JP4884109B2 - Moving locus calculation method, moving locus calculation device, and map data generation method - Google Patents

Description

  The present invention relates to a map data generation method for generating map data used in a navigation system, an automobile simulator, and the like, and more particularly, a map data generation method for generating map data by collecting position data while traveling on a road with a vehicle, and the map data generation method therefor The present invention relates to a movement trajectory calculation method and a movement trajectory calculation apparatus to be used.

In the navigation system, a traveling position is detected by a GPS device, a gyroscope, a speed (acceleration) sensor, and the like, and the traveling position is displayed on a map. When the GPS device has a good reception state, that is, when the number of visible satellites is four or more, the traveling position is mainly the traveling position detected by the GPS device, and the reception state of the GPS device is insufficient, that is, the number of visible satellites is four. If it is less than the machine, or if it can be judged that the detection accuracy is poor from the DOP value etc., the position calculated based on the moving direction (azimuth) detected by the gyro and the moving speed detected by the speed (acceleration) sensor is mainly obtained. ing. Here, the value of DOP is an index calculated from the positioning of GPS satellites, and is an amount calculated from the diagonal elements of the inverse matrix when solving the positioning equation using the least squares method, and is also related to the standard deviation. The smaller the DOP value, the better the detection accuracy. The GPS receiver is equipped with a function for outputting DOP as standard. However, GPS receivers that output standard deviation are limited.
Even if the DOP value is small, the detection error increases when a multipath is received. Therefore, the accuracy cannot be determined only by the DOP, but it can be said that the DOP corresponds to the detection accuracy in a place where the sky field of view without the multipath is opened. .

  The navigation system has map data created in advance, compares the detected driving position with the position of the map data on the road, and corrects the detected driving position to the position on the road when it is off the road. Is done.

  The navigation system is described in, for example, Patent Document 1. Patent Document 1 describes a configuration in which an error existing in the output of each sensor is estimated by a Kalman filter, and a position is calculated by performing correction based on the estimated value.

  The map data as described above is created based on aerial surveys, aerial photographs, ground surveys, and the like. Since roads are newly constructed or repaired, it is necessary to correct the map data accordingly. However, it is difficult in terms of cost to frequently perform the above-mentioned aerial survey, aerial photograph, and survey on the ground. Therefore, as the simplest method, it is conceivable to create a map data by actually traveling on a road with a vehicle equipped with a device for detecting a traveling position and detecting the movement trajectory. The device for detecting the travel position used here is similar to the navigation system, but the navigation system needs to output the travel position in real time, whereas if the travel position (movement trajectory) is obtained after travel, It is a good offline system.

  FIG. 1 is a diagram showing the configuration of a system that generates map data as described above. As shown in FIG. 1, the vehicle 1 includes a GPS device 2, a gyro 3 that detects a moving direction, a vehicle speed sensor 4, and a data storage device 5 that stores data output during driving. ing. After traveling, the data during traveling stored in the data storage device 5 is sent to the map data generating device 6, and the map data generating device 6 processes these data to calculate the traveling position, that is, the movement trajectory to map data. Output as. The GPS device 2 can receive and process a signal from a GPS satellite with higher accuracy than a normal device mounted on an automobile or the like, and is mounted integrally with an antenna on the top of the vehicle.

  The calculation of the movement trajectory in the system of FIG. 1 is basically performed by the same method as that performed in the conventional navigation system. By using the GPS device 2, when the number of visible satellites is four or more or when the DOP value is small, highly accurate position measurement can be performed. In such a case, the position detected by the GPS device 2 is used as the traveling position. To do. When traveling in the valley or between buildings, the sky may be blocked by a shield, and the number of visible satellites may be less than 4 or the DOP value may be large. A highly accurate positioning solution cannot be obtained. Therefore, position estimation is performed by a Kalman filter using detection data of the gyro 3 and the vehicle speed sensor 4.

  FIG. 2 is a diagram for explaining the calculation of the movement trajectory in the conventional system. As shown in FIG. 2A, in the movement trajectory when traveling in the direction of the arrow, A1 and A2 are travel positions detected by the GPS device 2 when the number of visible satellites is 4 or more or the DOP value is small The movement trajectory obtained from is shown. In such a case, the traveling position can be obtained with high accuracy by the GPS device 2 as described above.

