JP3624768B2 - Parallel parking assistance device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a parallel parking support apparatus, and more particularly to an apparatus for recognizing a target parking frame by image processing.
[0002]
[Prior art]
2. Description of the Related Art Conventionally, there has been proposed an apparatus that displays a rear view of a vehicle on a monitor when the driver cannot see a target location due to the blind spot of the vehicle when the vehicle is moving backward. For example, Japanese Patent Publication No. 2-36417 discloses a television camera that captures the rear of a vehicle, a monitor television that displays an image captured by the television camera, a sensor that outputs an information signal related to a tire steering angle, A vehicle rear monitoring and monitoring device is disclosed that includes a circuit that generates a marker signal in accordance with the information signal and displays the marker superimposed on a television screen. In this device, the steering angle data of the tire and the marker position data along the reverse direction of the vehicle corresponding to the steering angle are stored in the ROM, and the predicted backward trajectory of the vehicle according to the steering angle at that time is the marker position data. As a row, it is displayed on the television screen in a manner superimposed on the video photographed by the television camera.
[0003]
According to such a device, when the vehicle is moving backward, the predicted backward trajectory of the vehicle according to the steering angle is displayed on the monitor TV screen together with the field of view of the road behind the vehicle, so the driver turns backward. The vehicle can be moved backward by operating the steering wheel while watching the TV screen.
[0004]
[Problems to be solved by the invention]
For parallel parking, for example, the vehicle must be moved backward in parallel with the road, the handlebar must be turned in an appropriate position to enter the parking space, and the handlebar must be turned back in the opposite direction to guide the vehicle to the target parking position. There is. However, in the conventional rear monitoring and monitoring device, the driver can determine where to start or turn the steering wheel, and to determine the steering amount only by looking at the rear view and the expected backward trajectory of the vehicle on the TV screen. It was difficult to determine how much it should be, and there was a problem that sufficient support for parallel parking could not be provided.
This invention was made in order to solve such a problem, and provides the parallel parking assistance apparatus which can grasp | ascertain easily the timing and amount of steering when a driver | operator parks in parallel. Objective.
[0005]
[Means for Solving the Problems]
A parallel parking assist apparatus according to claim 1 of the present invention detects a steering angle of a steering wheel, a camera that captures the rear of the vehicle, a monitor that is disposed at the driver's seat of the vehicle and that displays an image by the camera when the vehicle is moving backward. Steering sensor, superimpose unit that superimposes and displays on the monitor screen a guide display for assisting vehicle driving during parallel parking, and image processing by the cameraBy recognizing the two straight lines included in the parking frame, the intersection point between them is used as the target point, and the target point is tracked according to the movement of the video as the vehicle moves backward,An image processing unit that outputs a steering timing and a steering amount necessary for parallel parking based on the positional relationship between the target point and the guide display is provided.
[0006]
Claim2The parallel parking support device according to claim1In this apparatus, the image processing unit predicts and corrects the position of the target point when the movement of the target point greatly deviates as the vehicle moves backward.
[0007]
Claim3The parallel parking support apparatus according to claim 1Or 2The apparatus further includes a voice selection unit that provides a steering guidance to the driver by emitting a voice corresponding to the steering timing and the steering amount output from the image processing unit.
Claim4The parallel parking support device according to claim3In this apparatus, the image processing unit outputs the steering timing and the steering amount to the superimposing unit together with the voice selection unit, and displays on the monitor screen the display corresponding to the steering guidance.
Claim5The parallel parking support apparatus according to claim 14The apparatus according to any one of the above, further comprising a parallel parking mode switch disposed in the driver's seat of the vehicle, and a guide display is displayed on the monitor screen by the superimposing unit when the parallel parking mode switch is turned on. Is.
According to a sixth aspect of the present invention, there is provided a parallel parking assist apparatus that detects a rear angle of a vehicle, a monitor that is disposed in a driver's seat of the vehicle and that displays an image of the camera when the vehicle moves backward, and detects a steering angle of the steering wheel. A steering sensor, a superimpose section that superimposes and displays on the monitor screen a guide display for assisting driving of the vehicle during parallel parking, and recognizes a target point for parking by image processing of the video from the camera An image processing unit that outputs a steering timing and a steering amount necessary for parallel parking based on the positional relationship between the target point and the guide display, and the guide display indicates the steering angle of the steering wheel detected by the steering sensor. It includes a handle mark that is moved and displayed on the monitor screen in response to the parking mark. Retracting the vehicle while the handle until the target point is coincident for, and performs parallel parking by then full off retracts the handle in the opposite direction.
