• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    APPLIANCE AND ROUTE ESTIMATE PROGRAM. The present invention relates to a feature point extraction section that acquires a captured image from an image capture device and extracts feature points from the captured image, from a vehicle track boundary point selection section that selects vehicle lane limit points that indicate vehicle lanes from the extracted characteristic points, a distribution determination section that determines the distribution of vehicle lane limit points, and a route parameter estimation section that if stable, it predicts the parameters of the route based on the vehicle's lane limit points, the results of the past estimate and the noise of the system that was established.
    公开号:BR112013006124B1
    申请号:R112013006124-3
    申请日:2011-09-26
    公开日:2021-03-09
    发明作者:Akihiro Watanabe；Takahiro Kojo；Theerawat Limpibunterng；Yoshiaki Tsuchiya
    申请人:Toyota Jidosha Kabushiki Kaisha；
    IPC主号:
    专利说明:

    Technical Field
    [0001] The present invention relates to a device and program for estimating the route, and in particular, with a device and program for estimating the route that estimates parameters of the route based on an image captured by an image capture device. Background Technique
    [0002] Conventionally, vehicle path recognition devices have been proposed that detect horizontal road signs in an entrance image captured in front of a vehicle by a CCD camera, and based on the results of the detection of horizontal road signs, calculate model parameters of the stay using a Kalman filter in order to represent the shape of the road in front of the vehicle (see Japanese Exposed Patent Application (JP-A) No 2002-109695). In the JP-A route recognition apparatus No 2002-109695, changes in the road model parameters are treated as having a probabilistic nature, and a random route model guided by a fixed Gaussian white noise is defined. DESCRIPTION OF THE INVENTION Technical Problem
    [0003] The accuracy of the estimation of each of the parameters to be estimated is affected by the distribution of the observation values, however, there is a problem with JP-A technology No 2002-109695 due to the fact that the system noise expressing the degree of variation in the road model parameters is established regardless of the observation values, so that a stable estimate of the road model parameter is not possible.
    [0004] The present invention addresses the above problem, and an objective is to provide a device and program for estimating the path that can estimate the parameters of the path in a stable way. Solution to the Problem
    [0005] In order to achieve the above objective, the travel estimation apparatus of the present invention includes: a acquisition section to acquire a captured image from a vehicle periphery; an extraction section to extract, from the captured image acquired by the acquisition section, characteristic points indicating the vehicle's tracks; an establishment section, based on a distribution of the characteristic points extracted by the extraction section, to establish the system noise expressing the variation of the path parameters when estimating the path parameters related to a position or an angle of the vehicle itself with respect to a route for travel by the vehicle itself and related to a shape or size of the route; and an estimation section to estimate the parameters of the path by processing the probability signal using a separate time signal based on the characteristic points extracted by the extraction section, in a previous result of the estimation of the parameters of the path and in the noise of the system established by the establishment section.
    [0006] According to the route estimation apparatus of the present invention, the acquisition section acquires an image that was captured from a vehicle periphery, and the extraction section extracts characteristic points indicating the vehicle's tracks from the image. acquired by the acquisition section. The extraction of the characteristic points indicating the vehicle's tracks is performed by first extracting edge points from the captured image, and then, selecting the characteristic points that indicate the vehicle's tracks from the edge points based on factors such as the shape and continuity of border points.
    [0007] Next, the establishment section establishes the noise of the system expressing the variation of the parameters of the route when estimating the parameters of the route related to the position or the angle of the vehicle itself with respect to a route for displacement by the vehicle itself and the shape or size of the route based on a distribution of the characteristic points extracted by the extraction section. The estimation section then estimates the parameters of the path by probability signal processing using the separate time signal based on the characteristic points extracted by the extraction section, the result of past estimate of the path parameters, and the noise of the system that was established by the establishment section.
    [0008] Consequently, it is possible to estimate the parameters of the route in a stable way since the noise of the system is established corresponding to each of the parameters of the route for estimation based on the distribution of characteristic points that indicate the tracks of the vehicle extracted from the captured image, in other words, based on the distribution of observation values.
    [0009] The configuration can be made so that the path parameters related to the position and the angle of the vehicle itself with respect to the course include a lateral position of the vehicle itself with respect to the course, a yaw angle with respect to a center line of the course and an angle of inclination with respect to a course plane, and where course parameters related to the shape and size of the course include a radius of curvature of the course and a track of the vehicle of the course.
