JPH1170884A - Self-advancing own vehicle position detecting device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To minimize the dispersion causing a fluctuation by calculating the displacement from a reference position by use of white line data of a plurality of areas to a distant position, setting a high confidentness in the white line information of a position having a good distance precision on an image, and changing the output signal to steering control side according to the confidentness and the road forward curvature radius. SOLUTION: When the displacement from a reference position is calculated by use of white line data to a distant position, and outputted to steering control side, the confidentness of the calculated data is set high by an output signal selecting means 109 when the white line information is present in the area close to the own vehicle with a good distance precision on an image. When no white information is present in the area near the own vehicle, the confidentness of the calculated data is set low. In the output signal selecting means 109, an optimum signal is selected and outputted to steering control according to the confidentenss and the degree of the road forward curve. Thus, a signal for performing a smooth steering control can be outputted.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for realizing smooth steering control in automatic driving, in which a vehicle-mounted camera captures a road image ahead of a vehicle, and a deviation from a reference position based on the obtained image information. The present invention relates to a self-vehicle position detection device for automatic traveling that calculates

[0002]

2. Description of the Related Art An on-vehicle camera captures an image of a road ahead of a vehicle and calculates a deviation of the own vehicle from a reference position, that is, a position of the own vehicle in a traveling lane, based on the obtained image information. Japanese Patent Laid-Open Publication No. 4-373,004 discloses an example of a device that performs steering control of an automatic traveling vehicle according to information.

In this vehicle position detection device, a white line on the road is imaged by a camera, and the vehicle travels automatically along the white line.
At this time, in order to be able to estimate the deviation from the reference position even if the white line is locally interrupted, a plurality of measurement points of the white line are set, and the input values at other measurement points are also reliable for each measurement point. A membership function that gives a degree is set, thereby expanding the effective range of the input value of each measurement point, and reflecting the input values of other measurement points in the control formula at each measurement point.

That is, even if the white line at a certain measurement point is interrupted, the deviation from the reference position is calculated using the white line data at another measurement point, and the steering control is always performed based on the calculation result.

[0005]

However, in such a conventional vehicle position detecting device, when the white line of the traveling path is a broken line, the break of the broken line is close to the vehicle. However, in the image processing, the deviation from the reference position is calculated using data of a plurality of measurement points up to a distant place.

Therefore, if white line data is not detected at a short distance with good distance accuracy on the image, the deviation from the reference position is calculated using long-distance white line data with poor distance accuracy. In addition, the obtained calculation result includes an error.

When this result is output to the steering control side, the effect of this error appears in a form such as "vehicle wobble". In particular, in the case where steering control is performed in a place where a state in which the vehicle travels on a straight road with a white line as a broken line is desirable, it is desirable to always travel without swaying on a target line, and for that purpose, a signal output to the steering control side Must be small. It is extremely difficult to remove this variation on the steering control side of the post-processing in consideration of the response of the control.

SUMMARY OF THE INVENTION The present invention has been made in view of such problems of the prior art, and has been made in consideration of the above-described problems. It is an object of the present invention to provide a traveling vehicle position detecting device.

[0009]

According to one aspect of the present invention, there is provided a self-vehicle position detecting device for an automatic vehicle related to steering control of a vehicle. An image pickup unit mounted to pick up a road image ahead of the vehicle, a road shape storage unit that stores a road shape as a road model, a coordinate conversion unit that converts a coordinate system to obtain a coordinate value on the image, Area setting means for setting an area in the vicinity of a straight line or a curved line at the time of calculation; feature point coordinate detection means for scanning image information taken by the imaging means to obtain position coordinates of feature points representing lanes; A lane extracting means for determining a lane formed by connecting pixels representing the characteristic points so as to form a straight line or a curve with reference to the coordinates of the characteristic points extracted by the point coordinate detecting means; And the coordinate values of the extracted feature points is compared with the image coordinate value converted by the coordinate converting unit of the road models stored in the road shape storing means,
Curve, slope and other parameters representing the road shape, and deviation from the reference position, height, yaw angle, pitch angle, roll angle and other parameters representing the attitude of the imaging means, at least the deviation from the reference position A change amount calculating means for calculating a change component of curvature, a road model stored in the road shape storage means from a change amount calculated by the change amount calculating means and a lane shape extracted by the lane extracting means; A parameter updating unit for updating the coordinate conversion unit and other parameters; and an output signal selecting unit for outputting a deviation from a reference position to a vehicle control side, wherein the output signal selecting unit is the feature point coordinate detecting unit. From the processing result in,
Using the information of the feature points representing the lane in the vicinity of the own vehicle, determine the confidence of the deviation from the reference position updated by the parameter updating means, and output a signal to the vehicle control side according to the confidence. It is characterized by updating.