Furthermore, the parallel parking assist device according to claim 7 is the device according to claim 6, wherein the guide display includes an eye mark fixedly displayed at a specific position with respect to the vehicle, and the eye mark is displayed on a monitor screen. When the vehicle is moved backward with the steering wheel operated until the steering wheel mark and the target point for parking coincide with each other, the target point for parking coincides with the eye mark so that the vehicle steering wheel is in the opposite direction. It indicates the arrival at the position where it is fully cut.
[0008]
In the parallel parking assistance device according to claim 1, the image processing unit includes:, Find the intersection of two straight lines included in the parking frame as a target point, track this target point as the vehicle moves backward,Based on the positional relationship between the target point and the guide display superimposed on the monitor screen by the superimposing unit, the steering timing and the steering amount necessary for parallel parking are output.
Claim2In the parallel parking support device according to claim 1, the claim1In this apparatus, the image processing unit predicts and corrects the position of the target point when the movement of the target point greatly deviates.
Claim3In the parallel parking assist device according to claim 1,Or 2In this apparatus, the voice selection unit performs steering guidance by generating a voice corresponding to the steering timing and the steering amount output from the image processing unit.
Claim4In the parallel parking support device according to claim 1, the claim3In this apparatus, the image processing unit outputs the steering timing and the steering amount also to the superimpose unit, and displays on the monitor screen the display corresponding to the steering guidance.
Claim5In the parallel parking assistance device according to claim 1,4In the device according to any one of the above, a guide display is displayed on the monitor screen when the parallel parking mode switch is turned on.
In the parallel parking assist apparatus according to claim 6, the image processing unit recognizes a target point for parking by performing image processing on the video image from the camera, and the target point and the superimposing unit superimpose and display on the monitor screen. The steering timing and the steering amount necessary for parallel parking are output based on the positional relationship with the displayed guide display, and are moved and displayed on the monitor screen from the parking stop position according to the steering angle of the steering wheel. Parallel parking is performed by retracting the vehicle while operating the handle until the handle mark coincides with the target point for parking, and then turning the handle fully in the opposite direction and moving backward.
The parallel parking assist device according to claim 7, wherein, in the device according to claim 6, when the vehicle is moved backward while operating the handle until the handle mark coincides with the target point for parking on the monitor screen. In addition, when the target point for parking coincides with the eye mark, the vehicle reaches the position where the steering wheel is fully turned in the opposite direction.
[0009]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of the present invention will be described below with reference to the accompanying drawings.
Embodiment 1 FIG.
As shown in FIG. 1, a camera 2 that captures a field of view behind the vehicle 1 is attached to the rear of the vehicle 1. A rear bumper 3 of the vehicle 1 is placed at the near side end of the field of view of the camera 2. A monitor 4 comprising a color type liquid crystal display is disposed in the driver's seat of the vehicle 1 and is normally used as a display device of a navigation device, and a shift lever 5 provided in the driver's seat is operated to R (reverse position). Then, the video from the camera 2 is displayed. The front wheel 6 as a steering wheel is steered by operating the handle 7.
[0010]
FIG. 2 shows a drawing form on the monitor 4. The vehicle 1 is stopped parallel to the roadside 8. The column guideline 9 guides the position where the steering operation is started, and the vehicle 1 is moved backward straight so that this and the target point TP overlap each other. When the vertical guideline 9 and the target point TP overlap, the operation is stopped, and the handle 7 is operated until the target mark TP and the handle mark 10 moving on the vertical guideline 9 according to the steering amount of the handle 7 overlap. When the handle mark 10 and the target point TP overlap each other, the handle 7 is held at the position as it is, and it moves backward until the eye mark 11 and the target point TP coincide. When the eye mark 11 coincides with the target point TP, the handle 7 is fully turned in the opposite direction and moved backward. The vehicle is retracted while paying attention to the rear, and the parallel parking is completed when the vehicle 1 is parallel to the roadside 8.