    [00010] Configuration can be made so that: when the characteristic points are only distributed in a distant region in the captured image, the establishment section reduces the noise of the system corresponding to the radius of curvature of the route, the width of the vehicle's lane. course, and the lateral position of the vehicle itself with respect to the course; when the characteristic points are only distributed in a region close to the captured image, the establishment section reduces the noise of the system corresponding to the radius of curvature of the path; when the characteristic points are in a distribution only expressing a limit on the left side of the vehicle lane or only expressing a limit on the right side of the vehicle lane, the establishment section decreases system noise corresponding to the width of the vehicle lane and the angle of inclination with respect to the plane of the course; and when the number of characteristic points present is the same or less than a specific predetermined number, the establishment section decreases the noise of the system corresponding to all the parameters of the route.
    [00011] Additionally, the route estimation program of the present invention is a program that causes a computer to perform the functions of: an acquisition section to acquire an image captured from a vehicle's periphery; an extraction section to extract, from the captured image acquired by the acquisition section, characteristic points indicating the vehicle's tracks; an establishment section for, based on a distribution of the characteristic points extracted by the extraction section, to establish the system noise expressing the variation of the path parameters when estimating the path parameters related to a position or an angle of the vehicle itself with respect a route for travel by the vehicle itself and related to a shape or size of the route; and an estimation section to estimate the path parameters by processing the probability signal using a separate time signal based on the characteristic points extracted by the extraction section, in a previous result of the estimate of the path parameters and in the system noise established by the establishment section. Advantageous Effects of the Invention
    [00012] According to the apparatus and program for estimating the path of the present invention as described above, the advantageous effect of being able to stably estimate path parameters is obtained, due to the establishment of the system noise corresponding to each parameter of the system. route that is to be estimated based on the distribution of characteristic points that indicate the vehicle tracks extracted from a captured image, namely, based on the distribution of observation values. BRIEF DESCRIPTION OF THE DRAWINGS
    [00013] Fig. Is a block diagram showing relevant parts of an electrical system of a travel estimation apparatus according to an embodiment of the present invention.
    [00014] Fig. 2A is a diagram schematically illustrating the parameters of the course to be estimated (lateral position, yaw angle, vehicle lane width).
    [00015] Fig. 2B is a diagram schematically illustrating a path parameter to be estimated (radius of curvature).
    [00016] Fig. 2C is a diagram schematically illustrating a parameter of the course to be estimated (inclination angle).
    [00017] Fig. 3 is a block diagram showing a functional configuration of a route estimation apparatus according to an embodiment of the present invention.
    [00018] Fig. 4A is an explanatory diagram to illustrate the extraction of the characteristic points.
    [00019] Fig. 4B is an explanatory diagram to illustrate the extraction of characteristic points.
    [00020] Fig. 4C is an explanatory diagram to illustrate the extraction of characteristic points.
    [00021] Fig. 5 is an explanatory diagram to illustrate the selection of vehicle lane limit points.
    [00022] Fig. 6 is an explanatory diagram to illustrate a distant and a close region.
    [00023] Fig. 7A is a diagram showing a distribution pattern of the vehicle's lane limit points.
    [00024] Fig. 7B is a diagram showing a distribution pattern of the vehicle's lane limit points.
    [00025] Fig. 7C is a diagram showing a distribution pattern of the vehicle's lane limit points.
    [00026] Fig. 7D is a diagram showing a distribution pattern of the vehicle's lane limit points.
    [00027] Fig. 7E is a diagram showing a distribution pattern of the vehicle's lane limit points.
    [00028] Fig. 7F is a diagram showing a distribution pattern of the vehicle's lane limit points.
    [00029] Fig. 7G is a diagram showing a distribution pattern of the vehicle's lane limit points.
    [00030] Fig. 7H is a diagram showing a distribution pattern of the vehicle's lane limit points.
    [00031] Fig. 7I is a diagram showing a distribution pattern of the vehicle's lane limit points.
    [00032] Fig. 7J is a diagram showing a distribution pattern of the vehicle's lane limit points.
    [00033] Fig. 7K is a diagram showing a distribution pattern of the vehicle's lane limit points.
    [00034] Fig. 7L is a diagram showing a distribution pattern of the vehicle's lane limit points.
    [00035] Fig. 7M is a diagram showing a distribution pattern of the vehicle's lane limit points.
    [00036] Fig. 7N is a diagram showing a distribution pattern of the vehicle's lane limit points.
    [00037] Fig. 7O is a diagram showing a distribution pattern of the vehicle's lane limit points.
    [00038] Fig. 7P is a diagram showing a distribution pattern of the vehicle's lane limit points.
    [00039] Fig. 8 is a table showing an example of system noise parameters for each path parameter according to the distribution of the vehicle's track limit points.
