JP4449701B2 - Parking assistance system - Google Patents

Description

  The present invention relates to a parking support system that supports driving when a user moves a host vehicle forward and stops at a parking start position, and then retreats and parks the vehicle.

An example of a parking support system that supports driving when the user moves the host vehicle forward and stops at the parking start position and then turns the vehicle backward to park the vehicle is disclosed in Patent Document 1, for example.
JP 2002-362271 A

  By the way, in the parking assistance system described in Patent Document 1 described above, the target parking route and the parking position while viewing the bird's-eye view image behind the host vehicle after the user moves the host vehicle forward and stops at the parking start position. Since (the position where the user wants to finally park the host vehicle) is set, there is a problem that the timing until the user starts the parking operation after stopping the host vehicle at the parking start position, There was a problem that it was inconvenient in a busy parking lot.

  The present invention has been made in view of the above-described circumstances, and an object thereof is to suppress a delay in timing from when the user stops the host vehicle to the parking start position until the parking operation is started. It is in providing the parking assistance system which can raise.

  According to the first aspect of the present invention, when the host vehicle is moving forward, the control unit causes the rear of the host vehicle to capture images at a plurality of timings, and the host vehicle moves forward and stops at the parking start position. After that, when the host vehicle is moving forward, the plurality of captured images taken by the imaging unit are subjected to image analysis of the plurality of bird's-eye images obtained by converting the viewpoints by the bird's-eye image creating unit, and parked at a position where the host vehicle can be parked. Search is made as a candidate position, and the search result is superimposed on the synthesized bird's-eye image created by the synthesized bird's-eye image creating means and displayed on the display means.

As a result, when the user advances the host vehicle and stops at the parking start position, a plurality of bird's-eye images obtained by converting the viewpoints of the plurality of captured images captured when the host vehicle is moving forward are subjected to image analysis. The position where the vehicle can be parked is searched as a parking candidate position, and if there is a position where the own vehicle can be parked, the position where the own vehicle can be parked is displayed as a parking candidate position on the synthetic bird's-eye view image to notify the user. Therefore, unlike the conventional one, after the user advances the host vehicle and stops it at the parking start position, it does not require the troublesome work of setting the target parking position while looking at the bird's-eye view image behind the host vehicle. Therefore, a delay in timing from when the user stops the own vehicle to the parking start position until the parking operation is started can be suppressed, and usability can be improved.
Further, the control means specifies the predicted traveling direction of the host vehicle based on the steering angle at the time when the host vehicle stops at the parking start position, and the host vehicle parks the range of the predicted traveling direction of the identified host vehicle. A possible position is searched as a parking candidate position. Thereby, a position where the own vehicle can be parked is searched as a parking candidate position with respect to a range of directions when it is assumed that the user stops the own vehicle at the parking start position and then moves backward without changing the steering angle. it can. Further, the control means searches for a position where the host vehicle can be parked as a parking candidate position for the range of the predicted traveling direction of the host vehicle, and if it is impossible to determine the parking candidate position, the control unit predicts the host vehicle. A position where the host vehicle can be parked is searched as a parking candidate position in a range close to the host vehicle or a direction far from the host vehicle based on the traveling direction. Thereby, although the user has searched the position where the own vehicle can be parked as the parking candidate position with respect to the range of the direction when assuming that the user stops the vehicle at the parking start position and then moves backward without changing the steering angle, Even if it is impossible to determine a parking candidate position, a position where the host vehicle can park in a range in a direction close to the host vehicle or in a direction far from the host vehicle based on the predicted traveling direction of the host vehicle is set as the parking candidate position. Can be searched as.
According to the second and third aspects of the present invention, the control means is configured to detect the range of the predicted traveling direction of the host vehicle or the range of the direction close to the host vehicle or the direction far from the host vehicle based on the predicted traveling direction of the host vehicle. If it is impossible to determine a parking candidate position by searching a position where the host vehicle can be parked as a parking candidate position, a direction opposite to the direction searched with respect to the predicted traveling direction of the host vehicle is determined. A position where the host vehicle can be parked for a range is searched as a parking candidate position. Thus, a position where the own vehicle can be parked is searched as a parking candidate position in a range close to the own vehicle or in a range far from the own vehicle based on the estimated traveling direction range of the own vehicle or the estimated traveling direction of the own vehicle. However, even if it is impossible to determine the parking candidate position, a position where the own vehicle can be parked is set as a parking candidate position in a range in the direction opposite to the searched direction based on the predicted traveling direction of the own vehicle. You can search.

According to the invention described in claim 4 , when the host vehicle is moving forward, the control unit causes the shooting unit to capture the rear of the host vehicle at a plurality of timings and simultaneously detects a three-dimensional object around the host vehicle. A plurality of captured images photographed by the photographing means when the own vehicle is moving forward after the own vehicle moves forward and stops at the parking start position. The bird's-eye view image is analyzed and the detection result of the three-dimensional object detection means is determined, and a position where the host vehicle can be parked is searched as a parking candidate position.

