light receiving units. By increasing the accuracy of estimating the road slope, it becomes possible to estimate the road slope even for a road where a white line exists only on one side, and it is possible to improve the accuracy of calculating the curvature of the white line. Further, it is possible to appropriately control the prime mover in the vehicle according to the road gradient. For example, in a so-called hybrid vehicle that includes an internal combustion engine and an electric motor, power distribution between the internal combustion engine and the electric motor can be appropriately controlled.

In the road slope estimating device 100 according to the first embodiment, in addition to increasing the resolution in the vertical direction VD, the resolution of the detection point group in the horizontal direction HD may be increased. As a result, it is possible to further improve the accuracy of extracting a plane, that is, the road surface, by using a high-density detection point group in the vertical direction VD and horizontal direction HD, and it is possible to further improve the estimation accuracy of the road slope. . Furthermore, it is possible to further improve the accuracy of estimating the gradient when the road gradient changes. The resolution of the detection point group in the horizontal direction HD can be increased, for example, by transmitting the rotation angle instruction signal to the motor driver more times than during normal processing, that is, by reducing the unit scan angle. Furthermore, the above-described process may be performed by replacing the vertical direction and the horizontal direction in the first embodiment. That is, the lidar 20 is arranged in a state rotated by 90 degrees vertically and horizontally, that is, the light receiving element array 22 shown in FIG. 3 is arranged in a state rotated by 90 degrees. The road gradient estimating device 100 may be implemented by reading the term "vertical direction" as the vertical direction.

In the gradient estimation process described above, the gradient was estimated using a plane. On the other hand, the gradient may be estimated using line segments. In this embodiment, since the resolution for detecting the line segment corresponding to the white line is increased, it is possible to define the line segment with high precision, even when a single line segment is used. Gradient estimation accuracy is improved. Specifically, using the three-dimensional coordinate information (x, y, z) of each detection point constituting the identified line segment, if the z-axis is the coordinate axis in the vertical direction VD in the coordinate system of the own vehicle, The increment or decrement of the z-coordinate of each detection point can be used to determine the slope of the line segment. In order to find a straight line from a group of points, the RANSAC method and the least squares method described above are used. A line segment is defined by ax+by+c=0 (Equation 1). Under this condition, RANSAC is used to calculate the coefficients a, b, and c of Equation 1, and a line segment indicating the road edge or white line on the road ahead is defined. The gradient of the road surface ahead is estimated by determining the angle formed by the defined line segment with respect to a predefined reference plane.

According to the road gradient estimating device 100 according to the first embodiment, the road gradient is estimated using both the camera 30 and the lidar 20, so even if a problem occurs in one, the road gradient can still be estimated. can be continued.

In the above description, as a specific detection point group, the resolution in the vertical direction VD is increased, and the number of times the lidar 20 acquires the detection point group is increased in the specific scanning angle range CR corresponding to the area of interest FA to increase the resolution. A high-resolution detection point cloud was obtained by compensating for the accompanying decrease in incident light intensity, and the road slope was estimated. In contrast, the signal intensity of the incident light intensity signal is increased by increasing the number of times the lidar 20 acquires a group of detection points in the specific scanning angle range CR corresponding to the region of interest without increasing the resolution in the region of interest. The road slope may be estimated by acquiring a group of high signal strength detection points as a group of detection points of interest. That is, an increase in the number of times the lidar 20 acquires a group of detection points in the specific scanning angle range CR corresponds to the acquisition of a group of detection points of interest, particularly a group of high signal strength detection points. In this case, even if the light receiving element 220 that cannot increase the resolution in the vertical direction VD is used, or the electric motor 40 or the scanning mirror 26 that cannot increase the resolution in the horizontal direction HD, It is possible to improve the estimation accuracy of road slope. In other words, by increasing the signal strength of the incident light intensity signal indicating the detection point group corresponding to a target object such as a white line, the detection accuracy of the target object such as a white line is improved, the extraction accuracy of the road surface is improved, and the road slope The estimation accuracy is improved.