[0011]
FIG. 3 shows the configuration of the parallel parking assist apparatus according to Embodiment 1 of the present invention. The rear image captured by the camera 2 is output from the external input processing unit 12 to the image processing unit 13 and the superimposing unit 14. The superimposing unit 14 superimposes a guide display as shown in FIG. 2 and outputs it to the monitor 4 together with the rear video. Further, the external input processing unit 12 receives a signal corresponding to the shift position from the shift switch 15, and if it is R (reverse position), instructs the superimpose unit 14 to display a rear image and performs image processing on the image processing unit 13. A signal is output so that When it becomes other than R, a signal is output so as to stop them. The steering sensor 16 outputs a pulse when the handle 7 is operated. The external input processing unit 12 receives this, counts it, and outputs it to the steering angle calculation unit 17 of the image processing unit 13.
[0012]
Start from a state where the vehicle 1 is parked in parallel with the roadside 8. In the stop image processing unit 13, first, the binarization / edge detection unit 18 binarizes the rear image and detects edges. As shown in FIG. 4, a histogram region 21 is set at a position where the vertical white line 20 of the substantially trapezoidal parking frame 19 crosses vertically, and a histogram is obtained. As shown in FIG. 5, the density level exceeding the number of pixels of p% for the first time from the maximum density level side of the obtained histogram is set as a binarization threshold Th. As shown in FIG. 6, binarization is performed only in a designated processing area 22 that does not include a horizontal line in front of the parking frame 19. Edges in the vertical and horizontal directions are obtained from the binarized image and are combined to obtain an edge image. As shown in FIG. 6, since processing is performed only in the limited processing region 22, processing is lighter than processing the entire image. Since the parking frame 19 is positioned at the top of the screen when parallel parking starts, the processing area 22 can be set at the top of the screen as shown in FIG.
Further, as shown in FIG. 7, the processing area 22 may be a small area including at least the corner A of the parking frame 19. However, it is assumed that the inclination of the straight line in the horizontal direction is substantially constant regardless of the front-rear position of the vehicle 1.
[0013]
The switching unit 23 switches the output destination of the edge image. From the start of processing, for example, the first frame to the tenth frame are output to the shape recognition unit 24 and thereafter are output to the approximate line extraction unit 25. The shape recognition unit 24 is for detecting the target point TP as reliably as possible at the start of the process, and recognizes the shape of the parking frame 19 to detect the target point TP. The positions of the points P <b> 1 to P <b> 4 in FIG. 8 within the processing region 22 vary depending on the distance of the vehicle 1 from the roadside 8. Therefore, the target point TP is detected by learning a plurality of states in advance and determining which state the processed image is close to.
[0014]
Next, specific processing will be described. The processed image is a binary image in which the edge portion indicated by the black line in FIG. 8 is white and the other portions are black. First, the points P1 and P2 are searched. A white pixel is searched for in the right direction from the origin O of the processing area 22, and a point P1 is found first. Similarly, a white pixel is searched upward from the origin O to find a point P2. Depending on the distance of the vehicle 1 from the roadside 8, the position of the points P1 and P2, the slope m1 of the straight line connecting the points P1 and P3, the slope m2 of the straight line connecting the points P2 and P4, and the average mc of these slopes m1 and m2 For example, as shown in FIG. Let me learn as 1-3. Then, it is determined which shape is close from the positions of the detected points P1 and P2. Here, it is assumed that no. The shape is 2.
[0015]
Next, the points P3 and P4 are searched. First, a middle point between the point P1 and the origin O is set as Pc, and a white pixel is searched in the direction of the inclination M2c therefrom. If a white pixel is found, the y coordinate of the point Pc ′ is stored. A point shifted by the number of pixels s from the point P1 to the left is defined as P1 ', and a white pixel is searched in the direction of the slope M21 from there to obtain a point P3'. Similarly, a white pixel is searched in the direction of the inclination M22 from a point P2 'that is shifted from the point P2 to the right by the number of pixels t to obtain a point P4'. At this time, if the difference between the y-coordinates of the points P3 'and P4' and the y-coordinate of the point Pc 'is large, it is determined that there is some noise and the search is started again. A point shifted from the point P3 'to the right by the number of pixels s is P3, and a point shifted from the point P4' to the left by the number of pixels t is P4. This process is repeated for 10 frames. FIG. 10 shows a flowchart of the shape recognition unit 24 and the result determination unit 26. If the detection is normally performed, the coordinates of each point, the slope m1 of the straight line connecting the points P1 and P3, and the slope of the straight line connecting the points P3 and P4 are output as data. When there is an undetectable point among the points P1 to P4, P1 ′ to P4 ′, Pc, and Pc ′ in all the processing for 10 frames, the result judging unit 26 provides a voice to that effect as an error. The signal is output to the voice selection unit 27, and the voice is emitted from the speaker 28.
