JP2001043377A - External monitoring device having failsafe function - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an external monitoring device which can secure an assured failsafe by detecting accurately the abnormality of a photographed image that is caused by the influence of light such as the weak backlight. SOLUTION: This external monitoring device can secure a failsafe to a decided fail and includes a pair of cameras 1 and 2 which photograph the outside scenes of a vehicle and a fail deciding part 12. The part 12 calculates the horizontal luminance distribution characteristics of the images photographed by the cameras 1 and 2 and then decides a fail if the calculated luminance distribution characteristics have the characteristic of a luminance distribution which occurs in the horizontal direction of an abnormal image that is to be decided as a fail.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an external monitoring apparatus that performs fail-safe when an abnormality occurs in a captured image due to the influence of light such as weak backlight.

[0002]

2. Description of the Related Art In recent years, attention has been paid to a vehicle exterior monitoring device using a vehicle-mounted camera having a solid-state imaging device such as a CCD.
This device recognizes a driving environment (for example, a distance between an object outside the vehicle and the own vehicle) based on an image captured by a vehicle-mounted camera, and calls attention to a driver as necessary. And vehicle control such as downshifting. For example, when recognizing a traveling environment using a distance measurement technique based on a stereo method, a positional difference (parallax) of a certain object in a pair of captured images is obtained, and the target object is determined using the principle of triangulation. (Distance information) can be obtained.

[0003] In order to put such a vehicle exterior monitoring device into practical use, it is necessary to provide a fail-safe function in order to guarantee the safe operation of the device. One of the failures to be detected by this type of device is a situation in which weak light is directly radiated from the sun to an in-vehicle camera and image recognition cannot be performed (a weak backlight situation). Here, a captured image in front of the vehicle as shown in FIG. 7 will be considered. When a weak backlight situation occurs,
As shown in FIG. 8, a ring-shaped luminance saturation region (hereinafter, referred to as a ring-shaped luminance saturation region)
This is called a backlight circle) occurs in a part of the image (the left side in the figure). If such a backlight circle occurs in an image, the luminance value of the pixel in the image area and the vicinity thereof reaches the saturation level, so that correct image information cannot be obtained. As a result, even if a three-dimensional object exists in the area, it is masked by the backlight circle, and it becomes difficult to correctly recognize the three-dimensional object. Further, the edge of the backlight circle has a large luminance change between adjacent pixels in the horizontal direction, so that incorrect distance information may be calculated at the edge portion. FIG. 9 shows FIG.
FIG. 7 is a diagram showing distance information obtained from a stereo image (a reference image and a comparison image) having such an image. In the figure, a portion indicated by a black point is a portion where the luminance change between pixels adjacent to each other in the horizontal direction of the image is large, and this portion has distance information (depth). In the area where the backlight circle has occurred, since the luminance of most pixels has reached the saturation level, the distance data to be calculated originally is not calculated. Further, at the edge of the backlight circle, since the luminance change in the horizontal direction is large, erroneous distance data is calculated. From the viewpoint of ensuring the safety of the exterior monitoring device at a high level, in a situation where a correct distance image cannot be obtained due to weak backlight, it is necessary to operate a fail-safe function for temporarily suspending monitoring control.

[0004]

As described above, a fail-safe function indispensable for the practical use of the vehicle exterior monitoring device has not been established so far, and a problem remains to secure the safety of the device at a high level. I was

[0005] Therefore, an object of the present invention is to provide a fail-safe system in which, when an abnormality occurs in a picked-up image due to the influence of light such as weak backlight or the like, such an image abnormality is accurately detected. It is an object of the present invention to provide a monitoring device outside the vehicle.