In the first embodiment, when high-resolution light reception processing is executed, the scanning mirror 26 may be repeatedly moved back and forth in the specific scanning angle range CR. In this way, the scanning mirror 26 repeats scanning within the specific scanning angle range CR, thereby making it possible to increase the signal strength of each detection point forming the detection point group in the specific scanning angle range CR.

Second embodiment:
In the first embodiment, the gradient estimation program Pr1 estimates the gradient of the road surface using a group of high-resolution detection points obtained in the region of interest FA, but in the second embodiment, the gradient estimation program Pr1 When the attention area FA cannot be determined, or when the attention area FA can only be determined in the vicinity and a high-resolution detection point cloud cannot be obtained, in addition to the normal detection point cloud, the ambient light obtained by the lidar 20 is used. The system is configured to use images in a complementary manner to estimate the slope of the road surface. Generally, there are cases where the reflection point of the rider 20 can be detected only in the vicinity of the road surface, and even in such cases, the ambient light image is used. Note that the configuration of the road gradient estimating device according to the second embodiment is similar to the configuration of the road gradient estimating device 100 according to the first embodiment, so the same reference numerals are given and the explanation will be omitted. Generally, the lidar 20 detects and acquires a group of detection points by emitting detection light from the light emitting unit 24 and by having the light receiving unit 21 receive reflected light corresponding to the detection light. The light receiving unit 21 identifies and detects the presence of a detection point in response to input of an incident signal intensity exceeding a predetermined threshold, but environmental light, also called disturbance light, also enters the light receiving unit 21 at any time. has been done. Since the lidar 20 receives incident light in the infrared region, which is different from the visible light region detected by the camera 30, the result of reception of environmental light by each light receiving pixel forms an infrared pixel image, as shown in FIG. It can be used as an ambient light image LF. In FIG. 8, the ambient light image LF includes a white line BL, and a road or road surface defined by the white line BL can be extracted.

The road slope estimating device 100 may use the environmental light image when the intensity of the environmental light is higher than a predetermined threshold. The environmental light image LF is a detection result image that does not depend on the detection light from the light emitting unit 24, and in shadow areas at night or during the day, it is not possible to obtain an environmental light image that satisfies the desired reliability, and the result is Therefore, accurate gradient estimation processing that complements the detected point group cannot be performed. The intensity of the environmental light is compared with a predetermined threshold using, for example, the maximum value or average value of the signal intensity received by the light receiving unit 21. When the intensity of the environmental light is stronger than a predetermined threshold, the road slope estimating device 100 executes the slope estimation process using the environmental light image in addition to the detection point group, and calculates the intensity of the environmental light. If it is weaker than a predetermined threshold, gradient estimation processing is performed using only the detection point group without using the environmental light image. Note that the gradient of the road surface in the environmental light image LF is obtained by calculating the extracted normal to the road surface and using the method described above. The road slope estimating device 100 can obtain the slope by averaging or weighted average using, for example, the slope obtained using the detection point group and the slope obtained using the environmental light image. In this case, even if the area of interest FA cannot be identified by the camera 30 and high-resolution processing cannot be performed, the slope of the road can be estimated with high accuracy.

Furthermore, as shown in FIG. 9, for a distant road surface area where the detection point group Dnp cannot be obtained by the lidar 20, the extracted road surface ESL obtained from the environmental light image LF is used to extrapolate the road surface. A gradient may also be estimated. As a result, it is possible to estimate the slope of a distant road surface where reflected light from the road surface that responds to the detected light cannot be obtained. Note that extrapolation is performed when the slope of the plane defined by the obtained detection point group Dnp and the slope of the extracted road surface ESL approximate or match. This ensures continuity between the road surface in front and the road surface in the distance.

Third embodiment:
The road slope estimation device according to the third embodiment is different from the first embodiment in that the slope estimation program Pr1 determines the attention area FA using the history of the attention area FA determined in the past when determining the attention area FA. This is different from the road gradient estimating device 100 according to the embodiment. Note that the other configurations of the road gradient estimating device according to the third embodiment are the same as those of the road gradient estimating device 100 according to the first embodiment, so the same reference numerals are given and the explanation will be omitted.