[0016]
Next, the approximate straight line extracting unit 25 performs the approximate straight line extracting process shown in the flowchart of FIG. 11 on the images after the eleventh frame after all the shape recognizing processes by the shape recognizing unit 24 are completed. . Here, approximate lines of the straight line L0 connecting the points P1 and P3 and the straight line L1 connecting the points P3 and P4 are extracted. First, a contour line is extracted from the input edge image. As shown in FIG. 12, a white pixel (for example, a density level 255) that is the starting point SP of the contour line is scanned and searched. When the starting point SP is found, the outline is traced in the horizontal direction. At this time, the contour line is labeled with a value other than the density level 255, and the white pixel (density level 255) positioned below the point is erased for each point on the contour line as indicated by the arrow. When the horizontal direction ends, the vertical outline is similarly labeled with a value other than the density level 255, and the white pixel located on the right side of the outline is deleted. Next, an approximate straight line of the contour line is obtained by Hough transform on the extracted contour line in the coordinate system in the processing region 22. Thereafter, the search for the start point SP is started again from the position of the start point SP thus obtained, but the outline is labeled with a value other than the density level 255, and the white pixels below and to the right of the outline are erased. A point on this edge is not detected as the start point SP. This process is continued until the search position of the start point SP comes to the lower right corner of the processing area 22.
[0017]
Next, Hough conversion will be described. One straight line uses the distance r from the origin and the inclination θ from the x-axis.
r = x · sin θ + y · cos θ (1)
It is represented by When a plurality of points on a certain straight line are mapped to the (r, θ) space using the equation (1), the curves in the (r, θ) space intersect at one point. This intersection (r, θ) represents the slope of the straight line and the distance from the origin. By using this property, the parameters of the approximate straight line of the contour line can be obtained by mapping the points on the contour line to the (r, θ) space and obtaining the intersection of the curves in the (r, θ) space.
[0018]
Actually, a two-dimensional array corresponding to θ and r is prepared. This becomes the (r, θ) space. In advance, for example, “0” is put at all the positions (r, θ), θ is changed by the expression (1), r at that time is obtained, and “1” is set at the position (r, θ). Add and count. If this process is performed on all points on the contour line, the count value of the position of the intersection (r, θ) becomes larger than the others, so this point may be extracted as a straight line parameter. At this time, the angle obtained in the previous approximate straight line extraction process is received from the data update unit 29 as a reference angle, and an intersection point is obtained within a range of several degrees before and after the angle. That is, a linear parameter is extracted only when the maximum count value (r, θ) is within the angle range. By doing so, only those close to the slope of the white line are extracted.
[0019]
When the two approximate lines L0 and L1 are extracted, the intersection calculation unit 30 calculates an intersection P (corresponding to the point A in FIG. 2) of these approximate lines. Here, since it is the coordinate system of the processing area 22, it is necessary to convert it to the coordinate system of all images. Assuming that the coordinate system of the processing area 22 is (x, y) and the coordinate system of all images is (X, Y), a relational expression is obtained from FIG.
X = sp_x+ X
Y = sp_y+ Y
R = X · sin θ + Y · cos θ
r = x · sin θ + y · cos θ
Can be used to convert intersection and straight line parameters to their respective coordinate systems. The intersection and straight line data are output to the target point calculation unit 31. If an approximate straight line cannot be extracted and no intersection point is obtained, the information is output to the target point calculation unit 31.
[0020]
The target point calculation unit 31 calculates the target point TP shown in FIG. 2 from the obtained intersection point P. As shown in FIG. 14A, a point POL on the vertical approximate line L0 is obtained from the intersection point P, and the coordinates of these points P and POL are transformed onto the road surface as shown in FIG. 14B. On the road surface, a point that is translated from the point P by the width W of the parking frame 19 in a direction perpendicular to the straight line connecting the point P and the point POL is obtained as the target point TP. The target point TP on the road surface is subjected to coordinate transformation again to obtain the target point TP on the screen as shown in FIG. In this process, until the target point TP overlaps the column guide line 9 in FIG. 2, the point POL is not calculated, but only the intersection point P is used to determine the target point TP. This is because the vehicle 1 is parallel to the roadside 8 in this portion, and the angle φ in FIG.
When the target point TP is calculated, the coordinate data is output to the coincidence determination unit 32 with the guide. Further, it is also output to the data updating unit 29 together with the parameters of the approximate straight lines L0 and L1. If information on no intersection is received from the intersection calculation unit 30, the information is output to the data update unit 29.
