JP2005032075A - Rear side alarm device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a rear side alarm device capable of performing a proper alarm operation by precisely detecting an obstacle existing in the neighborhood of a present vehicle with a simple configuration and reducing its cost. <P>SOLUTION: The rear side alarm device detects that an obstacle deviates from the visual field range of a camera 21, and observes the luminance change of a vertical edge face in an image to be picked up when the obstacle deviates from the visual field range of the camera 21, and predicts the location of the obstacle beyond the visual field range of the camera 21 based on the observation result. Thus, after the overall obstacle deviates from the visual field range of the camera 21, the back side alarm device predicts the location of the obstacle without detecting the obstacle by any image processing. Therefore, it is possible to realize simple sensor arithmetic configurations. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a rear side alarm device that detects an obstacle existing in the vicinity of a host vehicle and performs an alarm operation when necessary.

  2. Description of the Related Art Conventionally, for the purpose of improving the safety of driving operation of a vehicle, an apparatus that detects an obstacle present on the side of a traveling vehicle and notifies the driver (for example, is known) , See Patent Document 1).

Patent Document 1 discloses a vehicle side monitoring device that monitors a side vehicle based on a signal captured by a video camera having a visual axis directed to the side of the host vehicle. The vehicle side monitoring apparatus disclosed in Patent Document 1 extracts a feature point from each frame of a captured image, and based on the amount of movement of the feature point in the image between frames, The optical flow in the side image is calculated. Further, based on the speed of the host vehicle, a movement flow length corresponding to the amount of movement of the image as the host vehicle moves is calculated. Then, the calculated optical flow and the moving flow length are compared, and if they are different, it is determined that the vehicle is present on the side of the host vehicle, and an alarm is issued.
JP 7-296297 A

  Incidentally, in the conventional technique described in Patent Document 1 described above, it is necessary to determine the same feature point from the feature points extracted for each frame of the image based on image information such as luminance distribution. However, the image information such as the brightness distribution of the vehicle that exists on the side of the host vehicle depends on the difference in the projection plane due to the change in the relative position between the vehicle and the host vehicle, the change in the shadow due to surrounding structures, etc. Therefore, it is not constant. Therefore, in the conventional technique, the situation that the same feature point cannot be discriminated between the frames of the image is caused, and in this case, the optical flow cannot be calculated. In some cases, it is impossible to detect a vehicle existing in the direction.

  Also, in this type of conventional technology, it is difficult to perform processing in an environment where the processing load is large and processing must be performed by an inexpensive CPU (Central Processing Unit). There was also a problem of incurring a rise in the cost required to build the system.

  In particular, this type of prior art determines that an obstacle exists in the field of view of the imaging unit by image processing using an image captured by the imaging unit. Even if it deviates from the inside, if an attempt is made to determine the presence of an obstacle by the same image processing, the algorithm for the determination becomes more complicated, the CPU is required to have a very high processing capacity, and an increase in cost is inevitable. .

  The present invention has been made in view of the conventional situation as described above, and can accurately detect an obstacle existing in the vicinity of the host vehicle with a simple configuration and perform an appropriate alarm operation. It aims at providing the rear side warning device which can aim at cost reduction.

  The rear side alarm device according to the present invention detects an obstacle using an image picked up by an imaging means having a field of view behind the host vehicle, and performs an alarm operation as necessary. When an object deviates from the field of view of the imaging unit, this is detected, and the luminance change of the vertical end face in the image captured by the imaging unit is observed at that time. The position of the obstacle in is predicted.

  According to the rear side alarm device according to the present invention, when the entire obstacle is present within the field of view of the imaging means, the obstacle is detected using a predetermined image processing technique. After the whole deviates from the field of view of the imaging means, the position of the obstacle is predicted without performing detection by image processing, and an alarm operation based on this is performed, so the load required for calculation processing It is possible to perform an appropriate alarm operation when necessary, while reducing the cost and reducing the cost.

  Hereinafter, specific embodiments of a rear side alarm device to which the present invention is applied will be described in detail with reference to the drawings.

  A rear side warning device to which the present invention is applied uses an image captured by a camera (imaging means) mounted on the own vehicle to detect an obstacle existing on the rear side of the traveling own vehicle, and is necessary In this case, an alarm action is performed. Especially in this rear side warning device, after the entire obstacle has deviated from the field of view of the camera, detection by image processing such as general template matching is not performed, and the passing speed of the obstacle immediately after the departure, etc. Based on the above, the position of the obstacle is predicted.