  The range from the end point P of the movement locus A1 to the end point Q of A2 is when the number of visible satellites is less than 4 or the DOP value is large, and the position is estimated by the Kalman filter using the detection data of the gyro 3 and the vehicle speed sensor 4. To obtain the movement trajectory. Conventionally, as shown in FIG. 2A, a movement trajectory B1 from the point P is obtained, but since the error is accumulated in the trajectory calculated in this way, the vehicle travels at the point Q as shown. The point Q ′ is calculated as the travel position at the power point, and the point Q ′ is changed to the point Q. This means that the movement trajectory B1 calculated as described above deviates from the actual road. For example, the movement trajectory B1 is off the road from the middle and passes through the building.

  Since the system of FIG. 1 is an offline system as described above, the stored data of the gyro 3 and the vehicle speed sensor 4 are used to move from point Q to point P as shown in FIG. Although it is conceivable to calculate the backward movement trajectory B2, the point P ′ is calculated as the travel position at the time when the point P should be traveled, as described above, and the same problem occurs.

Japanese Patent Laid-Open No. 7-301541

  Gyros range from expensive ones that can measure with high accuracy for a long time to ones that are inexpensive but not very high in measurement accuracy. Similarly, if an expensive speed (acceleration) sensor is used, the error can be reduced accordingly. Therefore, if an expensive gyroscope or speed (acceleration) sensor is used, the error in the moving direction is small, and the difference between the calculated movement locus and the actual road can be reduced. However, this causes a problem that the cost increases.

  An object of the present invention is to realize a new calculation method capable of calculating a trajectory with a small error from an actual movement trajectory even when a relatively low-priced gyroscope and a speed (acceleration) sensor are used.

  In order to achieve the above object, the movement trajectory calculation method, the movement trajectory calculation apparatus, and the map data generation method of the present invention calculate a movement trajectory after moving a movement trajectory between a first point and a second point of a moving object. With respect to the first point, the movement direction and movement speed during movement from the first point to the second point are detected and stored as movement data, and the stored movement data is used in the forward direction from the first point. A forward movement trajectory to the second point is calculated, the backward movement trajectory from the second point to the first point is calculated using the stored movement data in the reverse direction, and the forward movement trajectory is calculated. And the reverse movement trajectory are combined with a weighting ratio that changes according to the position with respect to the first point and the second point.

  FIG. 3 is a diagram for explaining the principle of the present invention. In FIG. 3A, for example, the movement trajectories A1 and A2 are movement trajectories when the vehicle is moved, and the travel positions detected by the GPS device 2 when the number of visible satellites is four or more or the DOP value is small. The highly accurate movement trajectory obtained from Accordingly, the first point P and the second point Q, which are the end points of the movement trajectories A1 and A2, respectively, have high positional accuracy. When calculating the movement locus between the first point P and the second point Q based on the movement data detected and stored, the stored movement data is used in the forward direction from the first point P to the second point Q. The forward movement trajectory B1 is calculated, the stored movement data is used in the reverse direction, the reverse movement trajectory B2 from the second point Q to the first point P is calculated, and the forward movement trajectory B1 moves backward. The trajectory B2 is synthesized with a weighting ratio that changes according to the positions of the first point P and the second point Q.

  FIG. 3B is a diagram illustrating a change example of the weighting ratio of the synthesis, where the horizontal axis represents the movement distance or time, and the vertical axis represents the weighting ratio.

  As described above, the moving direction detected by the gyro accumulates errors, and the moving position is calculated by adding the displacement per unit time to the previously calculated moving position. If there is an error, the calculated movement locus also gradually increases. Accordingly, the forward movement trajectory B1 has a relatively small error near the point P, but the error increases as the distance from the point P increases. Similarly, the reverse direction movement locus B2 has a relatively small error near the point Q, but the error increases as the distance from the point Q increases. According to the present invention, the weighting ratio for the forward movement trajectory B1 is increased in the portion close to the point P, the weighting ratio for the backward movement trajectory B2 is decreased, and the weighting ratio for the forward movement trajectory B1 is near the point Q. And the weighting ratio with respect to the reverse direction movement locus B2 is increased, so that the influence of the accumulated error can be reduced and a locus close to the actual movement locus can be calculated.

  As shown in FIG. 3B, the movement trajectory is calculated also using the movement data of the portions outside the points P and Q. This is to obtain a smooth trajectory at points P and Q.