[0021]
The data update unit 29 stores past data, changes the position and size of the processing region 22 in accordance with the amount of movement of the intersection P, and outputs it to the binarization / edge detection unit 18. In addition, the data update unit 29 outputs the slope of the approximate line to the approximate line extraction unit 25. This is the reference angle in the next straight line extraction process. When information indicating that the intersection point P is not detected is received from the target point calculation unit 31, each data is continuously used without being updated. The coordinates of the points P1, P3, and P4 in FIG. 7 are received from the result determination unit 26 of the shape recognition process, and the inclination of the straight line connecting the points P1 and P3 and the inclination of the straight line connecting the points P3 and P4 are obtained. These average values are obtained and used as the initial value of the reference angle when the approximate straight line is extracted.
[0022]
The coincidence determination unit 32 with the guide determines the approach and coincidence of the target point TP to the various guides by calculating the distance between the calculated target point TP and the various guides. At this time, the steering angle calculation unit 17 calculates the steering angle based on the count value from the external input processing unit 12, and the coincidence determination unit 32 with the guide uses the steering angle input from the steering angle calculation unit 17 to handle the handle mark. Calculate the position of. When the handle 7 is operated, the handle mark 10 moves on the column guideline 9 and approaches the target point TP. For the column guideline 9 and the eye mark 11, the target points TP approach each other as shown in FIG. A signal is sent to the voice selection unit 27 according to the distance between the target point TP and various guides, and the voice selection unit 27 outputs a corresponding voice from the speaker 28.
[0023]
FIG. 16 shows the output timing of the voice guide. (A) is a state until the target point TP overlaps with the column guide line 9, and the target point TP approaches the column guide line 9 in the order of S1, S2, and S3. Yes), “Pompom” (sound 2), “Pawn” (sound 3), and so on, changing the pitch and / or length of the sound so that the degree of approach can be understood. In S3, a voice for guiding the next operation is output as a voice guide, such as “turn the handle until the handle mark (for example, red mark) matches the target point”. (B) is a state until the handle mark 10 is aligned with the target point TP. As in the case of (a), when the handle mark 10 approaches the target point TP in the order of S4, S5, and S6, according to the distance. Sound 1, sound 2, and sound 3 are output. In S6, a voice for guiding the next operation is output, such as “Retreat until the eye mark (for example, yellow mark) matches the target point”. (C) is a state until the target point TP coincides with the eye mark 11, and when the eye mark 11 approaches the target point TP in the order of S7, S8, S9, sound 1, sound 2, sound 3 is output. In S9, a guidance voice such as “Please turn the handle all the way in the opposite direction” is output. In (d), when the steering wheel is turned more than a certain angle in the opposite direction, a guidance voice such as “Please move backward while paying attention to the back” is output. Thereafter, when the white line of the parking frame 19 disappears from the screen, a guidance voice such as “Please pay attention to the vehicle behind and stop the car” is output.
[0024]
Embodiment 2. FIG.
In the second embodiment, for example, even if the target point TP is lost from the column guideline 9 to the eye mark 11, a prediction method of the target point TP is introduced so that the tracking can be continued to some extent. FIG. 17 shows the configuration of the parallel parking support apparatus according to the second embodiment. In this apparatus, the image processing unit 33 is connected to the external input processing unit 12 instead of the image processing unit 13 in the apparatus of the first embodiment shown in FIG. Is the same. This image processing unit 33 is provided with a target point prediction unit 34 instead of the intersection calculation unit 30 and the target point calculation unit 31 of the image processing unit 13 of FIG. FIG. 18 shows a flowchart of the target point prediction unit 33. The process until the target point TP overlaps the column guide line 8 and the handle mark 10 coincides with the target point TP is the first half process, and the process until the target point TP overlaps the eye mark 11 is the second half process. In the first half process, intersection calculation and target point calculation are performed in the same manner as in the first embodiment. After entering the second half of the process, switch to target point prediction.