(First embodiment)
First, a rear side alarm device according to a first embodiment to which the present invention is applied will be described.

  As shown in FIG. 1, the rear side warning device of the present embodiment includes a camera 21 mounted on the host vehicle 10, an image processing device 22 that performs image processing on an image captured by the camera 21, A position prediction device 23 that predicts the position of an obstacle outside the field of view of the camera 21, an alarm control device 24 that performs an alarm operation, a steering angle sensor 25 that detects the steering angle of the steering in the host vehicle 10, and the host vehicle 10 And a blinker switch 26 for operating the blinker and an alarm buzzer 27 for outputting a predetermined alarm sound.

  Of these units, the image processing device 22 is captured when the obstacle deviates from the visual field range of the camera 21 and the obstacle deviation detection means for detecting that the obstacle deviates from the visual field range of the camera 21. The position predicting device 23 predicts the position of an obstacle outside the visual field range of the camera 21 based on the observation result by the brightness change observing means. The alarm control device 24 includes a position of the obstacle within the visual field range of the camera 21 or the position of the obstacle predicted by the position prediction means, the steering operation direction of the host vehicle 10 by the driver, Based on the direction of operation of the winker by the driver, alarm means for performing an alarm operation is configured.

  The camera 21 is mounted on the host vehicle 10 so that the region indicated by the hatched portion A in the drawing, that is, the region having a predetermined viewing angle on the right rear side of the host vehicle 10 is the viewing range. When the overtaking vehicle 30 as a moving obstacle passes over the right side of the own vehicle 10, the camera 21 receives the overtaking vehicle when the overtaking vehicle 30 reaches the right rear side of the own vehicle 10. 30 images can be taken. The image signal output from the camera 21 is supplied to the image processing device 23.

  The image processing device 22 detects an obstacle such as the overtaking vehicle 30 that has entered the visual field range of the camera 21 by performing predetermined image processing on the image signal output from the camera 21. Further, the image processing device 22 detects when an obstacle such as the passing vehicle 30 deviates from the visual field range of the camera 21, and supplies the detection signal to the position prediction device 23.

  The position prediction device 23 predicts the position of an obstacle outside the visual field range of the camera 21 based on the detection signal supplied from the image processing device 22. The position prediction device 23 supplies a prediction signal indicating the predicted position of the obstacle to the alarm control device 24.

  The alarm control device 24 detects the position of the obstacle within the visual field range of the camera 21 detected by the image processing device 22 or the position of the obstacle outside the visual field range of the camera 21 predicted by the position prediction device 23 and the steering angle. Based on the detection signal detected by the sensor 25 and the winker operation direction signal supplied from the winker switch 26, the approaching state between the host vehicle 10 and an obstacle such as the passing vehicle 30 is determined, and both approach. When it is determined that the vehicle is in the state, an alarm operation is performed by operating the obstacle alarm buzzer 27.

  The steering angle sensor 25 detects the steering angle of the steering (not shown) by the driver. A detection signal detected by the steering angle sensor 25 is supplied to the alarm control device 24.

  The blinker switch 26 generates a blinker operation direction signal corresponding to the operation direction of the blinker (not shown) by the driver. The winker operation direction signal generated by the winker switch 26 is supplied to the alarm control device 24.

  The alarm buzzer 27 outputs a predetermined alarm sound under the control of the alarm control device 24.

  The rear side alarm device configured as described above performs obstacle detection and alarm operation through a series of processes as shown in FIGS.

  First, as shown in FIG. 2, the rear side warning device performs image processing in the image processing device 22 based on the image signal output from the camera 21 in step S <b> 1, and causes a failure within the visual field range of the camera 21. It is determined whether or not an object is present, and if an obstacle exists within the visual field range of the camera 21, the position of the obstacle in the image is calculated. As this technique, for example, processing by so-called template matching or optical flow calculation is conceivable. In order to reduce the load on the CPU (Central Processing Unit) as much as possible and to make a system configuration using an inexpensive CPU. It is desirable to employ template matching rather than optical flow calculation, which has a high calculation load. Specifically, the rear side warning device uses, as a template, a feature amount that can be extracted from an image that captures the front of an obstacle passing the host vehicle 10 from behind, such as a tire interval or a headlight interval. After setting, the position of the obstacle in the image is calculated by the matching process using the template.