According to a second aspect of the present invention, in the self-vehicle position detecting apparatus for automatic traveling, the output signal selecting means may be configured to determine a degree of curvature in front of the vehicle in addition to a certainty of deviation from a reference position. A signal output to the vehicle control side is updated.

Further, the self-vehicle position detecting device for automatic driving according to a third aspect of the present invention further comprises a road surface state detecting means for directly or indirectly detecting a road surface state of a traveling road, and the output signal selecting means comprises: A signal output to the vehicle control side is updated according to the degree of certainty of the deviation from the reference position, the degree of curvature ahead of the vehicle, and the road surface condition of the traveling road detected by the road surface condition detecting means. .

[0012] In the self-vehicle position detecting device for an automatic traveling according to any one of the first to third aspects, the image pickup means corresponds to the image pickup section 1 in the embodiment, and may be constituted by, for example, a video camera. The road shape storage unit, coordinate conversion unit, area setting unit, feature point coordinate detection unit, lane extraction unit, change amount calculation unit, parameter update unit, and output signal selection unit correspond to the image processing unit 2 in the embodiment. However, for example, it can be constituted by a microcomputer. The road surface state detecting means corresponds to the road surface state detecting unit 12 in the embodiment, and is, for example, an output signal of an ultrasonic Doppler sensor for directly measuring the wetness of the road surface or an indirect estimation of the road surface state. Process the status signal of the wiper switch or headlight switch.

[0013]

FIG. 1 is a functional block diagram showing a self-vehicle position detecting apparatus for automatic traveling according to the first invention. The operation of the first invention will be described with reference to this drawing. The imaging unit 101 is mounted on the vehicle and captures a road image ahead of the vehicle.
For example, a video camera. Road shape storage means 102
Stores the shape of the road as a road model, and is, for example, a memory. This memory stores captured images, initial conditions, road models, parameters representing camera attitude, and the like, and also functions as a work area when performing arithmetic processing.

The coordinate conversion means 103 converts a coordinate system of a point to be processed on a road to obtain a coordinate value on an image. The area setting means 104 sets a processing area near a straight line or a curved line representing the road shape at the time of the previous calculation.

The feature point coordinate detecting means 105 includes the image pickup means 1
In step 01, the captured image information is scanned, and for example, position coordinates of a feature point representing a lane are obtained using edge information obtained by spatial differentiation processing.

The lane extracting means 106 refers to the coordinates of the characteristic points extracted by the characteristic point coordinate detecting means 105 and approximates the pixels representing the characteristic points by using, for example, the least squares method so as to form a straight line or a curve. .

The change amount calculation means 107 calculates the coordinate values of the feature points extracted by the feature point coordinate detection means 105 and the image coordinates obtained by converting the road model stored in the road shape storage means 102 by the coordinate conversion means 103. Compare with the value
A change component of a parameter (curvature and gradient) representing the road shape and parameters (deviation from the reference position, height, yaw angle, pitch angle, and roll angle) representing the attitude of the imaging unit 101 are calculated.

The parameter updating means 108 calculates the change amount calculated by the change amount calculating means 107 and the lane extracting means 10.
6, the road model stored in the road shape storage means 102 and the coordinate transformation means 10 based on the lane shape extracted in Step 6.
The parameters such as 3 are updated.

The output signal selecting means 109 determines the degree of certainty of the information based on the processing result of the feature point coordinate detecting means 105, the deviation from the reference position processed by the parameter updating means 108 and the curvature of the road. Then, a signal serving as a basis for steering control is output to the following vehicle control side according to the certainty factor. Hereinafter, the vehicle control side performs the steering control based on the output of the output signal selection means 109.