  As a result, when the user advances the host vehicle and stops at the parking start position, the image analysis is performed on the plurality of bird's-eye images obtained by changing the viewpoints of the plurality of captured images captured when the host vehicle is moving forward. A position where the host vehicle can be parked is searched as a candidate parking position by determining a detection result of a three-dimensional object around the vehicle, and if there is a position where the host vehicle can be parked, the position where the host vehicle can be parked is determined as a parking candidate position. Since it is displayed on the synthetic bird's-eye image and notified to the user, the user moves the host vehicle forward and stops at the parking start position in the same manner as described above. The troublesome labor of setting the target parking position while viewing the bird's-eye view image can be eliminated. In this case, in addition to performing image analysis on a plurality of bird's-eye images obtained by changing the viewpoints of a plurality of captured images taken when the host vehicle is moving forward, a detection result of a three-dimensional object around the host vehicle is determined. Then, since the position where the host vehicle can be parked is searched as a parking candidate position, the position where the host vehicle can be parked can be searched with high accuracy.

According to the fifth aspect of the present invention, when the host vehicle is moving forward, the control unit causes the shooting unit to capture the rear of the host vehicle at a plurality of timings and simultaneously detects a three-dimensional object around the host vehicle. And detecting the traveling state of the host vehicle by the traveling state detecting unit. After the host vehicle moves forward and stops at the parking start position, a plurality of images taken by the shooting unit when the host vehicle is moving forward are captured. Image analysis is performed on a plurality of bird's-eye images obtained by converting the viewpoint of the captured image by the bird's-eye image creation unit, and the detection result of the three-dimensional object detection unit and the detection result of the traveling state detection unit are determined to determine a position where the host vehicle can park Search as a position.

  Thus, when the user advances the host vehicle and stops at the parking start position, the image analysis is performed on the plurality of bird's-eye images obtained by changing the viewpoints of the plurality of captured images captured when the host vehicle is moving forward. By determining the detection result of the three-dimensional object around the vehicle and the detection result of the running state of the host vehicle, a position where the host vehicle can be parked is searched as a parking candidate position, and if there is a position where the host vehicle can be parked, Is displayed on the synthetic bird's-eye view image as a parking candidate position to inform the user, so that the user advances the host vehicle and starts parking in the same manner as described in claim 1 above. It is possible to eliminate the troublesome work of setting the target parking position while viewing the bird's-eye view image behind the host vehicle after stopping at the position. Further, in this case, in addition to image analysis of a plurality of bird's-eye images obtained by converting viewpoints of a plurality of captured images taken when the host vehicle is moving forward, a detection result of the three-dimensional object around the host vehicle and Since the detection result of the running state of the vehicle is determined and a position where the host vehicle can be parked is searched as a parking candidate position, a position where the host vehicle can be parked can be searched with high accuracy.

According to the invention described in claim 6 , the control means generates a three-dimensional object outline representing the outline of the three-dimensional object by performing image analysis on a plurality of bird's-eye images, and the host vehicle detects the three-dimensional object outline from the generated three-dimensional object outline. A position where parking is possible is searched as a parking candidate position. Thereby, the position which can park the own vehicle can be searched as a parking candidate position by creating a solid object outline representing the outline of a solid object (for example, a parked vehicle or a building that has already been parked). .

According to the seventh aspect of the present invention, the control means detects a parking frame line by analyzing a plurality of bird's-eye images, and determines a position where the host vehicle can be parked from the detected parking frame line as a parking candidate position. Search as. Thereby, the position which can park the own vehicle can be searched as a parking candidate position by detecting a parking frame line (for example, the white line drawn on the road).

According to the invention described in claim 8 , the control means determines the detection result of the three-dimensional object detection means and the detection result of the traveling state detection means, and creates a three-dimensional object distance map representing the distance from the host vehicle to the three-dimensional object. In addition, a three-dimensional object outline created by image analysis of a plurality of bird's-eye images is corrected based on the three-dimensional object distance map, and a position where the host vehicle can be parked from the corrected three-dimensional object outline is determined as a parking candidate. Search as a position. Thereby, while creating the solid object distance map showing the distance from the own vehicle to the three-dimensional object, and correcting the three-dimensional object outline based on the three-dimensional object distance map, a position where the own vehicle can be parked is determined as a parking candidate. It can be searched as a position, and a position where the host vehicle can be parked can be searched with high accuracy.

According to the ninth aspect of the present invention, when the predetermined condition is satisfied in the situation where the parking candidate position is displayed on the display means, the control means determines the parking candidate position displayed at that time as the parking position. Determine as. Thereby, for example, when the user cancels the brake operation or emits a specific sound to establish a predetermined condition, the parking candidate position displayed at that time can be determined as the parking position.

According to the tenth aspect of the present invention, when the other predetermined condition is satisfied after the candidate parking position is determined as the parking position, the control means uses the automatic steering means from the parking start position to the parking position. Auto-steer. Thereby, for example, the user can automatically steer the vehicle from the parking start position to the parking position by the automatic steering means by releasing the brake operation or generating a specific sound to establish another predetermined condition. .