    [00040] Fig. 9 is a flowchart showing the content of a route estimation processing routine of a route estimation apparatus according to an embodiment of the present invention.
    [00041] Fig. 10 is an explanatory diagram illustrating an example using a particle filter. BEST MODE FOR CARRYING OUT THE INVENTION
    [00042] The following is the detailed explanation with respect to an illustrative embodiment of the present invention with reference to the drawings.
    [00043] As shown in Fig. 1, a route estimation apparatus 10 of the present illustrative embodiment includes an image capture device 12 for successively capturing images of a region in front of a vehicle, and a computer 16 for performing processing for estimate route parameters.
    [00044] Image capture device 12 includes an image capture section (not shown in the drawings) to capture an image of a target region in front of a vehicle and generate an image signal, an A / D converter section ( not shown in the drawings) to convert the analog image signal generated by the image capture section to a digital signal, and an image memory (not shown in the drawings) to temporarily store the converted A / D image signal.
    [00045] Computer 16 is configured including a CPU 20 that performs general controls of the route estimation apparatus 10; ROM 22 serving as a storage medium in which several programs are stored, such as a program for a route estimation processing routine, described later; RAM 24 that serves as a work area for temporarily storing data; a memory 26 serving as a storage section stored with various types of data; an input-output (I / O) port 28; and a bus that interconnects the sections above. The image capture device 12 is connected to the I / O port 28.
    [00046] The route estimation apparatus 10 of the present illustrative embodiment extracts characteristic points indicating the vehicle lanes (vehicle lane limit points) from images captured by the image capture device 12, and estimates parameters meters of the route using these characteristic points as observation values for employing a Kalman filter.
    [00047] The route parameters relating to the position and angle of the vehicle itself in relation to the route on which the vehicle itself is moving, and the route parameters relating to the shape and size of the route on which the vehicle itself is traveling. vehicle is moving are estimated as parameters of the route. More specifically, the parameters of the route relating to the position and the angle of the vehicle itself in relation to the route are taken as: a lateral position ek | k of the vehicle itself in relation to a vehicle lane as indicated by the lateral limit along the route, a vehicle lane as indicated by the right side limit, and a center line; a yaw angle θk | k relative to the center line of the course, and a tilt angle Φk | k relative to a course plane. The parameters of the route relating to the shape and size of the route are taken as the radius of curvature ck | k of the route and the width of the lane of the vehicle wk | k of the route. When these 5 parameters are referred to collectively as path parameters they are called path parameters xk | k (xk | k = (ek | k θk | k Φk | k ck | k Wk | k)). Fig. 2A to Fig. 2C schematically illustrate the lateral position, yaw angle, angle of inclination, radius of curvature (shape of the path) and width of the vehicle tracks that are to be estimated as parameters of the path.
    [00048] If computer 16 to perform such processing is described in terms of functional blocks divided by each execution section for each function determined by hardware and software, as shown in Fig. 3, computer 16 can then be represented as a configuration including: a feature point extraction section 30 to acquire a captured image that has been captured by the image capture device 12 and extract feature points from the captured image; a vehicle lane limit point selection section 32 for selecting vehicle lane limit points indicating vehicle lanes from the extracted characteristic points; a distribution determination section 34 for determining the distribution of vehicle lane limit points; a system noise establishment section 36 to establish the respective system noises based on the distribution of the vehicle lane limit points; and a section 38 parameter estimation section to estimate the parameters of the route based on the vehicle's lane limit points, past estimation results and system noises that have been established.
    [00049] The characteristic point extraction section 30, for example, extracts as characteristic points from a captured image, as shown in Fig. 4A, edge points which are points where the brightness of each pixel is changes when scanning in a horizontal direction, as shown in Fig. 4B. An example of extracted characteristic points is schematically illustrated in Fig. 4C.
    [00050] The vehicle lane limit point selection section 32 selects from the characteristic points extracted by the vehicle lane limit point extraction section 30 illustrating vehicle lanes by determining factors such as the shape, width and color of the border points that are aligned in a continuous row. When multiple vehicle lanes are present, the vehicle lane limit points representing the left and right innermost pair are selected. Fig. 5 schematically illustrates an example of the vehicle's lane limit points selected.
    [00051] The characteristic point extraction section 30 and the vehicle lane limit point selection section 32 are examples of an extraction section of the present invention.