As described above, the first aspect of the present invention calculates the deviation from the reference position using the white line data in a plurality of areas up to a distant place, and when outputting the deviation to the steering control side, the distance on the image is calculated. If the white line information exists in a place with high accuracy, set the data confidence to a high value.If it does not exist, set the data confidence to a low value. To change the signal output to the steering control. As a result, it is possible to output to the steering control a signal with little variation that causes fluctuations in the steering control, particularly with a small number of change points.

FIG. 6 is a functional block diagram showing a self-vehicle position detecting device for automatic traveling according to the second invention. The operation of the second invention will be described with reference to FIG. The configurations and functions of the imaging unit 101, road shape storage unit 102, coordinate conversion unit 103, region setting unit 104, feature point coordinate detection unit 105, lane extraction unit 106, change amount calculation unit 107, and parameter update unit 108 are described above. This is the same as the first invention.

The road surface condition detecting means 110 processes a signal which detects the wetness or illuminance of the road surface condition. For example, an output signal of an ultrasonic Doppler sensor for directly measuring the wetness condition of the road surface, or from a rainfall condition. A state signal of a wiper switch for indirectly estimating a wet state of a road surface or a state signal of a headlight switch for estimating that a traveling road in a tunnel or at night is low in illuminance is processed.

The output signal selecting means 109 outputs the processing result of the feature point coordinate detecting means 105, the deviation from the reference position processed by the parameter updating means 108, the curvature of the road, and the road surface state detecting means 110. Is determined on the basis of the output result of the above, and a signal serving as a basis for steering control is output to the subsequent vehicle control side according to the certainty. Hereinafter, the vehicle control side performs the steering control based on the output of the output signal selection means 109.

As described above, according to the second invention, the deviation from the reference position is calculated using the white line data in a plurality of areas up to a distance, and when the deviation is output to the steering control side, the distance on the image is calculated. If the white line information exists in a place with high accuracy, the data confidence is set high, and if it does not exist, the data confidence is set low. A signal to be output to the steering control side is changed according to three output results such as a road surface condition. As a result, it is possible to output to the steering control a signal with little variation that causes fluctuations in the steering control, particularly with a small number of change points.

[0025]

According to the first aspect of the present invention, since the steering control is performed based on the calculated value with high distance accuracy, for example, when the white line is a broken line on a straight road with almost no steering angle, Variations in output values due to estimation errors and the like, particularly the number of output value change points, can be reduced, and as a result, fine fluctuations during steering control can be suppressed.

According to the second aspect of the present invention, on a curved road having a small radius of curvature which must always travel with a steering angle, the latest white line information in a plurality of areas up to a long distance is always used. Since steering control is performed based on deviation information from the reference position, smooth steering control can be performed not only on a straight road but also on a curved road.

According to the third aspect of the present invention, the signal to be output to the control side is calculated in consideration of the state of the traveling road surface.
Smooth steering control is possible even in a bad environment such as nighttime or rainy weather.

[0028]

Embodiments of the present invention will be described below with reference to the drawings. First Embodiment FIG. 2 is a block diagram showing a first embodiment of the present invention. In FIG. 2, an imaging unit 1 is, for example, a video camera, is installed in front of a vehicle, and captures a road image ahead of the vehicle to convert it into an electric signal.

The image processing unit 2 is constituted by, for example, a microcomputer, and performs image processing by inputting a video signal from the imaging unit 1 to deviate from the reference position of the own vehicle, that is, in the traveling lane. Detect horizontal position.

The vehicle control unit 3 performs the steering control of the automatic traveling control based on the detection result of the image processing unit 2. The display unit 4 displays a detection result of the image processing unit 2, and is, for example, a CRT display device or a liquid crystal display device.

The processing contents of the information in the image processing unit 2 are as follows.
As described with reference to the functional blocks in FIG. 1, each processing content will be described in detail.

FIG. 3 is an overall flowchart of the present embodiment. First, when the self-vehicle position detecting device of the present embodiment is turned on by operating the self-vehicle position detecting device (S).
2), initialization is performed (S4). Here, the initialization includes processing such as clearing of a work area of a memory. In addition, as an operation performed by a person, initial adjustment of the imaging unit 1, setting of initial values of road model parameters representing the shape of the lane, and setting of initial values of camera posture parameters representing the mounting position and direction of the camera are performed. .