(First embodiment)
Hereinafter, a first embodiment of the present invention will be described with reference to FIGS. FIG. 1 shows the overall configuration of the parking assistance system as a functional block diagram. The parking assistance system 1 includes a parking assistance ECU 2 (bird's-eye image creation means, synthetic bird's-eye image creation means, control means in the present invention), an engine ECU 3, a yaw rate sensor 4 (traveling state detection means in the present invention), a steering sensor 5, A navigation ECU 6 and an automatic steering circuit 7 (automatic steering means in the present invention) are connected via an in-vehicle LAN 8.

  The parking assist ECU 2 receives a yaw rate detection signal input from the yaw rate sensor 4, a steering detection signal input from the steering sensor 5, and a vehicle speed input via the engine ECU 3 from the vehicle speed sensor 9 (traveling state detection means in the present invention). The parking assist operation is controlled by analyzing the detection signal and the like. The operation switch 10 is disposed in the driver's seat and is operated by the user to activate the parking support operation. When operated by the user, the operation switch 10 outputs an operation detection signal to the parking support ECU 2. When the operation detection signal is input from the switch 10, it is detected that the user has started the parking assist operation.

  The in-vehicle camera 11 (photographing means in the present invention) is attached to the rear part of the vehicle body of the host vehicle and captures the rear side of the host vehicle, and outputs a captured image of the rear side of the host vehicle to the parking support ECU 2. When a captured image is input from the in-vehicle camera 11, the input captured image is subjected to viewpoint conversion based on a predetermined viewpoint conversion algorithm to create a bird's-eye image, and a plurality of bird's-eye images are based on a predetermined synthesis algorithm. Create a composite bird's-eye view image. The display device 12 (display means in the present invention) is disposed in a position that is easy for the user to see in the vehicle interior, and when a display instruction signal is input from the parking assistance ECU 2, the composite bird's-eye image created by the parking assistance ECU 2 is displayed. indicate.

  The vehicle speed sensor 9 detects the vehicle speed and outputs a vehicle speed detection signal representing the detected vehicle speed to the engine ECU 3. When a vehicle speed detection signal is input from the vehicle speed sensor 9, the engine ECU 3 controls the engine based on the input vehicle speed detection signal, and the input vehicle speed detection signal is sent to the parking assist ECU 2 via the in-vehicle LAN 8. Output to. The yaw rate sensor 4 detects the yaw rate, and outputs a yaw rate detection signal representing the detected yaw rate to the parking assistance ECU 2 via the in-vehicle LAN 8. The steering sensor 5 detects the steering and outputs a steering detection signal representing the detected steering to the parking assist ECU 2 via the in-vehicle LAN 8.

  The navigation ECU 6 controls the operation of a navigation device (not shown) mounted on the vehicle. In the configuration described above, the function of accepting a user operation by the operation switch 10 and the function of displaying the synthesized bird's-eye image by the display device 12 may be substituted by the function of the navigation device. When an automatic steering control signal is input from the parking assist ECU 2, the automatic steering circuit 7 automatically steers the host vehicle based on the input automatic steering control signal.

  Next, the operation of the above configuration will be described with reference to FIGS. Here, FIG.2 and FIG.3 has shown the process which parking assistance ECU2 performs as a flowchart.

  First, when the user moves the own vehicle forward and stops at the parking start position and then turns back and parks the vehicle, an appropriate timing before stopping the own vehicle at the parking start position, that is, the own vehicle. By operating the operation switch 10 at an appropriate timing while the vehicle is moving forward, the parking assist operation is activated. In the present embodiment, the case where the parking assistance operation is activated by the user operating the operation switch 10 has been described. However, the parking assistance operation is activated when the user emits a specific sound. Alternatively, the parking assist operation may be automatically activated when the host vehicle decelerates and the forward speed of the host vehicle drops below a predetermined speed.

  When the operation detection signal is input from the operation switch 10, the parking assist ECU 2 starts a parking assist operation and determines whether or not the movement amount of the host vehicle has reached a predetermined distance (for example, 10 centimeters) ( Step S1), it is determined whether or not the host vehicle has stopped at the parking start position (Step S2). Here, when the parking assistance ECU 2 detects that the amount of movement of the host vehicle has reached the predetermined distance (“YES” in step S1), the rear of the host vehicle is photographed by the in-vehicle camera 11 (step S3). The captured image is stored (step S4). And parking assistance ECU2 repeats the above process until it detects that the own vehicle has stopped at the parking start position, and detects that the own vehicle has stopped at the parking start position (in step S2 or step S5). "YES"), it is determined whether or not the shift operation has been moved to the back position (step S6).

  Here, when the user stops the host vehicle at the parking start position, the user moves the shift operation to the back position while maintaining the brake operation (for example, foot brake operation). When the parking assist ECU 2 detects that the shift operation has been moved to the back position (“YES” in step S6), the parking assist ECU 2 proceeds to a parking candidate position display process (step S7).

  When the parking assistance ECU 2 proceeds to the parking candidate position display process, the plurality of captured images that have been stored so far, that is, the plurality of captured images captured at a plurality of timings when the host vehicle is moving forward. The viewpoint of the image is converted to create a plurality of bird's-eye images (step S11). And parking assistance ECU2 synthesize | combines these produced several bird's-eye images, and produces a synthetic bird's-eye image.