  Subsequently, in step S2, the rear side alarm device converts the position of the obstacle in the image into an actual spatial position. This conversion process can be easily realized by using a general geometric conversion method. By performing this conversion process, the rear side warning device can set the position of the obstacle relative to the host vehicle 10.

  Subsequently, in step S <b> 3, the rear side warning device determines whether or not an obstacle continues to exist within the visual field range of the camera 21 by the image processing device 22. That is, the rear side warning device detects that the obstacle deviates from the visual field range of the camera 21 by the image processing device 22. At this time, since the rear side warning device grasps the spatial positional relationship between the host vehicle 10 and the obstacle in step S2, it compares the viewing angle of the camera 21 at this spatial position. Thus, the moment when the obstacle starts to deviate from the visual field range of the camera 21 can be easily observed.

  Here, if the rear side alarm device determines that the obstacle is still present within the visual field range of the camera 21, the process proceeds to step S14 in FIG.

  On the other hand, when the rear side warning device determines that the obstacle has deviated from the visual field range of the camera 21, the process proceeds to step S <b> 4, and the image processing device 22 causes the obstacle to be seen by the visual field of the camera 21. Observe the vertical luminance change of the vertical end face in the image taken when deviating from the range. That is, for example, as shown in FIG. 4A, an image captured when the passing vehicle 30 as an obstacle is about to deviate from the field of view of the camera 21 is as shown in FIG. However, the rear side warning device observes the vertical luminance change of the vertical end surface ED corresponding to the traveling direction of the overtaking vehicle 30 in this image.

  Subsequently, in step S5, the rear side warning device obtains the inflection point of the time change graph of the luminance obtained by arranging the observation results in step S4 in time series by the position prediction device 23, and the luminance change starts to occur. The time Ty from the time until the time when the inflection point occurs is calculated.

  Specifically, for example, when the image shown in FIGS. 5A to 5C is captured by the camera 21, the luminance change occurs in the vertical direction of the vertical end surface ED. Since it is when it appears on the ED, the time when the image shown in FIG. When the rear side warning device obtains the luminance change of the vertical end face ED in the image in time series with this time as a reference, for example, a graph shown in FIG. In the graph shown in FIG. 4D, the vertical axis represents luminance expressed in 8-bit gradation, and the horizontal axis represents time (or relative speed of the overtaking vehicle 30). In this graph, the part indicated by a represents the luminance of the pixel in the vertical end face ED of the image shown in FIG. 5A, and the part indicated by b is in the vertical end face ED of the image shown in FIG. The portion indicated by c represents the luminance of the pixel in the vertical end face ED of the image shown in FIG.

  The rear side warning device obtains an inflection point, that is, a portion represented by b from the graph as shown in FIG. 4D by the position prediction device 23, and the inflection point is generated from the time when the luminance change starts to occur. The time Ty until the given time is calculated.

  Further, the rear side warning device calculates the passing speed of the obstacle by the position prediction device 23 in step S6. In general, since the width of the front of the vehicle is often 1.7 m, the rear side warning device determines this value as the default value of the width dimension of the obstacle and calculates this value in step S5. By dividing the time Ty, the speed at which the obstacle passes through the vertical end face ED in the image is calculated. For example, when the rear side warning device detects the entire obstacle existing in the visual field range of the camera 21, the obstacle is based on the tire width obtained from the template, the vehicle contour width in the front lateral direction, or the like. You may make it obtain | require the width dimension of a thing.

  Subsequently, in step S7, the rear side warning device assumes that the obstacle is moving at a constant speed at the passing speed obtained in step S6 by the position predicting device 23 in step S7. After the obtained inflection point occurs, the relative position of the obstacle until the obstacle passes through the host vehicle 10 is predicted. Specifically, when the passing speed V is obtained in step S6, the rear side warning device multiplies the passing speed V by the time T at which the inflection point occurs, for example, as shown in FIG. Thus, the relative position is calculated and set from the end face of the field of view of the camera 21.

  Subsequently, the rear side warning device observes the speed of the host vehicle 10 in step S8, and in step S9, the speed of the host vehicle 10 decelerates compared to when the obstacle passing speed is calculated in step S6. It is determined whether or not.