[0025]
Next, the principle of target point prediction will be described. In FIG. 19, the turning center C of the vehicle 1 is determined by the turning angle of the handle 7. The point P, the target point TP, and the point POL on the straight line in the vertical direction rotate around the center C when viewed from the vehicle 1. Finally, the target point TP overlaps the eye mark EM, and the point P overlaps the virtual eye mark VEM. The turning radii of these points P, TP and POL are respectively R_VEM, R_EM, R_POLIt is. FIG. 20 shows a method for correcting and estimating the point P. Although the point P rotates on the road surface, if the point P detected from the image is converted onto the road surface, it does not necessarily ride on the circumference. (A) shows a correction method when the detected point P deviates greatly from the circumference. Point P_iAnd the distance between the center C and R_EMWhen the difference between and increases, point P_iAnd the distance from the center C is R_EMModify to be. At this time, the x-axis and the line segment P_iAn angle θi with respect to C and an angle change φi are obtained. (B) shows an estimation method when the point P is not detected. The average value φ of the angle change is obtained from the past several processes. If point P is not detected, angle θ = θi + φ, radius R_EMTo find the point P. φ is used to obtain the next average angle change. Since the angle change of the point TP and the point POL is the same as that of the point P, the point TP and the point POL are obtained using the angle change obtained as a result of detecting and estimating the point P. At this time, the radius is R_VEMAnd R_POLIt is.
[0026]
When the point P is not corrected, the point POL needs to be corrected. As shown in FIG. 21, the distance D from the center C of the point P_PCIs radius R_EMIf it is larger but not corrected, if the point POL at the position of the point B is not corrected, an error occurs in the slope of the white line L1 connecting the point P and the point POL. Therefore, the distance D between the point P and the center C_PCRadius R_EMRatio to R_POLTo correct the point POL. That is, ΔABC is subjected to similarity conversion. By doing so, the inclination of the straight line L2 connecting the point P and the point POL can be made equal to the inclination of the straight line L3 connecting the point A and the point B.
Further, as shown in FIG. 22, as the processing proceeds, the calculated straight line L_i-1And the slope of the white line may shift. In order to absorb this, R at almost regular intervals._VEM, R_POLAnd update the angle data. Point P_iIf POL is detected, point POL_iAnd the radius is R_VEMi, R_POLiUpdate to At this time, the angle from the x-axis is also updated.
[0027]
FIG. 23 shows a flowchart of target point prediction. When the latter half of the process starts, the turning center C and the turning radius R of the eye mark from the turning angle of the handle_EMIs calculated. Subsequently, the point POL is calculated and the radius R_VEM, R_POLThe angle from the x axis is obtained as an initial value. If the point P is detected and the data is updated, the calculation of the point POL and the rotation radius R_VEM, R_POLAnd the angle data from the x-axis are updated. Thereafter, if the point P is detected, the point P is corrected, the angle data (angle change, angle from the x axis) is updated, and the target point TP and the point POL are calculated. When the point P is not detected, the point P is estimated, the angle data (an average angle change is adopted as the angle change) is updated, and the target point TP and the point POL are estimated. Each calculated point is converted on the screen of the monitor 4 to obtain inclination data. This inclination is used for straight line extraction. Thereafter, the next point P is detected.
FIG. 24A is a flowchart of the correction of the point P. When the correction is not performed, the ratio of the turning radii is obtained and used for the correction of the point POL. FIG. 24B is a flowchart for calculating the target point TP and the point POL, and FIG. 24C is a flowchart for estimating the point P.
[0028]
Embodiment 3 FIG.
FIG. 25 shows the configuration of the parallel parking assistance apparatus according to the third embodiment. This apparatus is the apparatus of the first embodiment shown in FIG. 3 in which a column mode switch 35 is additionally provided in the driver's seat and connected to the external input processing unit 12. When parking in parallel, the driver turns on the column mode switch 35 to enter the column mode. The column guideline 9 is displayed on the screen of the monitor 4 for the first time when the shift lever 5 is operated to R (retracted position) and the column mode switch 35 is turned on, and image processing is performed. That is, only necessary items are displayed when necessary. In addition, in order to notify that the column mode has been entered, the image processing unit 13 performs a mode guidance voice “This is a parallel parking mode” and an operation guidance voice “Retreat until the target point matches the blue line”. Is output from the speaker 28 via the voice selection unit 27.
Note that the tandem mode switch 35 may be connected to the apparatus of the second embodiment.