  In FIG. 3, the portions of the trajectories A1 and A2 are assumed to be highly accurate, but the present invention is applied when only the vicinity of the points P and Q is highly accurate and the other portions are not sufficiently accurate. It is also possible. In other words, the present invention can be applied to a case where a highly accurate movement position cannot be detected continuously and a trajectory between them is calculated based on movement data.

  The detection of the moving direction is performed by the gyro, but the moving direction is not limited to the gyro as long as the moving direction can be detected.

  The detection of the moving speed is not limited as long as the moving speed can be detected. For example, when the moving object is the vehicle shown in FIG. 1, a vehicle speed sensor can be used.

  Also, since the frequency of the received radio waves received from each satellite by the GPS receiver changes due to the Doppler effect, the movement speed in each direction can be detected by detecting the frequency of the received radio waves. It is also possible to do. In this case, the moving speed and the moving direction can be calculated from the moving speed in each direction, and it is not necessary to provide a gyro and a speed sensor (acceleration sensor), and only a GPS receiver may be provided.

  A Kalman filter is used to calculate the movement trajectory based on the movement direction and the movement speed.

  The movement trajectory calculated as described above is used for map data of the navigation system, creation of a visual database for a simulator, creation of teaching materials such as a driving school.

  According to the present invention, a new calculation algorithm capable of reducing calculation errors in off-line movement trajectory calculation is realized, and a trajectory having a small error from an actual movement trajectory even when a low-cost gyroscope and a speed (acceleration) sensor are used. Can be calculated.

  FIG. 4 is a diagram showing a configuration of a system for generating map data according to the first embodiment of the present invention, and is a system for generating map data from movement data mounted on a vehicle and collected as shown in FIG.

  As shown in FIG. 4, the system of the first embodiment includes a GPS device 11, a vibrator gyro 12, a vehicle speed sensor 13, a data storage unit 21, a forward trajectory calculation unit 31, and a reverse trajectory calculation unit. 32 and an integrated processing unit 33. The vibrator gyro 12 includes a plane gyro and a vertical gyro, and detects an angle (path angle) formed by a projection (azimuth) in a moving direction into a horizontal plane (xy plane) and a horizontal plane in the moving direction. The data storage unit 21 includes a GPS data storage unit 22, a movement direction data storage unit 23, and a movement speed data storage unit 24. The GPS device 11, the vibrator gyro 12, the vehicle speed sensor 13, and the data storage unit 21 are mounted on a vehicle traveling on a road, as in FIG. In other words, the system of the embodiment is different from the conventional example in the processing of the map data generation device, and the configuration of the other parts is the same. The forward trajectory calculation unit 31, the reverse trajectory calculation unit 32, and the integration processing unit 33 are realized by a computer.

  The integration processing unit 33 uses the movement locus obtained from the traveling position detected by the GPS device 11 as the movement locus for a portion where the number of visible satellites received by the GPS device 11 is four or more or a portion where the DOP value is small. For the portion where the number of visible satellites is less than four or the portion where the DOP value is large, the movement locus is calculated by synthesizing the forward locus and the backward locus output from the forward locus calculation unit 31 and the backward locus calculation unit 32. To do. The conventional example is the same in that the movement locus obtained from the traveling position detected by the GPS device 11 when the number of visible satellites is four or more is used as it is.

  The forward trajectory calculation unit 31 is configured to store the moving direction data and moving speed data (according to the movement direction data storage unit 23 and the moving speed data storage unit 24) for the portion of the GPS device 11 having less than four visible satellites. (Movement data) is used in the forward direction, position estimation is performed using a Kalman filter, and a forward movement trajectory is calculated. This is the same in the conventional example.

  The reverse direction trajectory calculation unit 32 uses the movement data stored in the movement direction data storage unit 23 and the movement speed data storage unit 24 in the reverse direction for the portion where the number of visible satellites of the GPS device 11 is less than four, The backward movement trajectory is calculated.

  As described above, since the error is accumulated in the moving direction detected by the gyro, the moving direction data at the position corresponding to the second point Q in FIG. 3 may include a large error. . Therefore, in FIG. 3, the reference movement direction of the second point Q predicted from the movement trajectory A2 is calculated, a correction amount is determined so that the movement direction data matches the reference movement direction, and the other movement data is determined as this movement data. Correct only the correction amount. The movement speed data itself detected by the vehicle speed sensor 13 is not corrected because no error is accumulated. After that, using the movement data in the reverse direction, the position is estimated by the Kalman filter in the same way as in the forward direction, and the reverse movement trajectory is calculated.