The gradient estimation process using the history of the attention area FA will be described in detail with reference to FIGS. 10 and 11. The attention area FA determined in the past is, for example, the previous attention area FAp determined using a previous frame image CFp temporally past the current frame image CFc in a frame image obtained by imaging with the camera 30. be. Note that the frame rate in the camera 30 is about 60 fps, and the time difference between the previous frame image CFp and the current frame image CFc is several tens of milliseconds. The road slope estimating device 100 according to the third embodiment determines the current area of interest FAc in the current frame image CFc using the previous area of interest FAp. Using the previous region of interest FAp means using the coordinate position information that defines the previous region of interest FAp, and by using the coordinate position information of the previous region of interest FAp in the current frame image CFc, The process of determining the current area of interest FAc can be started in the area where the white line BL exists or the approximate area of the white line BL, without scanning all pixels or the entire area. As a result, the amount of arithmetic processing for determining the attention area FA can be reduced, and the time required to determine the attention area FA can be reduced. Note that the past frame image CFp is obtained by being stored in the memory 102.

Fourth embodiment:
The road slope estimation device according to the fourth embodiment is different from the first embodiment in that the slope estimation program Pr1 determines the attention area FA using the history of objects detected in the past when determining the attention area FA. This is different from the road slope estimating device 100 according to the present invention. Note that the other configurations of the road gradient estimating device according to the fourth embodiment are the same as those of the road gradient estimating device 100 according to the first embodiment, so the same reference numerals are given and the explanation will be omitted.

With reference to FIG. 12, gradient estimation processing using the history of objects detected in the past will be described in detail. Note that the processing routine shown in FIG. 12 is repeatedly executed at predetermined time intervals in the same manner as the processing routine shown in FIG. The object detected in the past is, for example, an object on the road that was detected using the previous frame image CFp, which is temporally past the current frame image CFc, in the frame image obtained by imaging with the camera 30. . The road slope estimating device 100 according to the fourth embodiment, that is, the CPU 101 determines whether there is a history of the object (step S200). The object detection history is stored as object history OR in the area 102a of the memory 102, and information regarding detected objects, such as coordinate position information and shape information, is stored. If the detection of the target object is not recorded in the target object history OR (step S200: No), the CPU 101 moves to step S206. If detection of an object is recorded in the object history OR (step S200: Yes), the CPU 101 searches for the same object as the recorded history object (step 202). Specifically, the CPU 101 extracts the relevant object from the current frame image CFc using the coordinate position information and shape information stored in the object history OR. If the search is successful, that is, if an object that matches a previously detected object is detected in the current frame image Cfc (step S204: Yes), the CPU 101 uses the detected object to draw attention. It is decided to determine the area FA, and this processing routine is ended.

If a target object matching a previously detected target object is not detected (step S204: No), the CPU 101 moves to step S206. The CPU 101 executes a white line search for the current frame image CFc (step S206), determines to use the white line to determine the attention area FA, and ends this processing routine.

According to the road gradient estimating device 100 according to the fourth embodiment, it is possible to extract a target object from the current frame image CFc and determine the attention area FA using information on targets detected in the past. By using objects detected in the past, it becomes possible to suppress or prevent false detection of objects in the frame images of the camera 30, and also the time required to specify the object and determine the attention area FA. The amount of calculation processing and time can be reduced.

Fifth embodiment:
The road gradient estimation device according to the fifth embodiment is different from the road gradient estimation device 100 according to the first embodiment in that the gradient estimation program Pr1 increases the resolution in the vertical direction of the lidar as the distance from the vehicle 50 increases. different. Note that the other configurations of the road gradient estimating device according to the fifth embodiment are the same as those of the road gradient estimating device 100 according to the first embodiment, so the same reference numerals are given and the explanation will be omitted.