[0029]
FIG. 26 shows the configuration of the parallel parking assistance apparatus according to the fourth embodiment. In this apparatus, the image processing unit 13 is connected to the superimposing unit 14 in the apparatus of the first embodiment shown in FIG. The target judgment point 32 and approach information of various guides are sent to the superimposing unit 14 from the coincidence determination unit 32 with the guide of the image processing unit 13, and a display corresponding to the sound selected by the sound selection unit 27 is displayed on the screen of the monitor 4. To be done. For example, as shown in FIG. 27, an approach display unit 36 is provided at the bottom of the scene, and blue, yellow, red, etc. correspond to the sound sequentially changing to sound 1, sound 2, and sound 3 as approaching. The colors are sequentially changed and displayed on the approach display unit 36. Further, the operation steps to be performed by the driver can be displayed in characters. For example, the name of the guide that needs attention should be displayed, such as “guideline”, “handle mark”, and “eye mark”, superimposed on the approach display unit 36.
Note that the image processing unit 33 of the apparatus according to the second embodiment may be connected to the superimposing unit 14 so that the display corresponding to the sound is similarly performed on the screen of the monitor 4.
[0030]
【The invention's effect】
As described above, according to the tandem parking assistance apparatus of the first aspect, the image processing unit recognizes the target point for parking by image processing and is tandem based on the positional relationship between the target point and the guide display. Since the steering timing and the steering amount necessary for parking are output, the driver can perform the parallel parking operation of the vehicle very easily.further,As the image processing unit tracks the target point consisting of the intersection of two straight lines included in the parking frame as the vehicle retreats, the driver gradually determines the situation until the vehicle reaches the target point as the vehicle retreats. Can grasp.
Claim2According to the parallel parking assistance device described in claim1In this apparatus, the image processing unit predicts and corrects the position of the target point when the movement of the target point is greatly deviated. Therefore, even if the target point is lost, the driving operation support can be continued to some extent. it can.
Claim3According to the parallel parking assistance device according to claim 1,Or 2In this device, since the steering guidance is performed by producing a sound corresponding to the steering timing and the steering amount output from the image processing unit, the driver does not need to watch the monitor and confirms the safety of the surroundings. Parking can be done easily.
Claim4According to the parallel parking assistance device described in claim3In this device, the display corresponding to the steering guidance is displayed on the monitor screen, so that the driver can confirm the approaching state with his eyes, and even if he / she misses the voice through the character display, the current operation step Can be confirmed visually.
Claim5According to the parallel parking assistance device according to claim 1,4In the device of any one of the above, the guide display is displayed on the monitor screen when the parallel parking mode switch is turned on, so it is possible to consciously distinguish between parallel parking and other reverse driving. As well as being able to display only necessary drawings, the screen becomes easier to see.
According to the parallel parking assist device of the sixth aspect, the steering angle before the steering wheel is turned back easily by matching the steering wheel mark displayed in accordance with the steering angle of the steering wheel with the target point for parking. can do.
According to the parallel parking support apparatus of the seventh aspect, in the apparatus of the sixth aspect, the handle turning position can be easily achieved by matching the eye mark fixedly displayed on the monitor screen with the target point for parking. Can grasp.
[Brief description of the drawings]
FIG. 1 is a side view showing a vehicle equipped with a parallel parking assistance apparatus according to Embodiment 1 of the present invention.
FIG. 2 is a diagram showing a form of drawing on a monitor.
3 is a block diagram showing a configuration of a parallel parking assistance apparatus according to Embodiment 1. FIG.
FIG. 4 is a diagram showing a monitor screen on which a histogram area is displayed.
FIG. 5 is a graph showing the relationship between the density level of an image and the number of pixels.
FIG. 6 is a diagram showing a monitor screen showing a processing area.
FIG. 7 is a diagram showing a monitor screen showing other processing areas.
FIG. 8 is a diagram illustrating a state in a processing region when a target point is detected.
FIG. 9 is a diagram illustrating a relationship between a plurality of shapes of a processed image in a processing region, positions of points, and straight line inclinations.
FIG. 10 is a flowchart illustrating operations of a shape recognition unit and a result determination unit.
FIG. 11 is a flowchart showing approximate straight line extraction processing;
FIG. 12 is a diagram showing a contour line extraction method;
FIG. 13 is a diagram illustrating a coordinate conversion method.
14A and 14B are diagrams showing a method for calculating a target point, where FIG. 14A shows a monitor screen, and FIG. 14B shows a road surface.
FIG. 15 is a diagram illustrating movement of a target point on a monitor screen.
FIG. 16 is a diagram showing the output timing of the voice guide in stages.
FIG. 17 is a block diagram showing a configuration of a parallel parking support apparatus according to Embodiment 2.
FIG. 18 is a flowchart showing the operation of a target point prediction unit.