  Specifically, as shown in FIG. 4, the parking assist ECU 2 parks the host vehicle (indicated by “host vehicle A” in FIG. 4) via the travel position a, the travel position b, and the travel position c. Assuming a case where the vehicle stops at the start position, a captured image captured when the host vehicle is traveling at the traveling position a, a captured image captured when the host vehicle is traveling at the traveling position b, and the host vehicle. A bird's-eye view image is created by converting viewpoints of the captured image captured when the vehicle is traveling at the traveling position c and the captured image captured when the host vehicle is stopped at the parking start position. And parking assistance ECU2 synthesize | combines these produced several bird's-eye images, and produces a synthetic bird's-eye image.

  In this case, when the deviation between the bird's-eye images is large, the parking assistance ECU 2 calculates the deviation to detect a three-dimensional object (obstacle) and occupies the three-dimensional object in an area where the three-dimensional object exists. All bird's-eye images are synthesized (overlapped) so that the area is minimized, and a three-dimensional object is created by transforming the real image and filling the region of the three-dimensional object without converting the viewpoint. May be.

  Next, the parking assist ECU 2 detects a three-dimensional object based on a plurality of bird's-eye images and creates a three-dimensional object contour line representing the contour of the three-dimensional object (step S12). Also, a parking frame based on the plurality of bird's-eye images. A line is detected (step S13). Specifically, as shown in FIG. 4 described above, the parking assist ECU 2 has, for example, three parked vehicles parked around the parking start position where the host vehicle stopped, and a wall is erected, Also, assuming a case where a white line is drawn on the road, by analyzing a plurality of bird's-eye images, a contour line of a parked vehicle and a contour line of a wall are created as a three-dimensional object contour line as shown in FIG. In addition, a white line drawn on the road is detected as a parking frame line.

  In the present embodiment, the parking support ECU 2 has a function of creating a three-dimensional object contour line and a function of detecting a parking frame line, so that the parking support ECU 2 three-dimensionalizes the contour line of the parked vehicle and the contour line of the wall. Although the case where the white line drawn on the road is detected as the parking frame line has been described while being created as an object outline, by providing either one of the functions, as shown in FIG. It may be configured to create a contour line or a wall contour line as a three-dimensional object contour line, or, as shown in FIG. 7, only to detect a white line drawn on the road as a parking frame line. There may be.

  Then, the parking assist ECU 2 identifies the predicted traveling direction of the host vehicle based on the direction of the steering angle at that time, that is, when the host vehicle stops at the parking start position, and the predicted progress of the identified host vehicle. The parking candidate position is searched from the solid object outline and the parking frame line for the range of directions (step S14). Here, the parking assist ECU 2 searches for a parking candidate position as follows.

  That is, when the parking assist ECU 2 searches for a parking candidate position from the three-dimensional object outline, the parking assistance ECU 2 identifies the shape of the created three-dimensional object outline, and, for example, as shown in FIG. The line shape has a depth (own vehicle length + α) longer than the own vehicle length (total length in the longitudinal direction of the own vehicle) and a width (own vehicle length + β) wider than the own vehicle width (full width in the left-right direction of the own vehicle). When it is detected that the shape has a depression, it is detected that there is a parking candidate position. On the other hand, the shape of the three-dimensional object outline has a depression longer than the own vehicle length and wider than the own vehicle width. If it detects that it is not a shape, it will detect that there is no parking candidate position.

  Further, when searching for a parking candidate position from the parking frame line, the parking assist ECU 2 identifies the detected parking frame line shape and, for example, two parking frame lines as shown in FIG. Detects that there is a parking candidate position when detecting that the pair has a length longer than the own vehicle length and a width wider than the own vehicle width, while two parking frame lines When it is detected that there is no pair so as to have a length longer than the own vehicle length and a width wider than the own vehicle width, it detects that no parking candidate position exists.

  Further, when the parking assist ECU 2 searches for a parking candidate position from the solid object outline and the parking frame line, the parking assistance ECU 2 identifies the shape of the created solid object outline and the detected shape of the parking frame line, for example, When detecting that the shape of the three-dimensional object outline is a shape having a depression wider than the width of the host vehicle at a position away from one parking frame line, it detects that a parking candidate position exists, When it is detected that the shape of the three-dimensional object outline is not a shape having a depression wider than the width of the host vehicle at a position away from one parking frame line, it is detected that no parking candidate position exists. Note that the determination criterion as to whether or not there is a parking candidate position described above is an example, and it may be determined whether or not a parking candidate position exists based on a determination criterion other than these. Further, α and β shown in FIG. 6B and α and β shown in FIG. 7B may be different values depending on the own vehicle length and the own vehicle width.

  And if parking assistance ECU2 detects that a parking candidate position exists as a result of having searched the parking candidate position about the range of the estimated advancing direction of the own vehicle ("YES" in step S15), as shown in FIG. In addition, the host vehicle graphic representing the host vehicle and the parking candidate position diagram representing the parking candidate position are superimposed on the synthesized bird's-eye image and displayed on the display device 12 (step S16). In this case, the parking assist ECU 2 may also display the predicted route arrow indicating the predicted route from the parking start position of the host vehicle to the parking candidate position on the composite bird's-eye image and display it on the display device 12.