    [00052] The distribution determination section 34 determines which type of distribution the vehicle lane limit points extracted by the vehicle lane limit point selection section 32 does. In the present illustrative embodiment, a determination is made as to whether the vehicle's lane boundary points are distributed both in a distant and nearby region, whether they are distributed only in a distant region, or whether they are distributed only in a nearby region. A determination is also made as to whether this is a distribution in which both the vehicle's lane boundary points expressing a left-hand lane limit and the vehicle's lane boundary points expressing a right-hand boundary are present. it is a distribution in which only vehicle lane limit points expressing a left side limit are present, or if this is a distribution in which only vehicle lane limit points expressing a right side limit are present. A determination is also made as to whether or not the total number of vehicle lane limit points selected is the same or less than a specific predetermined number. Distributions in which the total number of vehicle lane limit points is the specific number or less are referred to as null observation value distributions.
    [00053] The following explanation concerns the determination of whether or not there are vehicle lane limit points present in a distant region and / or in a nearby region. First, as shown in Fig. 6, a geometric axis x in the horizontal direction and a geometric axis y in the vertical direction are determined with their origins in the pixel in the upper left corner of the captured image. The vanishing point position is established at the y ya coordinate, and the maximum value of the y coordinate of the captured image is set at yb, and a y yc coordinate is established so that yc = ya + k (yb - ya) . Note that k is a value so that 0 <k <1, and can, for example, be set at 1/3. A configuration can be made in which yc is established in consideration of factors such as the mounting angle of the image capture device 12 such as the y coordinate in the captured image corresponding to a position that is, for example, 20 m away from the image capture device 12. A distant region is then established as a range in which the y coordinates are from ya to yc, and a nearby region is established as a range in which the y coordinates are from yc until yb.
    [00054] Then, a determination is made as to whether there are vehicle lane limit points present in a distant region on the left side when the minimum value LF of the y coordinates for the vehicle lane limit points expressing the vehicle limit left side satisfies Lf <yc. A determination is made as to whether there are vehicle lane limit points present in a region close to the left side when the maximum value LN of the y coordinates of the vehicle lane limit points expressing the limit on the left side is LN> yc. Similarly, the minimum RF value and the maximum RN value of the y coordinates at the vehicle lane limit points expressing the right side limit are compared with yc to determine whether there are vehicle lane limit points present in the distant region of the vehicle. right side or in the region near the right side. Note that the configuration can be made in which the limit values of the coordinate y Tf and Tn are provided to respectively determine whether the limit points of the vehicle track are present or not in the distant or nearby region. Then, it is determined that there are vehicle lane limit points present in the distant region when Lf (Rf) <Tf, and it is determined that there are vehicle lane limit points present in the nearby region when LN> tN.
    [00055] As described above, in order to determine whether the vehicle lane limit points are a distribution in the distant or nearby region, and whether the vehicle lane limit points are a distribution on the left side or on the side right, patterns of the distribution of the vehicle's lane boundary points are categorized into patterns, such as those illustrated in Fig. 7A through Fig. 7P. Fig. 7P is a null observation value distribution.
    [00056] Based on the distribution of the vehicle's lane limit points determined by the distribution determination section 34, the system noise establishment section 36 then establishes the system noise corresponding to each of the xk | ka be estimated. The system noise indicates the variation in the parameter of the course when a parameter of the course is estimated at the current moment by the variation of the result of the estimate of the previous time based on the current observation.
    [00057] The stable estimate can be achieved for all parameters of the xk | k path when the vehicle's lane limit points are of a distribution both in the distant region and in the nearby region, and on both the left and right sides . However, when, for example, the vehicle's track boundary points are only distributed in the distant region, the estimation results become unstable due to a reduction in the accuracy of the estimation of the route parameters such as the lateral position and k | k of the vehicle itself and the track width of the vehicle wk | k. Consequently, the system noise is established according to each of the route parameters after first determining the observation conditions based on the distribution of the vehicle's track limit points.
    [00058] Fig. 8 illustrates an example of the methods of establishing system noise. The parameters provided can be comparably stable estimated under the following conditions: the lateral position ek | k when the vehicle's lane limit points are distributed in the nearby region; the yaw angle θk | k when the specific number or more of the vehicle's lane limit points are present (a valid observation value); the tilt angle Φk | k when the vehicle's lane limit points are distributed on both the left and right sides; the radius of curvature ck | k when the vehicle's lane boundary points are distributed from the near region to the distant region; and the vehicle lane width wk | k when the vehicle lane boundary points are distributed on both the left and right sides of the nearby region. Consequently, the pattern of distribution of the vehicle's lane limit points determined by the distribution determination section 34 is determined by determining which of any of the following classifications the distribution fits into: "near, far, left and right "at which vehicle lane limit points are present in all regions; "distant only" at which the vehicle's lane limit points are present only in the distant region; "only nearby" in which there are vehicle lane limit points present only in the nearby region; "left and right" in which there are both vehicle lane limit points expressing a left data limit and vehicle lane limit points expressing a right side limit present; "only one side" on which there are vehicle lane limit points expressing only one side within the left or right side limit present; "null observation value" in which the total number of vehicle lane limit points is the specific number or less. The alphabet codes on the bottom line of the classification names of the distribution in Fig. 8 correspond to the names of the respective distribution patterns in Fig. 7A through Fig. 7P. The patterns in Fig. 7A, Fig. 7F and Fig. 7J are not contained in "left and right". The patterns in Figs. 7H and Fig. 7N correspond to both "only distant" and "only one side". The patterns in Figs. 7L and Fig. 7O correspond to both "only near" and "only one side".