As shown in FIG. 13, in the conventional rider 20p, the density of detection points on the road surface in the vertical direction of the rider 20p decreases as the distance from the vehicle 50 increases. That is, when the vertical direction of the vehicle 50 is set to a vertical angle of 0°, as the vertical angle increases, the density of detection points of the rider 20p on the road surface in the vertical direction becomes lower, or the detection interval becomes wider, and the resulting detection result becomes smaller. Resolution decreases. For example, in the vicinity of the vehicle 50, that is, in an area where the vertical angle is small, the intervals between the detection points are narrow, and the distance from the rider 20p to the road surface is also short, so it is easy to obtain reflected light. On the other hand, in a region far away from the vehicle 50, that is, in a region where the vertical angle is large, the intervals between the detection points are wide and the distance from the rider 20p to the road surface is long, so it is difficult to obtain reflected light. As a result, as the distance from the vehicle 50 increases, the distance between the detection points increases and the detection point density decreases. Therefore, as shown in FIG. 14, the CPU 101 increases the resolution of the lidar 20 in the vertical direction as it moves away from the vehicle, and acquires a high-resolution detection point group as a detection point group of interest. Specifically, as shown in FIG. 15, the light reception control unit 23 controls whether the light reception pixels constituted by the light reception elements 220 arranged in the direction corresponding to the vertical direction of the lidar 20 in the light reception element array 22 are Light reception control is performed so that the number of light receiving elements 220 decreases as the distance from the light receiving element 50 increases. In the example of FIG. 15, the light receiving pixel 22e corresponding to the position closest to the vehicle 50 is composed of five light receiving elements 220, and the light receiving pixel 22a corresponding to the position farthest from the vehicle 50 is composed of one light receiving element 220. ing. From the light receiving element 22e toward the light receiving element 22a, there are a light receiving pixel 22d constituted by four light receiving elements 220, a light receiving pixel 22c constituted by three light receiving elements 220, a light receiving pixel 22b constituted by two light receiving elements 220, etc. The number of light-receiving elements constituting a light-receiving pixel is gradually reduced. That is, the number of light receiving elements constituting the light receiving pixels in the light receiving element array 22 is set so that the interval D1 between detection points is constant regardless of the distance from the vehicle 50. According to the fifth embodiment, the resolution increases in the direction away from the vehicle 50, so it is possible to obtain the same resolution regardless of the distance from the vehicle 50. Object detection accuracy can be maintained regardless of the distance from the object. Note that the number of light-receiving elements 220 constituting a light-receiving pixel is an example, and any number may be used as long as the number of light-receiving elements decreases as the distance from the vehicle 50 increases. In the above description, by reducing the number of light-receiving elements constituting the light-receiving pixels in the light-receiving element array 22 as the distance from the vehicle increases, the resolution in the vertical direction increases as the distance from the vehicle increases. A resolution detection point cloud has been acquired. On the other hand, when the lidar 20 that can scan vertically is used, the light-receiving pixels in the light-receiving element array 22 are not adjusted, and the unit scanning angle is decreased as the distance from the vehicle increases. The resolution in the vertical direction may be increased accordingly, and a high-resolution detection point group may be acquired as the detection point group of interest.

Other embodiments:
(1) An application example using the gradient estimated in the first to fifth embodiments will be described. For example, by obtaining the slope of the road in the direction of travel of the host vehicle with high precision in advance, the throttle opening and gears can be adjusted to suppress or prevent a decrease in vehicle speed if the slope is up, or if the slope is down. For example, it is possible to suppress or prevent an increase in vehicle speed.

(2) When extracting the target object,
- The captured image of the camera 30 and the environmental light image of the lidar 20 are used to extract white lines, median strips, and asphalt,
- The image captured by the camera 30, the environmental light image of the rider 20, and the detection point of the rider 20 are used to extract the guardrail and curb,
・For extraction of chatter bars and cat's eyes, strong reflection detection points of lidar 20 are used,
- Extraction of the preceding vehicle includes an image captured by the camera 30, an ambient light image of the lidar 20, a detection point of the lidar 20, and a detection point of the millimeter wave radar.

(3) In each of the above embodiments, the CPU 101 executes the gradient estimation program Pr1, thereby realizing the road gradient estimation device 100 that determines the attention area FA using software, executes high-resolution processing, and estimates the gradient. However, it may also be implemented in hardware by pre-programmed integrated circuits or discrete circuits. That is, the control unit and its method in each of the above embodiments are implemented by a dedicated computer provided by configuring a processor and memory programmed to execute one or more functions embodied by a computer program. May be realized. Alternatively, the controller and techniques described in this disclosure may be implemented by a dedicated computer provided by a processor configured with one or more dedicated hardware logic circuits. Alternatively, the control unit and the method described in the present disclosure are configured by a combination of a processor and memory programmed to perform one or more functions and a processor configured by one or more hardware logic circuits. may be implemented by one or more dedicated computers. The computer program may also be stored as instructions executed by a computer on a computer-readable non-transitory tangible storage medium.