[0006]

According to a first aspect of the present invention, there is provided an external monitoring apparatus for performing fail-safe when a failure is determined, an image capturing means for capturing a scene outside the vehicle, Calculating means for calculating the luminance distribution characteristics in the horizontal direction of the captured image obtained by the means, and the luminance distribution characteristics calculated by the calculating means have characteristics of the luminance distribution appearing in the horizontal direction of the abnormal image to be determined as a failure Further, there is provided an outside monitoring apparatus having a fail-safe function having a determination means for determining a failure.

A second invention is a vehicle exterior monitoring device that performs fail-safe when a failure is determined.
Image capturing means for capturing the scenery outside the vehicle, and by dividing the captured image obtained by the image capturing means into a plurality of pieces, a plurality of monitoring areas extending in the vertical direction of the image are set, and for each of the monitoring areas, Calculating means for calculating the overall brightness level of the monitoring area; and a brightness distribution characteristic in the horizontal direction of the captured image obtained from a relationship obtained from a relationship between the position of the monitoring area and the brightness level of each monitoring area. And determining means for determining a failure when the image has a characteristic of a luminance distribution appearing in the horizontal direction of the abnormal image to be determined as a failure.

Here, a typical example of an abnormal image is as follows.
Due to optical effects, in particular weak backlight situations, there are images in which there is a backlight circle in the image. Characteristics of the luminance distribution in the horizontal direction of the image where the backlight circle exists (backlight characteristic)
Examples include the following.

(1) Variation in luminance dispersion characteristics The luminance dispersion characteristics of an image in a weak backlight situation tend to be larger than that of a normal image in which the image does not occur. Focusing on such characteristics, the determination unit:
When the variance value of the luminance variance characteristic of the captured image is larger than a predetermined determination value, it is determined that a failure has occurred. In addition, the determination unit may determine that a failure has occurred when the difference between the maximum value and the minimum value of the luminance dispersion characteristics of the captured image is larger than a predetermined determination value.

(2) Change near the peak of the luminance dispersion characteristic In the vicinity of the peak of the luminance dispersion characteristic in an image in a weak backlight situation, the luminance sum tends to change more smoothly than in a normal image in which it does not occur. There is. In view of such a feature, the determination unit determines that a failure has occurred when the sum of the secondary differential values near the peak of the luminance dispersion characteristic in the captured image is smaller than a predetermined determination value.

(3) Peak value of luminance dispersion characteristic The peak of the luminance dispersion characteristic in an image in a weak backlight condition is as follows:
It tends to be larger than a normal image in which it has not occurred. From such a feature, the determination unit determines that a failure has occurred when the peak value of the luminance dispersion characteristic in the captured image is larger than a predetermined determination value.

(4) Difference in luminance sum between upper and lower areas of the monitoring area A mask area is set at an intermediate portion of each monitoring area. The calculating means calculates the sum of the luminance of the upper region excluding the mask region and the sum of the luminance of the lower region. The judging means judges a failure when the sum of the luminances in the upper region is larger than the predetermined judgment value in the lower region.

(5) Parallax of stereo camera Further, when a stereo camera composed of a plurality of cameras is used as an imaging means and the distance of a backlight circle generated in an image in a weak backlight situation is calculated, the distance is determined to be infinity. Become. Therefore, the positions of the backlight circles in the pair of stereo images are theoretically the same. In view of these characteristics,
The determining unit determines a failure when the difference between the peak position of the luminance sum in one captured image of the camera and the peak position of the luminance sum in the other captured image of the camera is within a predetermined threshold value.

By appropriately combining the above (1) to (5) in consideration of the above features, the detection accuracy of the weak backlight situation can be further improved.