Next, the signal for the road image ahead of the vehicle, which is received on the imaging surface by the imaging unit 1, is converted into an analog signal (S6), and the analog signal is converted into a digital signal by the A / D converter in the image processing unit 2. Convert (S
8).

Next, in the lane detecting process (S10), a white line indicating the boundary of the lane is detected, and the shape of the lane is approximated by a straight line or a curved line. Then, the amount of change of each parameter (road shape parameter and camera attitude parameter) from the previous time is obtained, a new road model is created from the amount of change, the result is stored in the memory, and the next image information is obtained. Therefore, the process returns to the image information capturing process (S6).

The output signal selecting means (S12) performs steering control in the current traveling lane based on the presence or absence of a white line near the own vehicle and the curvature of the road among the results of the lane detection processing (S10). The deviation data from the optimal reference position is selected and output to the vehicle control side.

In the steering control process (S14), the steering control is performed by controlling the steering actuator based on the output from the initial signal selection process (S12).

Note that the power supply is turned on (S
2) and the initial setting process (S4) need only be executed once immediately after the system is started, and the apparatus operates by repeating a series of processes thereafter.

Next, the lane detection processing (S10) and the output signal selection processing (S1), which are important points in this embodiment, are performed.
The details of the process 2) will be described in detail.

First, regarding the lane detection processing (S10),
The processing will be described with reference to FIG.

First, an edge image creation process (S100)
In 2), a horizontal edge image and a vertical edge image are created by performing an edge extraction process using, for example, a 3 × 3 pixel spatial differential filter (for example, a SOBEL filter).

Next, in the coordinate conversion process (S1004),
In the road model stored in the memory, a target point that is a predetermined distance from the own vehicle in the road coordinate system is subjected to coordinate conversion to an image coordinate system using a camera attitude parameter or the like, and the image coordinates of the target point are calculated. A value (hereinafter referred to as a model coordinate value) is calculated.

Next, in the detection area setting processing (S1006), the number of pixels corresponding to a fixed distance (for example, 2 m) in the road coordinate system is calculated using, for example, the focal length of the CCD camera, and the calculated number of pixels is calculated. A white line detection area centered on the model coordinate value is set on the edge image with the size of.

Next, in the matching processing (S1008), the area set in the detection area setting processing (S1006) is scanned left and right or up and down. At this time, the scanning direction and the type of the edge image to be scanned depend on the shape of the lane. If the white line appears to be nearly vertical on the image like a straight road, the vertical edge image is scanned in the horizontal direction, and the distant curved road is used. If it looks nearly horizontal like the white line, the horizontal edge image is scanned in the vertical direction. Then, a combination of a plus edge of a certain value or more and a minus edge of a certain value or less separated by the number of pixels corresponding to the width of the white line is extracted, and the coordinate value of one of the edges (hereinafter referred to as a calculated coordinate value) is calculated. Ask.

At this time, which edge coordinate value is used is the same throughout the processing. Then, the calculated coordinate values and the model coordinate values, which are the edge coordinate data thus obtained, are used in the subsequent processing.

Next, parameter estimation processing (S1010)
Then, a change amount of each parameter is calculated from the calculated coordinate value and the model coordinate value. At this time, the amount of deviation between the pixel of the model coordinate value and the pixel of the calculated coordinate value is the basis of the amount of change. A yaw angle, which is a camera posture parameter,
The amount of change in the pitch angle and the roll angle from the previous processing time is calculated.

Finally, the parameter update processing (S101)
In 2), each parameter is updated based on the amount of change calculated in the parameter estimation processing (S1010), and among the updated parameters, the deviation from the reference position and the curvature of the lane are output to the subsequent processing, and , And the approximate expression representing the shape of the lane is stored in the memory as a new road model.

Next, FIG. 5 is a flowchart showing the contents of the output signal selection processing (S12).
This will be described with reference to FIG. 9, which is an example of a road image in the present embodiment.

First, the white line position determination processing (S1202)
Then, the lane detection processing (S
A certainty factor is set as an index indicating the certainty of data with respect to the deviation from the reference position, which is one of the output results from 10).

This is because the white line 5 is detected in the lane detection processing (S10).
Among the detection areas 6 set for detecting the vehicle, for example, an area closest to the host vehicle is set as the determination area 7. If a white line exists in this region (FIG. 9A), the degree of certainty about the deviation from the reference position is set high, and if there is no white line (FIG. 9B), This certainty factor is set low.