Although the present disclosure has been described above based on the embodiments and modifications, the embodiments of the invention described above are for facilitating understanding of the present disclosure, and do not limit the present disclosure. This disclosure may be modified and improved without departing from its spirit or the scope of the claims, and this disclosure includes equivalents thereof. For example, the embodiments corresponding to the technical features in each form and the technical features in the modifications described in the column of the summary of the invention may be used to solve some or all of the above-mentioned problems, or to solve the above-mentioned problems. In order to achieve some or all of the effects, it is possible to replace or combine them as appropriate. Further, unless the technical feature is described as essential in this specification, it can be deleted as appropriate.

DESCRIPTION OF SYMBOLS 10... Road gradient estimation system, 20... Rider, 30... Camera, 40... Electric motor, 100... Road gradient estimation device, 101... CPU, 102... Memory, 103... Input/output interface, 104... Internal bus, 50... Vehicle, Pr1 ...Gradient estimation program.

Claims (14)

A road gradient estimation device (100) used in a vehicle (50), comprising:
an attention area determination unit (101) that determines an attention area (FA) on a road using an ambient light image or a captured image acquired from a lidar (20) or an imaging device (30);
A detection point group acquisition unit (101) that controls a lidar to acquire a detection point group, and is capable of acquiring a detection point group of interest by increasing the resolution or signal strength of the detection point group in the region of interest. a detection point group acquisition unit that is a group acquisition unit and increases the vertical or horizontal resolution of the lidar to acquire a high-resolution detection point group as the target detection point group;
comprising a slope estimating unit (101) that estimates a road slope using the group of detection points of interest ;
The lidar has a light receiving element array (22) composed of a plurality of light receiving elements (220),
The detection point group acquisition unit acquires the detection point group using a predetermined number of the plurality of light receiving elements as a light receiving unit,
The detection point cloud acquisition unit includes:
A road slope estimating device that increases vertical or horizontal resolution by decreasing the number of light receiving elements constituting the light receiving unit and increasing the number of the light receiving units . A road gradient estimation device (100) used in a vehicle (50), comprising:
an attention area determination unit (101) that determines an attention area (FA) on a road using an ambient light image or a captured image acquired from a lidar (20) or an imaging device (30);
A detection point group acquisition unit (101) that controls a lidar to acquire a detection point group, and is capable of acquiring a detection point group of interest by increasing the resolution or signal strength of the detection point group in the region of interest. a group acquisition unit;
comprising a slope estimating unit (101) that estimates a road slope using the group of detection points of interest;
a vertical direction of the rider corresponds to a direction away from the vehicle;
The detection point group acquisition unit is a road gradient estimating device that increases the resolution of the lidar in the vertical direction as the distance from the vehicle increases, and acquires a high-resolution detection point group as the detection point group of interest. The road slope estimating device according to claim 2 ,
The lidar has a light-receiving element array (22) composed of a plurality of light-receiving elements (220) arranged corresponding to a direction away from the vehicle,
The detection point group acquisition unit acquires the detection point group using a predetermined number of the plurality of light receiving elements as a light receiving unit,
The detection point cloud acquisition unit includes:
A road slope estimating device that increases resolution in the vertical direction by decreasing the number of the light receiving elements constituting the light receiving unit and increasing the number of the light receiving units as the distance from the vehicle increases. The road slope estimating device according to claim 1 or 3 ,
The detection point group acquisition unit further increases the number of times the detection point group is acquired per unit scanning angle in the region of interest. The road slope estimating device according to claim 1,
The detection point group acquisition unit is a road slope estimating device that increases the signal strength of the detection point group by increasing the number of times the detection point group is acquired per unit scanning angle in the region of interest. The road slope estimating device according to any one of claims 1 to 5 ,
The detection point group acquisition unit is a road gradient estimating device that acquires the detection point group of interest when a detection rate of the detection point group is equal to or lower than a predetermined reference detection rate. The road slope estimating device according to claim 6 ,