  On the other hand, the parking assist ECU 2 searches for the parking candidate position for the range of the predicted traveling direction of the host vehicle, but detects that the parking candidate position does not exist ("NO" in step S15), the parking assist ECU 2 A candidate parking position is searched from the three-dimensional object outline and the parking frame line by performing the same procedure as the procedure described above for the range in the direction close to the host vehicle based on the predicted traveling direction (step S17). More specifically, as shown in FIG. 9A, the parking assist ECU 2 sets a range for searching for a parking candidate position in a direction close to the own vehicle based on the predicted traveling direction of the own vehicle (in the left direction in FIG. 9). ) To search for a parking candidate position from a three-dimensional object outline or a parking frame line in a range in a direction close to the own vehicle based on the predicted traveling direction of the own vehicle.

  And parking assistance ECU2 will detect that a parking candidate position exists as a result of having searched the parking candidate position about the range of the direction close | similar to the own vehicle on the basis of the estimated advancing direction of the own vehicle (in step S18, " YES ”), as described above, the own vehicle graphic representing the own vehicle and the parking candidate position graphic representing the parking candidate position are superimposed on the synthesized bird's-eye image and displayed on the display device 12 (step S16).

  On the other hand, the parking assist ECU 2 searches for a parking candidate position for a range in a direction close to the own vehicle with reference to the predicted traveling direction of the own vehicle, but detects that no parking candidate position exists (in step S18). "NO"), the parking candidate position is searched from the three-dimensional object outline and the parking frame line by performing the same procedure as described above for the range in the direction far from the host vehicle based on the predicted traveling direction of the host vehicle. (Step S19). More specifically, as shown in FIG. 9B, the parking assist ECU 2 sets a range for searching for a parking candidate position in a direction far from the own vehicle based on the predicted traveling direction of the own vehicle (in the right direction in FIG. 9). ) To search for a parking candidate position from a three-dimensional object outline or a parking frame line in a range in a direction far from the own vehicle based on the predicted traveling direction of the own vehicle.

  And parking assistance ECU2 will detect that a parking candidate position exists as a result of having searched the parking candidate position about the range of the direction far from the own vehicle on the basis of the estimated advancing direction of the own vehicle (in step S20, " YES ”), as described above, the own vehicle graphic representing the own vehicle and the parking candidate position graphic representing the parking candidate position are superimposed on the synthesized bird's-eye image and displayed on the display device 12 (step S16).

  On the other hand, the parking assist ECU 2 searches for a parking candidate position for a range in a direction far from the host vehicle based on the predicted traveling direction of the host vehicle, but detects that no parking candidate position exists (in step S20). "NO"), an undetected message indicating that no parking candidate position exists is displayed on the display device 12 (step S21). In this way, when the parking assist ECU 2 searches the range in any direction and detects that the parking candidate position exists, the parking support ECU 2 displays the parking candidate position on the display device 12, while the range in any direction is displayed. If it is detected that there is no parking candidate position for the searched item, an undetected message is displayed on the display device 12.

  Then, when the parking candidate position is displayed in this way, the user determines the displayed parking candidate position as a parking position (a position where the user finally wants to park the host vehicle). The parking assist ECU 2 determines whether or not the brake operation has been released by the user (whether a predetermined condition referred to in the present invention is satisfied, or whether another predetermined condition referred to in the present invention is satisfied) (step S8). ) When it is detected that the brake operation is released by the user (“YES” in step S8), the candidate parking position displayed at that time is determined as the parking position (step S9), and the automatic steering control signal is determined. Output to the automatic steering circuit 7 to automatically steer the host vehicle from the parking start position to the parking position (step S10).

  In the present embodiment, the case where the user releases the brake operation to determine the parking candidate position as the parking position and automatically steers the vehicle from the parking start position to the parking position has been described. The conditions for determining the position and the conditions for automatically steering the host vehicle from the parking start position to the parking position may be other things such as a user making a specific sound or performing a specific switch operation, for example. Further, the conditions may be different from each other.

  By performing the series of processes described above, the user stops the host vehicle at the parking start position when the host vehicle moves forward and stops at the parking start position, and then moves backward while parking. If the position where the host vehicle can be parked is displayed as the parking candidate position, and there is no problem even if the displayed parking candidate position is determined as the parking position, the brake operation is released. Thus, the parking candidate position can be determined as the parking position, and the host vehicle can be parked from the parking start position to the parking position by automatic steering. That is, the user can easily park the host vehicle from the parking start position to the parking position without performing the troublesome task of setting a target parking position while viewing the bird's-eye view image behind the host vehicle.

  By the way, the above describes the case where there is one parking candidate position as a result of searching for a parking candidate position, but if there are a plurality of parking candidate positions as a result of searching for a parking candidate position, As shown in FIGS. 10 to 12, only the parking candidate positions in the expected traveling direction specified from the steering angle among the plurality of searched parking candidate positions are displayed, and the user operates the steering to change the steering angle. The parking candidate positions may be sequentially displayed in response to this.