    [00059] In addition, as shown in Fig. 8, a method of establishing system noise corresponding to each of the path parameters is determined separately for each of the distribution classifications. For example, for "only distant", the noise of the system corresponding to the lateral position ek | k, the radius of curvature ck | k, and the width of the lane of the vehicle wk | k is determined to be "low noise". For "low noise" cases, the system noise is set low (for example, for "0"), and for cases other than "low noise", the system noise is set according to a conventional method. Setting the system noise down means making the variation small when estimating the path parameter, and includes an arrangement so that the path parameter in question is not updated. The path parameter is more likely to vary the higher the system noise, and as the system noise gets lower the variation is less likely to occur, stable estimation results can be obtained.
    [00060] The distribution determination section 34 and the noise establishment section of the system 36 are examples of an establishment section of the present invention.
    [00061] The section parameter estimation section 38 estimates the parameters of the path xk | k according to the Kalman filters presented below.

    [00062] Where: Xk | k is the internal state (path parameter) in time k, yk is an observation value (coordinate of the vehicle lane limit points) expressed by -Vk 'yk 1;) fk is a state transition function, hk is an observation function, Fk is a state transition matrix in time k, Gk is an orientation matrix in time k, Hk is an observation matrix, ∑k | k is the error covariance matrix predicted at time k, ∑wk is a system noise covariance matrix at time k, and ∑vk is the observation noise covariance matrix at time k. The system noise established by the system noise establishment section 36 is ∑wk from Equation (5). The path parameters xk | k are estimated by entering the coordinates of the vehicle's runway limit point as the observation value yk.
    [00063] Next, if the explanation, with reference to Fig. 9, with respect to the route estimation processing routine performed by the computer 16 of the route estimation apparatus 10 according to the present illustrative embodiment. The present routine is performed by CPU 20 by executing a route estimation program stored in ROM 22.
    [00064] In step 100, an image that was captured with image capture device 12 is acquired, then in step 102, edge points that are points where there is a change in the brightness of each pixel of the captured image are extracted as characteristic points.
    [00065] Then, in step 104, vehicle lane limit points indicating vehicle lanes are selected by determining from the characteristic points extracted in step 102 of the shape, width and color of the border points aligned in a continuous row.
    [00066] Then, in step 106, a determination is made of what type of distribution is adopted by the vehicle lane limit points selected in step 104. First, a determination is made as to whether the total number of limit points of the vehicle lane is the specific or lesser number or not, then a determination is made that the distribution is the null distribution value pattern of Fig. 7P. Then, if the minimum value LF of the y coordinate of the vehicle lane limit points indicating the limit on the left side satisfies LF <vc, it is de-terminated that there are vehicle lane limit points present in the distant region on the left side. If the maximum value LN of the y coordinate of the vehicle lane limit points indicating the limit on the left side satisfies LN> yc, then the determination is made that there are vehicle lane limit points present in the region close to the left side. Similarly, yc is compared with the y coordinates of the minimum RF value and the maximum RN value of the vehicle lane limit points indicating the right side limit to determine whether or not there are vehicle lane limit points present in the region far from the right side or in the region near the right side. These determinations are used to determine which of the lane limit point patterns of the vehicle from Fig. 7A to Fig. 7 corresponds to the distribution. For example, when there are vehicle lane boundary points present in all regions including the region on the far right, the region on the far left, the region on the far right and the region on the far right, then the distribution is defined as the pattern in Fig. 7A. Or, when the vehicle's lane boundary points are only present in the distant region on the left side and in the distant region on the right side, the distribution is determined to be the pattern in Fig. 7F.