  Further, as shown in FIG. 3B, not only the movement data between the points P and Q but also the movement data outside the points P and Q are used to calculate the movement trajectories B1 and B2. However, this is to calculate the smooth movement trajectories B1 and B2 at the points P and Q. However, since the portions of the trajectories A1 and A2 outside the points P and Q are highly accurate, A1 and A2 are used as the trajectories.

  The integration processing unit 33 displays the forward movement trajectory output by the forward trajectory calculation unit 31 and the backward movement trajectory output by the reverse trajectory calculation unit 32 for the portion where the number of visible satellites of the GPS device 11 is less than four. 3 is synthesized with a weighting ratio that changes as in (B) of FIG. For example, if the position of the forward movement trajectory at a certain point in time is J and the backward movement trajectory is K, the composite position L is represented by L = RJ + SK. For example, R and S change as shown in FIG. 3B, but are not limited thereto.

  As described above, the method of estimating the position of the backward movement by estimating the position using the Kalman filter using the movement data is the same as the conventional example and is well known, but will be briefly described here.

FIG. 5 is a diagram for explaining a coordinate system in this method. Here, the center of gravity position of the vehicle in three-dimensional orthogonal coordinates x, y, and z is represented by x, y, and z, the azimuth angle in the traveling direction on the xy plane is ψ, and the path angle that is made with the xy plane in the traveling direction is γ. To express. Assuming that the movement position is calculated for each minute time Δt, the values at time k are represented by x _k , y _k , z _k , ψ _k , and γ _k , respectively. The moving position and moving direction (azimuth angle and path angle) at time k + 1 are expressed by equation (1) with respect to the value at time k.

However, V _k and ω _k are the vehicle speed and the azimuth velocity at the time k. Here, the positions x _k , y _k , the azimuth angle ψ _k , and the path angle γ _k are outputs of the GPS device 11, and the speed V _k , the azimuth angular speed ω _k , and the path angular speed α _k are the vehicle speed sensor, horizontal gyro, and vertical, respectively. This is the output of the gyro.

Here, if Expression (1) is expressed as Expression (2), x _{k + 1} can be expressed by Expression (3).

Here, w _k in the driving noise of u _k, it follows a normal distribution N (0, Q _k).

  The observation equation of the GPS device is expressed by equations (4) and (5).

Here, v _k is an observation noise and follows N (0, R _k ).

R _k can be obtained as a function of latitude, longitude standard deviation or DOP if GPS positioning results are used.

In the formula (3), x _k when given observation sequence of the covariance matrix M _{k + 1} estimate error of x _k at the time when the observation sequence of P _k to time k given to the time k _If the covariance matrix of ₊₁ one-stage prediction error is obtained, the following equation (6) is approximately obtained from the equation (3).

  Here, formulas (7) and (8) were further adopted.

Also, x _k and w _k are not correlated.

  From the equation (5), the following equation (9) is obtained.

  Therefore, the estimation algorithm based on the extended Kalman filter becomes the following equation, and the one-stage predicted value is represented by the following equation (10).

  The estimated value that accompanies the update of the observed value is expressed by Equation (11).

  Further, the estimated error covariance matrix is expressed by Expression (12).

  The Kalman gain is expressed by Expression (13).

Further, from the equation (4), the covariance matrix of the observed value z _{k + 1} when z _k (k = 0, 1, 2,..., K) is given is the following equation (14).

That is, the observation prediction error expressed by the equation (15) follows the normal distribution N (0, S _{k + 1} ).

  The process of calculating the reverse movement trajectory by estimating the position using the Kalman filter using the movement data is performed as described above, but here the reliability of the observation value is calculated based on the Mahalanobis distance of the observation prediction error If it is low, a reject process is performed. Further, the position and the azimuth angle of the GPS device are considered to be independent because they are amounts obtained from the pseudorange and the Doppler speed. Therefore, as described below, the Mahalabis distance related to the position and the Mahalabis distance related to the bearing are calculated separately.

  First, the Mahalabis distance related to the position will be described.

  It is assumed that the azimuth angle variance of the observation error covariance R is infinite as shown in Equation (16), assuming that the azimuth reliability is low.

  The Maharabis distance regarding the position in this case is as shown in Equation (17).

  Next, the Mahalanobis distance regarding the azimuth and the route angle will be described.