  More specifically, when there is no parked vehicle under the environment shown in FIG. 4, there are a plurality of parking candidate positions, so the display image shown in FIG. 10 is displayed. When the user turns the steering wheel to the left while maintaining the brake operation from the state, the parking candidate position may be moved and displayed in the left direction as shown in FIG. When the steering wheel is turned to the right while maintaining the brake operation, the parking candidate position may be moved to the right and displayed as shown in FIG.

  As described above, according to the first embodiment, in the parking support system 1, when the host vehicle is moving forward, the rear of the host vehicle is photographed, and the host vehicle moves forward and stops at the parking start position. Later, a plurality of bird's-eye images obtained by converting a plurality of captured images taken when the host vehicle is moving forward are subjected to image analysis, and a position where the host vehicle can be parked is searched as a parking candidate position. If there is a position where the vehicle can be parked, a position where the host vehicle can be parked is displayed as a parking candidate position on the synthesized bird's-eye view image to inform the user.

  Thereby, unlike the conventional one, it is possible to eliminate the troublesome work of setting the target parking position while viewing the bird's-eye view image after the user advances the host vehicle and stops at the parking start position. And the delay of the timing after a user stops the own vehicle to a parking start position until it starts parking operation can be suppressed, and usability can be improved.

(Second Embodiment)
Next, a second embodiment of the present invention will be described with reference to FIGS. In addition, description is abbreviate | omitted about the same part as above-mentioned 1st Embodiment, and a different part is demonstrated. Compared to the first embodiment described above, the second embodiment detects three-dimensional objects around the host vehicle by sonar and detects the running state of the host vehicle by a vehicle speed sensor or a yaw rate sensor. The point which correct | amends a solid-object outline based on a result differs. That is, in the parking assist system 21, the sonar 22 (three-dimensional object detection means in the present invention) is attached to, for example, a side portion of the vehicle body of the host vehicle, and transmits ultrasonic waves in a predetermined direction to receive reflected waves. Based on the above, a three-dimensional object around the host vehicle is detected.

  In this case, when the parking assistance ECU 2 detects that the amount of movement of the host vehicle has reached the predetermined distance (“YES” in step S1), the rear of the host vehicle is photographed by the in-vehicle camera 11 and at the same time, The three-dimensional object is detected by the sonar 22, the vehicle speed is detected by the vehicle speed sensor 9, and the yaw rate is detected by the yaw rate sensor 4 (step S31). At the same time, the captured image is stored, and at the same time, the detected surroundings of the own vehicle are detected. The detection result of the three-dimensional object is saved, and the detection result of the vehicle speed and the yaw rate is saved (step S32). Also in this case, the parking assist ECU 2 repeats the above processing until it detects that the host vehicle has stopped at the parking start position.

  Next, when the parking assistance ECU 2 proceeds to the parking candidate position display process, the parking support ECU 2 creates a plurality of bird's-eye images (step S11), detects a three-dimensional object based on the created bird's-eye images, and outlines the three-dimensional object. Is created (step S12), and the parking frame line is detected based on a plurality of bird's-eye images (step S13). In this case, the three-dimensional object around the stored vehicle is stored. Detection results and vehicle speed and yaw rate detection results, that is, the three-dimensional objects around the host vehicle detected when the host vehicle is moving forward, the vehicle speed and yaw rate, and the vehicle is close to the actual situation (high accuracy). A three-dimensional object distance map representing the relative positional relationship between the vehicle and the three-dimensional object is created (step S41), and the created three-dimensional object outline is corrected based on the three-dimensional object distance map (step S42).

  Specifically, as shown in FIG. 16, the parking assist ECU 2 creates a three-dimensional object distance map according to the array of measured points of the sonar 22 and corrects the three-dimensional object outline based on the three-dimensional object distance map. In FIG. 6, the position of the parked vehicle and the position of the wall are shown as positions on the synthesized bird's-eye image as being able to be accurately detected. In this case, since the three-dimensional object (obstacle) is detected by calculating the deviation, the actual point of the three-dimensional object distance map is shown in FIG. 6 due to the influence of the error generated in creating the synthetic bird's-eye view image. It does not necessarily match the position of the parked vehicle or the position of the wall shown. In addition, it is not necessary to correct all portions of the three-dimensional object outline created by image analysis of a plurality of bird's-eye images based on the three-dimensional object distance map, and may be partially corrected.

  And parking assistance ECU2 utilizes the solid object outline corrected based on the solid object distance map in this way, and thereafter performs the same processing as that described in the first embodiment described above. Thus, the parking candidate position is searched from the three-dimensional object outline and the parking frame line.

  By the way, the configuration in which the three-dimensional object around the own vehicle is detected by the sonar has been described above, but the configuration in which the three-dimensional object around the own vehicle is detected by the radar may be used. The configuration may be such that an object is detected by combining sonar and radar. In addition, the configuration in which the three-dimensional object around the own vehicle is detected by the sonar and the traveling state of the own vehicle is detected by the vehicle speed sensor or the yaw rate sensor has been described, but only the three-dimensional object around the own vehicle is determined and the three-dimensional object distance map is displayed. The structure to create may be sufficient.