    [00067] Then, in step 108, the system noise corresponding to each of the parameters of the path xk | k is established for estimation based on the distribution of the limit points of the vehicle track determined in step 106. As shown in Fig. 8 , the establishment of the system noise that was associated with the vehicle lane limit point distribution pattern is read with reference to the predetermined associations between the vehicle lane limit points and the system noise, and ∑wk of the Equation (5) is established. For example, when the distribution of the vehicle track limit points is determined in step 106 as the pattern in Fig. 7F, then the distribution classification is "only distant", and thus, the system noise corresponding to the lateral position of the path parameters ek | k, the radius of curvature ck | k and the track width of the vehicle wk | k is set low. When the distribution of the vehicle's lane limit points is the standard in Fig. 7N, the distribution classification is "only distant" and "only one side", and thus, in addition to the parameters of the above path, the system noise corresponding to the yaw angle θk | k is also set low.
    [00068] Then, in step 110, the coordinates of the vehicle track limit points selected in step 104 are replaced as the observation values yk and the path parameters xk | k are estimated according to Equation (1 ) and Equation (5) and the estimate results issued. The estimate results issued can therefore be displayed on a video device, not shown in the drawings, and used as input data, such as for a vehicle motion control vehicle to control the movement of the vehicle.
    [00069] As explained above, according to the route estimation apparatus of the present illustrative embodiment, based on the distribution of the vehicle track limit points, namely, the distribution of the observation values, a determination is made as to a whether or not the observation conditions are such that the estimation accuracy for each of the path parameters must be reduced. The stable estimate of the path parameter can thus be obtained by decreasing the noise of the system corresponding to the path parameters in order to reduce the accuracy of the estimate.
    [00070] In the illustrative embodiment above, an explanation is given of a case in which the path parameters are estimated using a Kalman filter, however, a configuration can be made in which another filter is used based on the processing of probability signal processing ( statistic) of a separate time signal. A particle filter can, for example, be used. In such cases, as shown in Fig. 10, (1) the probability that a path parameter will be expressed as a particle size (weighting), and (2) the path parameter in the next time will be estimated. Such a case considers a dynamic model of the vehicle in which, for example, the lateral position of the next time also shifts when the vehicle is at an angle with respect to the vehicle's lane. Then, (3) the probabilities of the path parameter are diffuse. Diffusion is done over a wide range where there is a large variance, and diffusion is done over a narrow range where there is a small variance. The diffusion width corresponds to the noise of the system of the present invention. (4) Then, the weighting is applied using the probabilities of the values of each path parameter and the observation values (vehicle track limit points), and (5) a probability distribution of renewed path parameters is then calculated for the observation values.
    [00071] Although the explanation in the present illustrative embodiment is a case in which a determination is made as to whether the distribution of the vehicle lane limit points has points present in a distant region and / or a nearby region, and on the left side and / or on the right side, there is no limitation in such cases. The determination of the distribution can be made in more refined region divisions, the determination can only be if there are points present in the distant region and / or in the nearby region, and any determination capable of determining the distribution of the limit lines of the vehicle track, the what are the vehicle tracks, can be used according to the characteristics of the parameters of the route to be estimated.
    [00072] The program of the present invention can be provided stored in a storage medium, or a mode in which the program of the present invention is provided via wired or wireless communication devices can be adopted. There is also no limitation for the implementation through a software configuration, and the implementation can be done through a hardware configuration, or a combination of a configuration and software with a hardware configuration. Explanation of Reference Numbers 10 ROUTE ESTIMATE APPARATUS 12 IMAGE CAPTURE DEVICE 16 COMPUTER 30 CHARACTERISTIC POINT EXTRACTION SECTION 32 VEHICLE TRACK LIMIT SELECTION SECTION 34 DISTRIBUTION DETERMINATION SECTION 36 STATEMENT OF DETERMINATION SECTION 36 SYSTEM 38 ROUTE PARAMETER ESTIMATE SECTION
    权利要求:
    Claims (3)
    [0001]
    1. Course estimation device (10), characterized by the fact that it comprises: an acquisition section (12) to acquire a captured image from a vehicle periphery; an extraction section (30) to extract, from the captured image acquired by the acquisition section (12), characteristic points indicating the vehicle's tracks; an establishment section (36) for, based on a distribution of the characteristic points extracted by the extraction section, to establish the system noise expressing the variation of the path parameters when estimating the path parameters related to a position or with an angle of the vehicle itself with respect to a route for travel by the vehicle itself and related to a shape or size of the route; and an estimation section (38) to estimate the path parameters by processing the probability signal using a separate time signal based on the characteristic points extracted by the extraction section, in a previous result of the estimate of the path parameters and in the noise of the system established by the establishment section (36).