The detection point cloud acquisition unit determines, as the detection rate, at least one of a ratio of a maximum detection distance in current detection point cloud data to reference detection point cloud data or past detection point cloud data, and a ratio of the number of detection points on the road. A road slope estimation device that uses either one of the two. The road slope estimating device according to any one of claims 1 to 7 ,
The captured image is formed by a plurality of frame images,
The attention area determining unit determines the current attention area (FAc) using a previous attention area (FAp) determined using a previous frame image (CFp) that is older than the current frame image (CFc). Road gradient estimation device. A road gradient estimation device (100) used in a vehicle (50), comprising:
An attention area determination unit that determines an attention area (FA) on a road using an ambient light image acquired from a lidar (20) or an imaging device (30) or a captured image formed by a plurality of frame images. (101), in which the current area of interest (FAc) is determined using the previous area of interest (FAp) determined using the previous frame image (CFp) that is past the current frame image (CFc). an area determining section;
A detection point group acquisition unit (101) that controls a lidar to acquire a detection point group, and is capable of acquiring a detection point group of interest by increasing the resolution or signal strength of the detection point group in the current region of interest. a detection point cloud acquisition unit;
comprising a slope estimating unit (101) that estimates a road slope using the group of detection points of interest;
The attention area determination unit includes:
If an object on the road is detected in the previous frame image, detecting the detected object on the road from the current frame image to determine the current region of interest;
A road gradient estimating device that detects a marking line (BL) from the current frame image to determine the current region of interest when the object on the road is not detected. The road slope estimating device according to any one of claims 1 to 9 ,
When the region of interest cannot be determined using the captured image, or when the region of interest can only be determined in the vicinity of the region of interest, the gradient estimator is configured to estimate the road using the detection point group and the environmental light image. A road slope estimation device that estimates slope. A road slope estimation system (10) mounted on a vehicle (50), comprising:
A road slope estimating device (100) according to any one of claims 1 to 10 ,
the imaging device;
The rider;
A road gradient estimation system. A road gradient estimation method in a vehicle (50), comprising:
An area of interest (FA) on the road is determined using a captured image or an ambient light image acquired from an imaging device (30) or a lidar, and the lidar is a light-receiving area composed of a plurality of light-receiving elements (220). having an element array (22);
By decreasing the predetermined number of the light-receiving elements constituting a light-receiving unit and increasing the number of light-receiving units, the vertical or horizontal resolution of the lidar is increased and detection in the region of interest is performed. Obtain a high-resolution detection point cloud as a target detection point group with increased resolution of the point cloud,
A road gradient estimation method, comprising: estimating a road gradient using the group of detection points of interest. A road slope estimation method in a vehicle (50), comprising:
determining an area of interest (FA) on the road using a captured image or an ambient light image acquired from an imaging device (30) or lidar;
The vertical direction of the lidar corresponds to a direction moving away from the vehicle, and the resolution of the lidar in the vertical direction increases as the lidar moves away from the vehicle, thereby increasing the resolution of the detection point group in the region of interest. Obtain a high-resolution detection point cloud as a group ,
A road gradient estimation method, comprising: estimating a road gradient using the group of detection points of interest. A road gradient estimation method in a vehicle (50), comprising:
determining an area of interest (FA) on the road using an ambient light image acquired from an imaging device (30) or lidar or a captured image formed by a plurality of frame images ;
A current area of interest (FAc) is determined using a previous area of interest (FAp) determined using a previous frame image (CFp) that is past the current frame image (CFc), and the current area of interest (FAc) is ) is determined by
If an object on the road is detected in the previous frame image, detecting the detected object on the road from the current frame image to determine the current region of interest;
If the object on the road is not detected, detecting a lane marking (BL) from the current frame image to determine the current region of interest;
increasing the resolution or signal strength of the detection point group in the current region of interest to obtain a detection point group of interest;
A road slope estimation method, comprising: estimating a road slope using the group of detection points of interest.

Images