  As described above, according to the second embodiment, in the parking support system 21, when the host vehicle is moving forward, the rear of the host vehicle is photographed and a solid object around the host vehicle is detected and the host vehicle is detected. After the host vehicle moves forward and stops at the parking start position, a plurality of bird's-eye images obtained by converting the viewpoints of the plurality of captured images captured when the host vehicle is moving forward are analyzed. In addition, the detection result of the three-dimensional object and the detection result of the traveling state of the host vehicle are determined, and a position where the host vehicle can be parked is searched as a parking candidate position. If there is a position where the host vehicle can park, the host vehicle is parked. A possible position is displayed as a parking candidate position on the synthetic bird's-eye view image to inform the user.

  Accordingly, in the same manner as described in the first embodiment, it is troublesome to set a target parking position while viewing the bird's-eye view image after the user advances the host vehicle and stops at the parking start position. Time can be eliminated, and a delay in timing from when the user stops the host vehicle to the parking start position until the parking operation is started can be suppressed, and usability can be improved. In this case, image analysis is performed on a plurality of bird's-eye images, and a detection result of a three-dimensional object around the host vehicle and a detection result of the running state of the host vehicle are determined to search for a position where the host vehicle can be parked as a parking candidate position. Therefore, the position where the host vehicle can be parked can be searched with high accuracy.

(Other embodiments)
The present invention is not limited to the above-described embodiment, and can be modified or expanded as follows.
The configuration may be such that the parking support operation is automatically activated when the host vehicle enters the parking lot by taking in the map information from the navigation device.
Not only is the configuration in which the parking candidate position display process is performed when the user moves the shift operation to the back position, but other conditions such as the user making a specific voice or performing a specific switch operation are established. In this case, the configuration may be such that parking candidate position display processing is performed.

The configuration is not limited to detecting a white line drawn on the road as a parking frame line, but may be a configuration detecting a rope stretched on the road as a parking frame line.
Converts viewpoints not only of the images taken by the in-vehicle camera that captures the back of the vehicle but also the images captured by the in-vehicle camera that captures the side of the vehicle, and detects three-dimensional objects and creates a composite bird's-eye view image You may do it.
In 2nd Embodiment, the structure detected not only by the structure which detects the solid object around the own vehicle by the sonar which sends out an ultrasonic wave but by the sensor which sends out electromagnetic waves, and the sensor which sends out laser light may be sufficient.

The functional block diagram which shows the 1st Embodiment of this invention and shows the whole parking assistance system flowchart 2 equivalent diagram The figure which shows the aspect by which a several bird's-eye view image is synthesize | combined roughly The figure which shows the aspect in which a solid-object outline is created roughly The figure which shows the aspect in which a parking frame line is detected roughly The figure which shows schematically the aspect in which a solid frame outline is created and a parking frame line is detected The figure which shows the example of the display picture Diagram showing the movement of the search range Equivalent to FIG. Equivalent to FIG. Equivalent to FIG. The functional block diagram which shows the 2nd Embodiment of this invention and shows the whole parking assistance system 2 equivalent diagram 2 equivalent diagram The figure which shows the aspect by which a solid-object outline is correct | amended roughly

Explanation of symbols

In the drawings, 1 is a parking assistance system, 2 is a parking assistance ECU (bird's-eye image creation means, synthetic bird's-eye image creation means, control means), 4 is a yaw rate sensor (running state detection means), and 7 is an automatic steering circuit (automatic steering means). ), 9 is a vehicle speed sensor (running state detection means), 11 is an in-vehicle camera (photographing means), 12 is a display device (display means), 21 is a parking support system, and 22 is a sonar (three-dimensional object detection means).

Claims (10)