  Here, when the rear side warning device determines that the speed of the host vehicle 10 is decelerating, the speed of the obstacle is relatively higher than the speed of the host vehicle 10, so that the step is performed. In S10, the speed δV × T for deceleration is added to the passing speed V in Step S6, and the process proceeds to Step S13 in FIG.

  On the other hand, if the rear side warning device determines that the speed of the host vehicle 10 is not decelerated, as shown in FIG. 3, the speed of the host vehicle 10 is determined as the obstacle in step S6 in step S11. It is determined whether or not the vehicle is accelerating more than when the passage speed is calculated.

  Here, if the rear side warning device determines that the speed of the host vehicle 10 is not accelerated, the rear side warning device proceeds to the process of step S13.

  On the other hand, if the rear side warning device determines that the speed of the host vehicle 10 is accelerating, the speed of the obstacle is relatively slower than the speed of the host vehicle 10, so step S12. , The acceleration speed δV × T is subtracted from the passing speed V in step S6.

  Then, in step S13, the rear side warning device determines whether or not the predicted position is within a preset range.

  Here, when the rear side alarm device determines that the predicted position is within the preset range, the process from step S7 in FIG. 2 is repeated while the predicted position is within the preset range. If it is determined that the obstacle is not within the range, it is determined that the obstacle has already deviated from the visual field range of the camera 21, and the process proceeds to processing as a rear side alarm after step S14.

  When the rear side warning device shifts to the processing as the rear side warning, in step S14, the distance between the host vehicle 10 and the obstacle is calculated, and in step S15, the driver approaches the host vehicle 10 to the obstacle. It is determined whether or not the turn signal indicating the direction has been operated.

  Here, when it is determined that the winker is operated, the rear side warning device shifts the process to step S17, while when it is determined that the winker is not operated, step S16. In step (1), it is determined whether or not the driver has steered the host vehicle 10 in a direction to approach the obstacle.

  When the rear side warning device determines that the steering operation is not performed, it determines that there is no possibility of contact with the obstacle, and repeats the processing from step S1 in FIG.

  On the other hand, if the rear side warning device determines that the steering operation has been performed, in step S17, it is determined whether or not the distance between the host vehicle 10 and the obstacle calculated in step S14 is smaller than a predetermined value. judge.

  Then, when it is determined that the distance between the host vehicle 10 and the obstacle is equal to or greater than the predetermined value, the rear side warning device determines that there is no possibility of contact with the obstacle, and the step in FIG. The processing from S7 is repeated.

  On the other hand, when it is determined that the distance between the host vehicle 10 and the obstacle is smaller than the predetermined value, the rear side warning device determines that there is a possibility of contact with the obstacle when the traveling lane is changed. In step S18, a predetermined alarm operation is performed using the rear side alarm device 24 in order to call the driver's attention, and the series of processes is terminated.

  The rear side alarm device detects an obstacle existing behind the host vehicle 10 through a series of processes as described above, and the host vehicle 10 may come into contact with the obstacle. An appropriate alarm action can be performed.

  In the above description, for convenience of explanation, it is assumed that the obstacle passes from the rear to the front of the host vehicle 10 and there is no relative speed change when the obstacle passes through the end face of the field of view of the camera 21. Although it is assumed, as a form of passing through the side vehicle, it is conceivable that the host vehicle 10 overtakes a vehicle existing on the side.

  Thus, when the own vehicle 10 overtakes the vehicle existing on the side, the rear of the vehicle enters the field of view of the sensor. However, in this case, it is easy for the driver to recognize the vehicle obliquely ahead. Therefore, it is not necessary to use the rear side warning device when the host vehicle 10 passes the obstacle, and this rear side warning device is used until the obstacle completely enters the field of view of the sensor. There is no problem even if the function is stopped.

  For the driver, the range to be noted in the rear side warning is that the position of an obstacle that has approached the driver 21 from the rear in the visual field range of the camera 21 and the front side deviating from the visual field range of the camera 21. It is considered that the obstacle has moved to the front range up to about 20 °. Therefore, it is considered that the rear side warning device should be able to reliably detect an obstacle approaching from the rear so as to cope with the presence of the obstacle.