    [0002]
    2. Route estimation device (10), according to claim 1, characterized by the fact that the route parameters related to the position and the angle of the vehicle itself with respect to the route include a lateral position of the vehicle itself with with respect to the course, a yaw angle with respect to a center line of the course and an angle of inclination with respect to a plane of the course, and where the parameters of the course related to the shape and size of the course include a radius of curvature of the route and a track of the vehicle on the route.
    [0003]
    3. Course estimation device (10), according to claim 2, characterized by the fact that: when the characteristic points are only distributed in a distant region in the captured image, the establishment section (36) decreases the system noise corresponding to the radius of curvature of the course, the width of the vehicle's lane of the course, and the lateral position of the vehicle itself with respect to the course; when the characteristic points are only distributed in a region close to the captured image, the establishment section (36) reduces the noise of the system corresponding to the radius of curvature of the path; when the characteristic points are in a distribution only expressing a limit on the left side of the vehicle lane or only expressing a limit on the right side of the vehicle lane, the establishment section (36) decreases system noise corresponding to the width of the vehicle's lane course and the angle of inclination with respect to the course plane; and when the number of characteristic points present is the same or less than a specific predetermined number, the establishment section (36) decreases the noise of the system corresponding to all the parameters of the route.
    类似技术:
    公开号 | 公开日 | 专利标题
    BR112013006124B1|2021-03-09|route estimation device
    CN107330376B|2020-01-21|Lane line identification method and system
    JP4157620B2|2008-10-01|Moving object detection apparatus and method
    EP3581890A2|2019-12-18|Method and device for positioning
    CN105809149A|2016-07-27|Lane line detection method based on straight lines with maximum length
    CN108921089A|2018-11-30|Method for detecting lane lines, device and system and storage medium
    KR20150112656A|2015-10-07|Method to calibrate camera and apparatus therefor
    KR101761586B1|2017-07-27|Method for detecting borderline between iris and sclera
    JP4964171B2|2012-06-27|Target region extraction method, apparatus, and program
    CN108280450B|2020-12-29|Expressway pavement detection method based on lane lines
    CN106250824A|2016-12-21|Vehicle window localization method and system
    KR20090043416A|2009-05-06|Surveillance camera apparatus for detecting and suppressing camera shift and control method thereof
    JP2005275500A|2005-10-06|Vanishing point decision method
    CN109697860A|2019-04-30|Parking stall measure and tracking system and method and vehicle
    CN108171695A|2018-06-15|A kind of express highway pavement detection method based on image procossing
    CN109584294A|2019-04-05|A kind of road surface data reduction method and apparatus based on laser point cloud
    CN104866838B|2018-08-03|A kind of front vehicles automatic testing method of view-based access control model
    JP6844235B2|2021-03-17|Distance measuring device and distance measuring method
    US20170178341A1|2017-06-22|Single Parameter Segmentation of Images
    JP2014041427A|2014-03-06|Object detection device and program
    JP3605955B2|2004-12-22|Vehicle identification device
    JP2019218022A|2019-12-26|Rail track detection device
    JP2008084109A|2008-04-10|Eye opening/closing determination device and eye opening/closing determination method
    JP3999345B2|2007-10-31|Own vehicle position recognition device, own vehicle position recognition method and program
    CN108765456B|2020-10-30|Target tracking method and system based on linear edge characteristics
    同族专利:
    公开号 | 公开日
    CN103098111B|2015-12-16|
    WO2012039496A1|2012-03-29|
    EP2620930B1|2016-08-31|
    CN103098111A|2013-05-08|
    EP2620930A4|2015-09-16|
    US20130177211A1|2013-07-11|
    JP2012068961A|2012-04-05|
    US8948455B2|2015-02-03|
    BR112013006124A2|2016-05-31|
    JP5258859B2|2013-08-07|
    EP2620930A1|2013-07-31|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    JP3288566B2|1994-11-10|2002-06-04|株式会社豊田中央研究所|Travel lane recognition device|