A parking assistance system that assists driving when the host vehicle moves forward and stops at the parking start position and turns backward while parking.
Photographing means for photographing the rear of the vehicle;
A bird's-eye image creation means for creating a bird's-eye image by converting the viewpoint of the photographed image photographed by the photographing means;
A synthesized bird's-eye image creation means for creating a synthesized bird's-eye image by combining a plurality of bird's-eye images created by the bird's-eye image creation means;
When the host vehicle is moving forward, the rear of the host vehicle is photographed at a plurality of timings by the photographing means, and after the host vehicle moves forward and stops at the parking start position, the shooting is performed when the host vehicle is moving forward. A plurality of bird's-eye images obtained by converting the plurality of captured images taken by the means by the bird's-eye image creation means are image-analyzed to search for a position where the host vehicle can be parked as a parking candidate position, and the search result is the composite Control means for superimposing the synthesized bird's-eye image created by the bird's-eye image creation means on the display means ,
The control means identifies the predicted traveling direction of the host vehicle based on the steering angle at the time when the host vehicle stops at the parking start position, and the host vehicle can park the range of the identified predicted traveling direction of the host vehicle. If it is impossible to determine a candidate parking position and determine a candidate parking position, a range in a direction near or far from the own vehicle based on the predicted traveling direction of the own vehicle A parking support system that searches a position where the host vehicle can be parked as a parking candidate position . In the parking assistance system according to claim 1 ,
Before SL control means, parking candidate positions available parking position the vehicle is about the expected traveling direction of the range or far range of directions from the direction or the vehicle near the expected traveling direction of the host vehicle on the basis of the vehicle of the vehicle If it is impossible to determine the candidate parking position, the vehicle is parked at a position where the host vehicle can park in a range in the direction opposite to the searched direction based on the predicted traveling direction of the host vehicle. A parking support system characterized by searching as a candidate position. A parking assistance system that assists driving when the host vehicle moves forward and stops at the parking start position and turns backward while parking.
Photographing means for photographing the rear of the vehicle;
A bird's-eye image creation means for creating a bird's-eye image by converting the viewpoint of the photographed image photographed by the photographing means;
A synthesized bird's-eye image creation means for creating a synthesized bird's-eye image by combining a plurality of bird's-eye images created by the bird's-eye image creation means;
When the host vehicle is moving forward, the rear of the host vehicle is photographed at a plurality of timings by the photographing means, and after the host vehicle moves forward and stops at the parking start position, the shooting is performed when the host vehicle is moving forward. A plurality of bird's-eye images obtained by converting the plurality of captured images taken by the means by the bird's-eye image creation means are image-analyzed to search for a position where the host vehicle can be parked as a parking candidate position, and the search result is the composite Control means for superimposing the synthesized bird's-eye image created by the bird's-eye image creation means on the display means,
The control means identifies the predicted traveling direction of the host vehicle based on the steering angle at the time when the host vehicle stops at the parking start position, and the range of the predicted traveling direction of the identified host vehicle or the predicted progress of the host vehicle. When it is impossible to search a position where the host vehicle can be parked as a parking candidate position for a range in a direction close to the host vehicle or a direction far from the host vehicle based on the direction, and determining the parking candidate position, A parking support system, wherein a position where the host vehicle can be parked is searched as a parking candidate position for a range in a direction opposite to the direction searched with respect to the predicted traveling direction of the host vehicle. In the parking assistance system according to any one of claims 1 to 3,
Provided with a three-dimensional object detection means for detecting a three-dimensional object around the host vehicle with electromagnetic waves, ultrasonic waves or laser light,
The control means causes the three-dimensional object around the host vehicle to be detected by the three-dimensional object detection means at the same time as the rear of the host vehicle is photographed at a plurality of timings by the photographing means. After moving forward and stopping at the parking start position, image analysis is performed on a plurality of bird's-eye images obtained by converting a plurality of captured images photographed by the photographing unit when the host vehicle is moving forward by the bird's-eye image creating unit. A parking support system characterized by determining the detection result of the three-dimensional object detection means and searching for a position where the host vehicle can be parked as a parking candidate position. In the parking assistance system according to claim 4 ,
A driving state detecting means for detecting the driving state of the host vehicle;
The control means causes the three-dimensional object around the host vehicle to be detected by the three-dimensional object detection means while the rear of the host vehicle is photographed at a plurality of timings by the photographing means when the host vehicle is moving forward. After the traveling state is detected by the traveling state detection means and the host vehicle moves forward and stops at the parking start position, a plurality of captured images captured by the photographing unit when the host vehicle is moving forward are the bird's-eye view images. Analyzing a plurality of bird's-eye images whose viewpoints have been converted by the creating means and determining a detection result of the three-dimensional object detection means and a detection result of the traveling state detection means to search for a position where the host vehicle can be parked as a parking candidate position A parking assistance system characterized by that. In the parking assistance system according to any one of claims 1 to 5,
The control means, the bird's-eye image of the multiple and image analysis to create a three-dimensional object contour line representing the contour of the three-dimensional object, the vehicle from created three-dimensional object contour that is capable parking position as a parking candidate position A parking support system characterized by searching. In the parking assistance system according to any one of claims 1 to 6,
The control means detects a parking frame line by performing image analysis on a plurality of bird's-eye images, and searches for a position where the host vehicle can be parked as a parking candidate position from the detected parking frame line. system. In the parking assistance system according to any one of claims 4 to 7,
The control means determines the detection result of the three-dimensional object detection means and the detection result of the traveling state detection means to create a three-dimensional object distance map representing the distance from the host vehicle to the three-dimensional object, and also displays a plurality of bird's-eye images. The three-dimensional object outline created by image analysis is corrected based on the three-dimensional object distance map, and a position where the host vehicle can be parked is searched as a parking candidate position from the corrected three-dimensional object outline. Parking assistance system. In the parking assistance system according to any one of claims 1 to 8,
The control means determines the parking candidate position displayed at that time as the parking position when a predetermined condition is satisfied in a situation where the parking candidate position is displayed on the display means. Support system. In the parking assistance system according to claim 9 ,
Equipped with automatic steering means for automatically steering the vehicle from the parking start position to the parking position;
The control means automatically steers the own vehicle from the parking start position to the parking position by the automatic steering means when another predetermined condition is satisfied after determining the parking candidate position as the parking position. Support system.

Images