  On the other hand, in the rear side warning device, there is no guarantee that a simple inflection point is recognized when an obstacle passes through a vertical end face in an image. That is, the graph of the luminance temporal change obtained by arranging the observation results of the vertical luminance change of the vertical end face in the image in time series is, for example, according to the relative speed change when the obstacle passes through the end face. As shown in FIG. 7, a plurality of inflection points may occur.

  In such a case, if the rear side warning device changes the process of step S7 in FIG. 2 and detects a scene in which the slope of the change in luminance shown in FIG. 7 becomes gentle through a plurality of inflection points. Good.

  The scene where the slope of the change in brightness becomes gentle means that all of the front of the obstacle has deviated out of the visual field range of the camera 21 even when the relative speed change occurs. . Therefore, the rear side warning device may predict the position of the obstacle based on the position from the time when the gradient of the change in luminance becomes gentle. In this case, the rear side warning device uses an average value during acceleration / deceleration as the speed for predicting the behavior of the obstacle by dividing the time until the inclination changes from the vehicle width.

  As described above in detail, the rear side alarm device shown as the first embodiment uses the image processing device 22 to detect that an obstacle deviates from the visual field range of the camera 21 and The luminance change of the vertical end face in the image captured when the object deviates from the visual field range of the camera 21 is observed, and based on the observation result, the position prediction device 23 is used to detect the obstacle outside the visual field range of the camera 21. Predict location. That is, this rear side warning device normally detects the obstacle using a predetermined image processing method when the entire obstacle is within the visual field range of the camera 21, When the obstacle deviates from the visual field range of the camera 21, it is necessary to use another image processing method that requires high CPU performance due to the complexity of the necessary algorithm. After deviating from the visual field range of the camera 21, the position of the obstacle is predicted without performing detection by image processing. Therefore, this rear side alarm device can realize a simple sensor calculation configuration.

  In addition, the rear side warning device obtains an inflection point of the temporal change in luminance of the vertical end face in the image by the position predicting device 23, and calculates the passing speed of the obstacle based on the inflection point. That is, the rear side warning device can calculate the speed parameter immediately before the entire obstacle deviates from the visual field range of the camera 21 based on a simple temporal change of the viewing angle end face. Therefore, this rear side alarm device can easily predict the position of the obstacle even in an inexpensive CPU environment, and the speed of the obstacle just before deviating from the visual field range of the camera 21. By measuring, the position prediction accuracy can be improved.

  Further, the rear side warning device is based on the obstacle position predicted by the position prediction device 23, the steering operation direction of the host vehicle 10 by the driver, and the winker operation direction by the driver. Since the alarm operation is appropriately performed under the control of 24, the driver can be appropriately alerted.

  As described above, the rear side warning device can detect an obstacle present on the rear side of the host vehicle with high accuracy and reliability with a simple configuration, while realizing cost reduction. Appropriate alarm action can be performed when necessary.

(Second Embodiment)
Next, a rear side alarm device according to a second embodiment to which the present invention is applied will be described. The rear side warning device of the present embodiment basically detects both ends of the visual field range of the camera 21 when detecting an obstacle existing in the visual field range of the camera 21 by template matching from the image captured by the camera 21. If a window for matching is set, and the inflection points to be found are not found within a predetermined time, or if a large number of inflection points are found within a predetermined time, the window width is increased. The field of view is scanned over a wide range. The following description will focus on the characteristic features of the rear side warning device of the present embodiment, and the same components as those in the first embodiment will be described with the same reference numerals and the same step numbers. Is omitted.

  As shown in FIG. 8, the rear side alarm of this embodiment includes a camera 21, an image processing device 22, a position prediction device 23, an alarm control device 24, a steering angle sensor 25, and a winker switch 26 mounted on the host vehicle 10. In addition to the alarm buzzer 27, a window setting device 28 for setting a window for performing template matching, and a setting window for the window setting device 28 based on the result of observing the luminance change of the visual field end face in the image processing device 22. And a window width designating device 29 for designating the width of the image.

  The window setting device 28 sets a range (window) for performing template matching for detecting an obstacle existing in the visual field range of the camera 21 in the image processing device 22. The window width designating device 29 designates the window width to be set for the window setting device 28. In particular, the window width designating device 29 is an inflection point to be obtained in step S7 in FIG. Is not obtained within a predetermined time, or when a large number of inflection points are obtained within a predetermined time, the window setting device 28 is scanned so that a wide area other than both ends of the field of view of the camera 21 is also scanned. On the other hand, it is specified to set the window with a wide width.