  Assuming that the reliability of the position is low, the variance of the position (X, Y, Z) is infinite as shown in the equation (18).

  In this case, the Mahalanobis distance regarding the azimuth and the route angle is expressed by Equation (19).

  The estimated value of Equation (11) and the covariance matrix of Equation (12) are updated according to the following criteria.

  If the value on the left side of Equation (17) is greater than the value on the left side of Equation (19), the previous value is used without updating.

When the value on the left side of Expression (17) is larger than the threshold value, it is considered that the reliability of (x, y, z) is low, and P _k is _obtained using Expression (18) in Expression (12). . The previous value is used for the position estimate, and only the direction of movement is updated using the observed value.

If the value on the left side of Equation (19) is larger than the threshold value, it is considered that the reliability of (ψ _k , γ _k ) is low, and P _k is _obtained using Equation (16) in Equation (12). . The previous value is used for the estimated value of the moving direction, and only the position is updated using the observed value.

  When the value on the left side of Expression (17) is larger than the threshold value and the value on the left side of Expression (19) is larger than the threshold value, the observed value is updated in the same manner as in a normal Kalman filter.

  In the first embodiment, the vibratory gyroscope 12 having a planar gyroscope and a vertical gyroscope and the vehicle speed sensor 13 for detecting the moving speed are used to detect the moving direction. The GPS device 11 has a Doppler effect in the frequency of the received radio wave due to the movement of the vehicle with respect to each satellite, and the reception frequency changes. This change in the reception frequency cannot be detected unless the vehicle is moving at a certain speed or more, but if the vehicle is limited to a moving condition at a certain speed, the GPS device 11 can detect the moving speed. . The second embodiment is an embodiment under such conditions.

  FIG. 6 is a diagram showing the configuration of the GPS device 11 used in the map data generating system according to the second embodiment of the present invention. The second embodiment is the first embodiment except that the moving direction and moving speed calculated by this configuration are used as the moving direction detected by the vibrator gyro 12 and the moving speed detected by the vehicle speed sensor 13 of the first embodiment. The configuration is the same as the example. Therefore, in the second embodiment, it is not necessary to provide the vibrator gyro 12 and the vehicle speed sensor 13. The GPS device 11 needs to detect the moving direction and moving speed when the number of visible satellites is less than four, but it is desirable to always detect the moving direction and moving speed by the GPS device 11.

As shown in FIG. 6, the reception frequency analysis unit 41 analyzes the frequency (cycle) of the received radio waves from the plurality of satellites received by the GPS reception unit 40, and calculates the displacement of the frequency from the reference frequency for each satellite. calculate. Since this displacement is caused by the Doppler effect for each satellite, the moving speed calculation means 42 calculates the speed for each satellite from the displacement for each satellite and combines them to calculate the moving speed (movement vector) including the direction. To do. The V _k calculation unit 43 calculates the movement speed V _k from the absolute value of the movement vector. The ψ _k , γ _k calculation unit 44 calculates the azimuth angle ψ _k and the path angle γ _k from the ratio of the three-axis components of the movement vector.

  Based on the movement direction and movement speed calculated as described above, the movement locus is calculated in the same manner as in the first embodiment. In this case, the azimuth angular velocity and the path angular velocity are constant, and equation (20) is used as a state variable instead of equation (1).

The azimuth angle ψ _k and the path angle γ _k are calculated according to equation (21).

  Here, even if an error is included in the pseudo distance due to multipath, the delta range or Doppler speed used to determine the speed does not include an error, and thus an accurate speed can be determined.

  In the second embodiment, the equations (7) and (8) in the first embodiment become the equation (22). A Kalman filter is constructed for this expression.

  The present invention can be applied to the case where the trajectory calculation is calculated offline.

It is a figure which shows the system configuration | structure containing the vehicle for obtaining the data regarding a movement path | route. It is a figure explaining the prior art example of a movement locus | trajectory calculation. It is a figure explaining the principle of the movement locus | trajectory calculation of this invention. It is a figure which shows the structure of the map data generation system of 1st Example of this invention. It is a figure explaining the coordinate system in 1st Example. It is a figure which shows the structure of the GPS apparatus used with the map data generation system of 1st Example of this invention.