    JP3538476B2|1995-05-12|2004-06-14|本田技研工業株式会社|Recognition device for vehicle lane markings|
    US6091833A|1996-08-28|2000-07-18|Matsushita Electric Industrial Co., Ltd.|Local positioning apparatus, and a method therefor|
    JPH1123291A|1997-07-04|1999-01-29|Nissan Motor Co Ltd|Picture processing device for car|
    JP3319399B2|1998-07-16|2002-08-26|株式会社豊田中央研究所|Roadway recognition device|
    JP3521860B2|2000-10-02|2004-04-26|日産自動車株式会社|Vehicle travel path recognition device|
    JP3695315B2|2000-11-14|2005-09-14|日産自動車株式会社|Vehicle display device|
    JP4086759B2|2002-11-05|2008-05-14|ダイハツ工業株式会社|Road model estimation apparatus and method|
    US7660436B2|2003-06-13|2010-02-09|Sarnoff Corporation|Stereo-vision based imminent collision detection|
    JP3864945B2|2003-09-24|2007-01-10|アイシン精機株式会社|Road lane detection device|
    JP4390631B2|2004-06-02|2009-12-24|トヨタ自動車株式会社|Boundary line detection device|
    JP4659631B2|2005-04-26|2011-03-30|富士重工業株式会社|Lane recognition device|
    CN100535954C|2005-08-30|2009-09-02|珠海金联安软件有限公司|Multifunctional running and stopping guide alarm system|
    CN100403332C|2006-11-02|2008-07-16|东南大学|Vehicle lane Robust identifying method for lane deviation warning|
    JP2009154647A|2007-12-26|2009-07-16|Aisin Aw Co Ltd|Multi-screen display and program of the same|
    JP5058002B2|2008-01-21|2012-10-24|株式会社豊田中央研究所|Object detection device|
    JP2010191661A|2009-02-18|2010-09-02|Nissan Motor Co Ltd|Traveling path recognition device, automobile, and traveling path recognition method|JP3712760B2|1995-05-17|2005-11-02|Ｔｄｋ株式会社|Organic EL device|
    JP3861400B2|1997-09-01|2006-12-20|セイコーエプソン株式会社|Electroluminescent device and manufacturing method thereof|
    US6756165B2|2000-04-25|2004-06-29|Jsr Corporation|Radiation sensitive resin composition for forming barrier ribs for an EL display element, barrier rib and EL display element|
    JP4378186B2|2004-02-06|2009-12-02|キヤノン株式会社|Organic EL element array|
    JP5970811B2|2011-12-28|2016-08-17|セイコーエプソン株式会社|LIGHT EMITTING ELEMENT, LIGHT EMITTING DEVICE, AND ELECTRONIC DEVICE|
    US9740942B2|2012-12-12|2017-08-22|Nissan Motor Co., Ltd.|Moving object location/attitude angle estimation device and moving object location/attitude angle estimation method|
    BR112016006666A2|2013-09-27|2017-08-01|Nissan Motor|information presentation system|
    JP6046666B2|2014-06-24|2016-12-21|トヨタ自動車株式会社|Runway boundary estimation device and runway boundary estimation method|
    KR20160059376A|2014-11-18|2016-05-26|엘지전자 주식회사|Electronic appartus and method for controlling the same|
    JP6363518B2|2015-01-21|2018-07-25|株式会社デンソー|Lane marking recognition system|
    EP3435352B1|2016-03-24|2021-11-17|Nissan Motor Co., Ltd.|Travel path detection method and travel path detection device|
    CN106092121B|2016-05-27|2017-11-24|百度在线网络技术（北京）有限公司|Automobile navigation method and device|
    JP2019011971A|2017-06-29|2019-01-24|株式会社東芝|Estimation system and automobile|
    US10586118B2|2018-01-13|2020-03-10|Toyota Jidosha Kabushiki Kaisha|Localizing traffic situation using multi-vehicle collaboration|
    US10916135B2|2018-01-13|2021-02-09|Toyota Jidosha Kabushiki Kaisha|Similarity learning and association between observations of multiple connected vehicles|
    US10963706B2|2018-01-13|2021-03-30|Toyota Jidosha Kabushiki Kaisha|Distributable representation learning for associating observations from multiple vehicles|
    法律状态:
    2018-12-26| B06F| Objections, documents and/or translations needed after an examination request according [chapter 6.6 patent gazette]|
    2019-10-01| B06U| Preliminary requirement: requests with searches performed by other patent offices: procedure suspended [chapter 6.21 patent gazette]|
    2020-10-13| B06A| Patent application procedure suspended [chapter 6.1 patent gazette]|
    2021-01-19| B09A| Decision: intention to grant [chapter 9.1 patent gazette]|
    2021-03-09| B16A| Patent or certificate of addition of invention granted [chapter 16.1 patent gazette]|Free format text: PRAZO DE VALIDADE: 20 (VINTE) ANOS CONTADOS A PARTIR DE 26/09/2011, OBSERVADAS AS CONDICOES LEGAIS. |
    优先权:
    申请号 | 申请日 | 专利标题
    JP2010-214025|2010-09-24|
    JP2010214025A|JP5258859B2|2010-09-24|2010-09-24|Runway estimation apparatus and program|
    PCT/JP2011/071898|WO2012039496A1|2010-09-24|2011-09-26|Track estimation device and program|
    [返回顶部]