  Such a rear side warning device basically performs the same processing as the series of processing shown in FIGS. 2 and 3 in order to detect an obstacle existing behind the host vehicle 10. However, in the process at step S1 in FIG. 2, the window setting device 28 performs a window setting process, and the window width by the window width instruction device 29 is set after the process at step S7 in FIG. The specified process is added.

  That is, the rear side warning device of this embodiment sets a window for performing template matching in the process of step S1 in FIG. Here, the rear side warning device considers that the obstacle enters from both ends of the visual field range of the camera 21 when there is no obstacle in the visual field range of the camera 21. A window is set at both ends of the visual field range, and both ends of the visual field range (both ends of the image captured by the camera 21) are scanned to monitor whether an obstacle enters the visual field range.

  Specifically, as shown in FIG. 9, in step S31, the window setting device 28 reads the window width specified by the window width specifying device 29, and sets a window for template matching. In step S32, it is determined whether or not an obstacle has appeared as a result of scanning the areas in the window set at both ends of the visual field of the camera 21.

  Here, when it is determined that no obstacle has appeared, the rear side warning device repeats the process of step S32, while when it is determined that an obstacle has appeared, in step S33. Then, template matching is performed within the set window, and the process proceeds to step S2 and subsequent steps in FIG.

  Then, when the rear side alarm device completes the processing from step S2 to step S7 in FIG. 2, whether or not an inflection point has been obtained by the window width designating device 29 in step S51 as shown in FIG. Determine.

  Here, if the rear side alarm device determines that the inflection point has not been obtained, after the window width designating device 29 sets the window width set in FIG. Then, the process proceeds to step 8 and the subsequent steps in FIG.

  On the other hand, when it is determined that the inflection point has been obtained, the rear side alarm device sets the window width set in FIG. 9 to be small by the window width designation device 29 in step S53, and FIG. The process proceeds to middle step 8 and subsequent steps.

  As described above, the rear side warning device uses the function of monitoring the inflection point to set the window by the window setting device 28 when performing template matching, and the inflection point to be obtained is within a predetermined time. If a large number of inflection points are obtained within a predetermined time, the window width designating device 29 designates the window setting device 28 to perform setting with a wide window width. .

  The rear side warning device knows that it is difficult to detect the behavior of the obstacle from the image due to the influence of optical disturbance etc. when setting with a wide window width is specified. In order to improve the reliability, the image processing device 22 scans all the regions in the field of view of the camera 21 again.

  At this time, the rear side alarm device does not always scan, but performs special processing only when a specific abnormality occurs in order to reduce the CPU processing speed and realize a system at low cost. To.

  Even in the case described above, the rear side alarm device scans all areas in the field of view and detects an obstacle, and from that point on, the rear side alarm device shown as the first embodiment is used. The process of step S1 in FIG. In this case, the rear side warning device only needs to pay attention to the one having a great influence on the own vehicle 10 even when there are a plurality of obstacles on the rear side in the processing of the rear side warning. Therefore, only the obstacle is traced using a window having a small window width.

  As explained in detail above, the rear side warning device shown as the second embodiment is not able to obtain an inflection point within a predetermined time or many inflection points are obtained within a predetermined time. In this case, since the image processing device 22 scans all the regions in the field of view of the camera 21 again, it is possible to calculate the speed of the obstacle based on the inflection point due to the influence of an optical disturbance or the like. Even when it is difficult to detect the behavior of the obstacle from the image, the obstacle can be reliably detected, and the detection reliability can be improved.

  In addition, when detecting an obstacle from the image captured by the camera 21 by template matching, the rear side warning device sets a window for performing matching, so that the width of the window can be appropriately changed. It is possible to trace only the obstacle using a window having an appropriate window width while paying attention only to the obstacle having a large influence on the host vehicle 10, and to reduce the load on the CPU.

  Although the embodiment of the present invention has been specifically described above, it is needless to say that the present invention is not limited to the above-described embodiment. For example, in each of the above-described embodiments, the alarm operation is performed by outputting a predetermined alarm sound. However, the present invention performs the alarm operation by displaying predetermined visual information on the display screen. Any method may be applied as the alarm operation.