The determination area 7 does not need to be one area closest to the host vehicle, but may be set at a plurality (for example, four) as shown in FIG. At this time, if the interval between the first determination region 8, the second determination region 9, the third determination region 10, and the fourth determination region 11 is, for example, 2 m, the solid line portion is 8 m and the discontinuous portion is 1
For the broken line of 2 m, the degree of certainty when the white line information is present in all four areas is the highest, and hereinafter, the first determination area 8, the second determination area 9, and the third determination area 10, if it exists in the first determination area 8 and the second determination area 9, if it exists only in the first determination area 8, the second determination area 9 and the third determination area 10 Fourth determination area 1
1, the third determination area 10 and the fourth determination area 11, the fourth determination area 11 only, and the non-existence of any of the four determination areas. May be set low.

Next, in the lane shape determining means (S1204), based on the curvature information of the road model, which is one of the output results from the lane detection processing (S10), the lane shape is determined in accordance with the radius of curvature of the curved road. , Classify the degree of the curve. For example, based on a radius of curvature of 300 m as a reference, it is roughly classified into a case where the radius is larger (closer to a straight road) and a case where the radius is smaller (a sharper curve). It should be noted that the degree of the curve is not necessarily limited to being roughly divided into two.

Finally, output signal update processing (S1206)
Then, data to be output to the steering control side is determined from the degree of certainty of the deviation from the reference position and the degree of the curve. For example, when the degree of the curve is close to a straight road (for example, the radius of curvature is 300 m or more) and the degree of certainty of the deviation from the reference position is high, the deviation from the reference position is used as the overall output value for steering control. Output to

When the degree of the curve is close to a straight road and the confidence of the deviation from the reference position is low, the deviation from the reference position with the latest confidence in the processing up to that time is low. Is output to the steering control as an overall output value. That is, when the white line is a dashed line on a running road close to a straight road, the signal output to the steering control may be updated anew, or may retain the previous result without being updated.

On the other hand, when the degree of the curve is steep (for example, the radius of curvature is 300 m or less), a newly updated signal is always used as the entire output value regardless of the degree of certainty of deviation from the reference position. Output to the side.

According to the above-described embodiment, when calculating the deviation from the reference position using the white line data in a plurality of areas up to a distant place and outputting the deviation to the steering control side, the distance accuracy on the image is calculated. If the white line information exists in the area near the own vehicle, the confidence level of the calculated data is set high. If the white line information does not exist in the area near the own vehicle, the calculated data Since the optimal signal is selected according to the confidence and the degree of the curve ahead of the road by setting the confidence to be low, it is possible to realize a device capable of outputting a signal for performing smooth steering control. Can be.

Second Embodiment FIG. 7 is a block diagram showing a second embodiment of the present invention. The second embodiment differs from the first embodiment in that a road surface state detector 12 is added. The configuration is the same.

The road surface condition detecting unit 12 is, for example, an ultrasonic Doppler sensor, a wiper switch, or a headlight switch, and is a sensor that directly or indirectly detects a wet state or an illuminance state of the road surface.

In the self-vehicle position detecting device for automatic traveling according to the present embodiment, the operation processing is executed in accordance with the overall flowchart shown in FIG. 3 as in the first embodiment described above.

In particular, since the processing in the output signal selection processing routine in step 12 in FIG. 3 is different, it will be described with reference to FIG. 8 and FIG. 9 which is an example of a road image.

First, the white line position determination processing (S1202)
Then, the lane detection processing (S
A certainty factor is set as an index indicating the certainty of data with respect to the deviation from the reference position, which is one of the output results from 10).

This is because the white line 5 in the lane detection processing (S10)
Among the detection areas 6 set for detecting the vehicle, for example, an area closest to the host vehicle is set as the determination area 7. If a white line exists in this region (FIG. 9A), the degree of certainty about the deviation from the reference position is set high, and if there is no white line (FIG. 9B), This certainty factor is set low.