Explanation of symbols

DESCRIPTION OF SYMBOLS 11 GPS apparatus 12 Vibrator gyro 13 Vehicle speed sensor 21 Data storage part 31 Forward direction trajectory calculation part 32 Reverse direction trajectory calculation part 33 Integration processing part

Claims (17)

A movement trajectory calculation method for calculating a movement trajectory between a first point and a second point of a moving object after movement,
Detecting the moving direction and moving speed during movement from the first point to the second point and storing it as movement data;
Using the stored movement data in the forward direction, a forward movement trajectory from the first point to the second point is calculated,
Using the stored movement data in the reverse direction to calculate a reverse movement trajectory from the second point to the first point,
The moving trajectory calculation method comprising combining the forward moving trajectory and the backward moving trajectory with a weighting ratio that changes in accordance with a position with respect to the first point and the second point.   The first point and the second point are points detected by a GPS receiver using four or more visible satellites, or points that are determined to be accurate based on a DOP value, and the second point from the first point. 2. The movement trajectory calculation method according to claim 1, wherein between the points is a portion where the number of visible satellites by the GPS receiver is less than four, or a portion where accuracy is determined to be poor based on a DOP value or the like.   The movement trajectory calculation method according to claim 1, wherein the movement direction is detected by a gyro. The moving object is a vehicle;
The movement trajectory calculation method according to claim 1, wherein the movement speed is detected by a vehicle speed sensor.   The movement trajectory calculation method according to claim 2, wherein the movement speed and the movement direction are calculated from the movement speed for each direction detected by the GPS receiver.   The movement trajectory calculation method according to claim 1, wherein the forward movement trajectory and the backward movement trajectory are calculated by a Kalman filter. A movement trajectory calculation device for calculating a movement trajectory between a first point and a second point of a moving object after movement,
A storage device that detects a moving direction and a moving speed during movement from the first point to the second point and stores them as movement data;
A forward trajectory calculating unit that calculates a forward trajectory from the first point to the second point using the stored movement data in the forward direction;
Using the stored movement data in the reverse direction, a reverse direction trajectory calculation unit for calculating a reverse direction movement trajectory from the second point to the first point;
A movement trajectory calculation apparatus comprising: a synthesis processing unit that synthesizes the forward movement trajectory and the reverse movement trajectory with a weighting ratio that changes according to a position with respect to the first point and the second point. . With a GPS receiver,
The first point and the second point are points detected by four or more visible satellites with the GPS receiver, or points that are determined to have high accuracy based on the DOP value, etc. 8. The movement trajectory calculation apparatus according to claim 7, wherein between two points is a portion where the number of visible satellites by the GPS receiver is less than four, or the DOP value is determined to be large and the accuracy is low. Equipped with a gyro,
The movement trajectory calculation apparatus according to claim 7 or 8, wherein the movement direction is detected by the gyro. The moving object is a vehicle;
The movement locus calculation apparatus according to claim 7, further comprising a vehicle speed sensor that detects a movement speed of the vehicle.   The movement trajectory calculation apparatus according to claim 8, further comprising a movement speed calculation unit that calculates the movement speed and the movement direction from a movement speed for each direction detected by the GPS receiver.   The movement trajectory calculation apparatus according to claim 7, wherein the trajectory calculation unit includes a Kalman filter. A map data generation method for generating map data,
While traveling on the road with the vehicle, the GPS device mounted on the vehicle detects the traveling position, detects the moving direction and moving speed during movement, and stores it as movement data,
For the point detected by the GPS receiver with four or more visible satellites or the point determined to have high accuracy by the DOP value or the like, the traveling position detected by the GPS device is used as road map data,
For the parts where the number of visible satellites by the GPS receiver is less than four, or the parts judged to be inaccurate due to the DOP value, etc., the position was detected by four or more visible satellites at the GPS receivers at both ends of the part. Using the movement data stored in the forward direction as the first point and the second point, the forward movement trajectory from the first point to the second point is calculated,
Using the stored movement data in the reverse direction to calculate a reverse movement trajectory from the second point to the first point,
The map data generation method, wherein the forward movement trajectory and the backward movement trajectory are combined with a weighting ratio that changes according to the position with respect to the first point and the second point.   The map data generation method according to claim 13, wherein the movement direction is detected by a gyro.   The map data generation method according to claim 13, wherein the movement speed is detected by a vehicle speed sensor.   The map data generation method according to claim 13, wherein the movement speed and the movement direction are calculated from a movement speed for each direction detected by the GPS receiver.   The map data generation method according to any one of claims 13 to 16, wherein the forward movement locus and the backward movement locus are calculated by a Kalman filter.

Images