  Further, in the above-described embodiment, each part constituting the rear side alarm device has been described as a physically separate device. However, the present invention is configured as software that can be executed by a CPU or the like, for example. Of course you can also. In addition, even if the present invention is in a form other than the above-described embodiment, various modifications can be made depending on the design and the like as long as the technical idea of the present invention is not deviated. Not too long.

It is a block diagram which shows the structure of the rear side warning apparatus of 1st Embodiment. It is a flowchart explaining a series of processes which perform an alarm operation when an obstacle existing on the rear side of the host vehicle is detected by the rear side alarm device of the first embodiment and there is a possibility of contact. It is a flowchart explaining a series of processes which perform an alarm operation when an obstacle existing on the rear side of the host vehicle is detected by the rear side alarm device of the first embodiment and there is a possibility of contact. It is a figure for demonstrating a mode that the vertical direction luminance change of the vertical end surface in an image when an obstacle deviates from the visual field range of a camera, (A) is an overtaking vehicle as an obstacle from the visual field range of a camera. FIG. 4B is a diagram illustrating a state in which the camera is about to deviate, and FIG. 5B is a diagram illustrating an image captured by the camera in the situation illustrated in FIG. It is a figure for demonstrating the process which analyzes the result of having observed the vertical direction luminance change of the vertical end surface in the image when an obstacle deviates from the visual field range of a camera, and (A) is the time when the luminance change began to occur. It is a figure which shows the imaged image, (B) is a figure which shows the image imaged at the time after the image shown to (A), (C) is further after the image shown to (B). (D) is a diagram showing a graph in which the luminance change of the vertical end face in the images shown in (A) to (C) is obtained in time series. It is a figure for demonstrating the process which estimates the relative position of the said obstacle after the inflection point shown in FIG.5 (D) arises until an obstacle passes the own vehicle. It is a figure which shows the graph which calculated | required the luminance change of the vertical end surface in an image in time series, Comprising: It is a figure for demonstrating a mode that several inflection points have arisen. It is a block diagram which shows the structure of the rear side warning apparatus of 2nd Embodiment. It is a flowchart explaining the process which detects the obstruction which exists in the visual field range of a camera with the rear side warning apparatus of 2nd Embodiment. It is a flowchart explaining the process which designates a setting window width | variety according to the observation result of a luminance change in the rear side warning apparatus of 2nd Embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 10 Own vehicle 21 Camera 22 Image processing apparatus 23 Position prediction apparatus 24 Alarm control apparatus 25 Steering angle sensor 26 Winker switch 27 Alarm buzzer 28 Window setting apparatus 29 Window width designation apparatus 30 Passing vehicle

Claims (5)

In a rear side warning device that detects an obstacle using an image picked up by an imaging means having a field of view on the rear side of the host vehicle, and performs a warning operation as necessary,
Obstacle departure detecting means for detecting an obstacle deviating from the field of view of the imaging means;
A luminance change observation means for observing a luminance change of a vertical end face in an image captured when the obstacle deviates from the visual field range of the imaging means;
A rear side warning device comprising: a position prediction unit that predicts the position of the obstacle outside the field of view of the imaging unit based on an observation result by the luminance change observation unit. The said position prediction means calculates | requires the inflection point of the time change of the brightness | luminance of the vertical end surface in the said image, and calculates the passage speed of the said obstacle based on this inflection point. Rear side alarm device. When the inflection points are not obtained within a predetermined time, or when a large number of inflection points are obtained within a predetermined time, all the areas in the image are scanned again, and the imaging means The rear side warning device according to claim 2, wherein whether or not the obstacle exists within a visual field range is confirmed. Window setting means for setting a window for performing template matching for calculating the position of the obstacle in the field of view of the imaging means;
4. The rear apparatus according to claim 1, further comprising: a window width designating unit that designates a width of a setting window for the window setting unit based on an observation result by the luminance change observation unit. Side alarm device. The position of the obstacle within the field of view of the imaging means or the position of the obstacle outside the field of view of the imaging means predicted by the position prediction means, the steering operation direction of the host vehicle by the driver, and the driving The rear side warning device according to any one of claims 1 to 4, further comprising warning means for performing a warning action based on an operation direction of the winker by a person.

Images