The determination area 7 does not need to be one area closest to the host vehicle, but may be set at a plurality (for example, four) as shown in FIG. At this time, if the interval between the first determination region 8, the second determination region 9, the third determination region 10, and the fourth determination region 11 is, for example, 2 m, the solid line portion is 8 m and the discontinuity portion is 1
For the broken line of 2 m, the degree of certainty when the white line information is present in all four areas is the highest, and hereinafter, the first determination area 8, the second determination area 9, and the third determination area 10, if it exists in the first determination area 8 and the second determination area 9, if it exists only in the first determination area 8, the second determination area 9 and the third determination area 10 Fourth determination area 1
1, the third determination area 10 and the fourth determination area 11, the fourth determination area 11 only, and the non-existence of any of the four determination areas. May be set low.

Next, in the lane shape determining means (S1204), based on the curvature information of the road model, which is one of the output results from the lane detection processing (S10), the lane shape is determined in accordance with the radius of curvature of the curved road. , Classify the degree of the curve. For example, based on a radius of curvature of 300 m as a reference, it is roughly classified into a case where the radius is larger (closer to a straight road) and a case where the radius is smaller (the curve is tighter). It should be noted that the degree of the curve is not necessarily limited to being roughly divided into two.

Next, in the wet state determination processing (S1206), the wet state of the road surface is estimated from an output signal from a road surface state detection sensor, for example, an ultrasonic Doppler sensor or a state signal of a wiper switch. For example, when the wiper switch has four stages of OFF, intermittent, Lo, and Hi, the wet state of the road surface is roughly classified into two stages in association with the state of the wiper switch. And the output signal “0” is changed to Lo or H
In the case of i, the road surface is estimated to be wet, and the output signal “1” is assigned. Similarly, the output signal from the ultrasonic Doppler sensor is roughly divided into two stages, dry and wet. However, the wet state is not necessarily limited to roughly divided into two.

Next, in the illuminance state determination process (S1208), the illuminance of the road surface is estimated from a state signal of, for example, a headlight switch which is a road surface state detection sensor. For example, when the headlight switch has four stages of OFF, small, Lo, and Hi, the illuminance state of the road surface is roughly classified into two stages in association with the state of the headlight switch. Estimates that the output signal "0" is bright up to the far side of
In the case of o or small, it is presumed that the far side of the traveling path is dark, and the output signal “1” is assigned. However, the illuminance state is not necessarily limited to two.

Finally, output signal update processing (S1210)
Then, data to be output to the steering control side is determined using the degree of certainty of the deviation from the reference position and the degree of the curve, as well as the wet state and the illuminance state of the road surface. For example, when the degree of the curve is close to a straight road (for example, the radius of curvature is 300 m or more) and the degree of certainty of the deviation from the reference position is high, the deviation from the reference position is used as the overall output value for steering control. Output to

If the degree of the curve is close to a straight road and the degree of certainty of the deviation from the reference position is low, the deviation from the reference position with the highest degree of certainty in the processing up to that time is performed. Is output to the steering control as an overall output value. That is, when the white line is a dashed line on a running road close to a straight road, the signal output to the steering control may be updated anew, or may retain the previous result without being updated.

On the other hand, when the degree of curve is sharp (for example, the radius of curvature is 300 m or less), when the wet state signal is “0” and the illuminance state signal is “0”,
Regardless of the degree of certainty of deviation from the reference position, a newly updated signal is always output to the steering control side as an overall output value. When either the wet state signal or the illuminance state signal is “1”, similarly to the case of the straight road, when the certainty of the deviation from the reference position is high, the deviation from the reference position is determined. Is output to the steering control as an overall output value. When the degree of certainty of the deviation from the reference position is low, the deviation from the reference position with the highest degree of certainty in the processing up to that point is output to the steering control as the entire output value.

According to the above-described embodiment, when calculating the deviation from the reference position using the white line data in a plurality of areas up to a distant place and outputting the deviation to the steering control side, the distance accuracy on the image is calculated. If the white line information exists in the area near the own vehicle, the confidence level of the calculated data is set high. If the white line information does not exist in the area near the own vehicle, the calculated data Since the optimal signal is selected according to the confidence and the degree of the curve ahead of the road by setting the confidence to be low, it is possible to realize a device capable of outputting a signal for performing smooth steering control. Can be.

The embodiments described above are described for facilitating the understanding of the present invention, and are not described for limiting the present invention. Therefore, each element disclosed in the above embodiment is intended to include all design changes and equivalents belonging to the technical scope of the present invention.