FIG. 19 is a diagram illustrating the principle of target point prediction.
20 is a diagram illustrating a correction method and an estimation method for a point P. FIG.
FIG. 21 is a diagram illustrating a method for correcting a point POL.
FIG. 22 is a diagram illustrating a data update method.
FIG. 23 is a flowchart showing a target point prediction process.
24A is a flowchart showing a method for correcting a point P, FIG. 24B is a flowchart showing a method for calculating a target point and a point POL, and FIG. 24C is a flowchart showing a method for estimating a point P;
25 is a block diagram showing a configuration of a parallel parking assistance apparatus according to Embodiment 3. FIG.
FIG. 26 is a block diagram showing a configuration of a parallel parking assistance apparatus according to Embodiment 4.
FIG. 27 is a diagram showing a monitor screen in the fourth embodiment.
[Explanation of symbols]
1 Vehicle, 2 Cameras, 3 Rear bumper, 4 Monitor, 5 Shift lever, 6 Front wheel, 7 Handle, 8 Road side, 9 Column guideline, 10 Handle mark, 11 Eye mark, 12 External input processing unit, 13, 33 Image processing unit , 14 Superimpose section, 15 Shift switch, 16 Steering sensor, 17 Steering angle calculation section, 18 Binarization / edge detection section, 19 Parking frame, 20 White line, 21 Histogram area, 22 Processing area, 23 Switching section, 24 Shape recognition unit, 25 approximate line extraction unit, 26 result determination unit, 27 voice selection unit, 28 speaker, 29 data update unit, 30 intersection calculation unit, 31 target point calculation unit, 32 match determination unit with guide, 34 target point Prediction unit, 35 column mode switch, 36 approach display unit.

Claims (7)

A camera that captures the back of the vehicle,
A monitor arranged in the driver's seat of the vehicle and displaying an image from the camera when the vehicle is retracted;
A steering sensor for detecting the steering angle of the steering wheel;
A superimpose unit that superimposes and displays a guide display for assisting driving of the vehicle during parallel parking on the screen of the monitor;
Recognizing two straight lines included in the parking frame by performing image processing on the video from the camera and using the intersection as a target point, tracking the target point according to the movement of the video as the vehicle moves backward, A parallel parking support apparatus comprising: an image processing unit that outputs a steering timing and a steering amount necessary for parallel parking based on a positional relationship between the vehicle and the guide display. The parallel parking support device according to claim 1 , wherein the image processing unit predicts and corrects the position of the target point when the movement of the target point greatly deviates as the vehicle moves backward. Parallel parking assist apparatus according to claim 1 or 2 further comprising an audio selection unit for performing guidance of the steering to the driver issues a sound corresponding to the output timing of the steering steering amount and from the image processing unit. The image processing unit according to claim 3 to perform a display timing and steering amount of the steering corresponding to the guide of the steering outputs to the superimposing unit together with the audio selection unit on the screen of the monitor Parallel parking assistance device. Further comprising a parallel parking mode switch located in the driver's seat of the vehicle, any claim 1-4 in which the guide displayed on the screen of the monitor by the superimposing unit when charged with parallel parking mode switch is displayed The parallel parking assistance device according to claim 1. A camera that captures the back of the vehicle,
A monitor arranged in the driver's seat of the vehicle and displaying an image from the camera when the vehicle is retracted;
A steering sensor for detecting the steering angle of the steering wheel;
A superimpose unit that superimposes and displays a guide display for assisting driving of the vehicle during parallel parking on the screen of the monitor;
Image processing for recognizing a target point for parking by image processing the video from the camera and outputting a steering timing and a steering amount necessary for parallel parking based on the positional relationship between the target point and the guide display Department and
The guide display includes a handle mark that is displayed on the monitor screen according to the steering angle of the handle detected by the steering sensor.
From the parking stop position, move the vehicle back while operating the handle until the handle mark matches the target point for parking on the monitor screen, and then turn the handle fully in the opposite direction and move backward. A parallel parking support apparatus characterized by performing parallel parking by means of the above. The guide display includes an eye mark fixedly displayed at a specific position with respect to the vehicle,
This eye mark indicates that the target point for parking coincides with the eye mark when the vehicle is moved backward with the handle operated until the handle mark coincides with the target point for parking on the monitor screen. The parallel parking assist device according to claim 6, which indicates reaching a position where the steering wheel of the vehicle is fully turned in the opposite direction.

Images