[Brief description of the drawings]

FIG. 1 is a functional block diagram showing a self-vehicle position detecting device for automatic traveling according to a first invention.

FIG. 2 is a block diagram showing a first embodiment of the present invention.

FIG. 3 is a flowchart illustrating the calculation processing of the entire first embodiment.

FIG. 4 is a flowchart showing a sub-routine for lane detection in FIG. 3;

FIG. 5 is a flowchart showing a subroutine for selecting an output signal in FIG. 3;

FIG. 6 is a functional block diagram showing a self-vehicle position detecting device for automatic traveling according to a second invention.

FIG. 7 is a block diagram showing a first embodiment of the present invention.

FIG. 8 is a flowchart illustrating a subroutine for selecting an output signal according to the first embodiment.

FIGS. 9A and 9B are diagrams illustrating an example of a road image according to the embodiment, in which FIG. 9A illustrates an example of high data certainty and FIG. 9B illustrates an example of low data certainty.

FIG. 10 is a diagram illustrating another example of the certainty factor setting in the embodiment.

[Explanation of symbols]

 REFERENCE SIGNS LIST 1 imaging unit 2 image processing unit 3 vehicle control unit 4 display unit 5 white line 6 detection region 7 determination region 8 first determination region 9 second determination region 10 third determination region DESCRIPTION OF SYMBOLS 11 ... Fourth determination area 12 ... Road surface state detection part 101 ... Imaging means 102 ... Road shape storage means 103 ... Coordinate conversion means 104 ... Area setting means 105 ... Feature point coordinate detection means 106 ... Lane extraction means 107 ... Change amount calculation Means 108: Parameter updating means 109: Output signal selecting means 110: Road surface state detecting means

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁶ Identification code FIG08G 1/16 G06F 15/62 380 // B62D 137: 00

Claims (3)

[Claims] 1. A self-vehicle position detecting device for automatic driving related to steering control of a vehicle, comprising: an image pickup means mounted on the vehicle for picking up a road image ahead of the vehicle; and a road shape storing a shape of the road as a road model. Storage means, coordinate conversion means for converting a coordinate system to obtain a coordinate value on an image, area setting means for setting an area near a straight line or a curvilinear expression in a previous calculation, and an image taken by the imaging means A feature point coordinate detection unit that scans information and obtains position coordinates of a feature point representing a lane; and refers to the coordinates of the feature point extracted by the feature point coordinate detection unit, thereby defining a pixel representing the feature point as a straight line or a curve. Lane extraction means for obtaining a lane constituted by connecting to form a coordinate value of the feature point extracted by the feature point coordinate detection means,
By comparing the road model stored in the road shape storage unit with the image coordinate values converted by the coordinate conversion unit, a curve, a slope, and other parameters representing a road shape,
A deviation amount from a reference position, a height, a yaw angle, a pitch angle, a roll angle, and other parameters representing the attitude of the imaging unit, and at least a deviation amount from the reference position and a change amount calculating unit that calculates a change component of curvature. A parameter for updating the road model and the coordinate conversion means and other parameters stored in the road shape storage means from the change amount calculated by the change amount calculation means and the lane shape extracted by the lane extraction means; Updating means, output signal selecting means for outputting the deviation from the reference position to the vehicle control side, the output signal selecting means, based on the processing result of the feature point coordinate detecting means, in an area near the own vehicle A signal to be output to the vehicle control side according to the degree of certainty determined from the reference position updated by the parameter updating means using the feature point information representing the lane of the vehicle. Automatic traveling vehicle position detecting apparatus characterized by update. 2. The method according to claim 1, wherein the output signal selection means updates a signal to be output to the vehicle control side in accordance with the degree of curvature ahead of the vehicle in addition to the certainty of deviation from the reference position. Item 4. The self-vehicle position detecting device for automatic traveling according to Item 1. 3. A road surface condition detecting means for directly or indirectly detecting a road surface condition of a traveling road, wherein the output signal selecting means includes a certainty degree of deviation from a reference position and a degree of curvature in front of the vehicle. 3. A signal output to a vehicle control side is updated according to a road surface condition of a traveling road detected by the road surface condition detecting means.
The self-vehicle position detecting device for automatic driving according to the above.