[0015]

FIG. 1 is a block diagram of a vehicle exterior monitoring apparatus according to this embodiment. A pair of cameras 1 and 2 having a built-in image sensor such as a CCD are attached at predetermined intervals in a vehicle width direction of a vehicle such as an automobile, and capture an image of a scene in front of the vehicle. The main camera 1 captures a reference image (right image) necessary for performing stereo processing, and the sub camera 2 captures a comparison image (left image) in this processing. In a state where the cameras 1 and 2 are synchronized with each other,
Each of the analog images output from 2 is converted into a digital image of a predetermined luminance gradation (for example, 256 gray scales) by A / D converters 3 and 4. The digitized image is subjected to luminance correction, geometric conversion of the image, and the like in the image correction unit 5. Usually, since the mounting positions of the pair of cameras 1 and 2 have an error to some extent, a shift caused by the error exists in the left and right images. In order to correct this displacement, geometric transformation such as rotation or translation of the image is performed using affine transformation or the like. The reference image and the comparative image corrected in this way are stored in the original image memory 8.

On the other hand, the stereo image processing unit 6 calculates the three-dimensional position (including the distance from the vehicle to the object) of the same object in the image from the reference image and the comparison image corrected by the image correction unit 5. calculate. This distance can be calculated based on the principle of triangulation from the relative displacement of the position of the same object in the left and right images. The distance information of the image thus calculated is stored in the distance data memory 7.
Is stored in

The microcomputer 9 recognizes a road condition or the like in front of the vehicle based on the information stored in the original image memory 8 and the distance data memory 7 (road recognition unit 10), or a three-dimensional object (vehicle) in front of the vehicle. (Three-dimensional object recognition unit 1)
1). Then, the processing unit 13 outputs these recognition units 10, 1
When it is determined from the information from 1 that an alarm is required, the driver is alerted by an alarm device 19 such as a monitor or a speaker, or the various control units 14 to 18 are controlled as necessary. For example, it instructs an AT (automatic transmission) control unit 14 to execute downshifting. Further, it may instruct the engine control unit 18 to reduce the engine output. In addition, an anti-lock brake system (ABS) controller 15, a traction control system (TCS) controller 16, or
Appropriate vehicle control can also be instructed to the vehicle behavior control unit 17 that controls the torque distribution and rotation speed of each wheel.

Further, the fail determining unit 12 performs a fail determination based on the original image information stored in the original image memory 8 and the distance information stored in the distance data memory 7 in accordance with a routine described later. During a period in which a failure is determined, that is, a period in which the fail flag NG is set to 1 by a failure determination routine described later,
In order to prevent malfunction of the device due to erroneous recognition of a road or a three-dimensional object, the above-described vehicle control or the like is temporarily suspended.

FIG. 2 is a flowchart for explaining a procedure for determining a failure due to weak backlight. This routine is repeatedly executed by the fail determination unit 12 at a predetermined control cycle (for example, 100 ms). Hereinafter, one image captured by one of the stereo cameras will be described, but the same determination is performed on an image captured by the other camera. This fail determination is based on the assumption that the shutter speed is relatively low (daytime or the outside of the vehicle is bright). Then, when the fail flag NG is set to 1 with respect to any one of the images, the processing unit 13 executes the fail safe of the monitoring apparatus outside the vehicle.

First, the failure determining unit 12 calculates the luminance sum A of each monitoring area set in the image area based on the luminance value of the original image stored in the original image memory 8 (step 1). . Whether or not a backlight circle due to a weak backlight situation occurs in the image area is determined based on the overall luminance distribution characteristics in each monitoring area. Here, the monitoring area is set in the image area as shown in FIG. 20
An image area having a size of 0 × 512 pixels is divided into 16 pixels in the horizontal direction, and a rectangular (extends in the vertical direction of the image) monitoring area Ni (1 ≦ 1) each having a size of 200 × 16 pixels. i ≦ 32) is set. Then, an intermediate portion of each monitoring area Ni is defined as a mask area (an area indicated by hatching in the figure). The luminance values of the pixels in these mask areas are not considered in the fail determination.
Therefore, the fail determination is performed based on the luminance values of the pixels in the image upper and lower regions excluding the mask region. The first reason for providing the mask region is to improve the detection accuracy of a weak backlight situation. Even if there is a three-dimensional object with extremely high brightness (for example, a car with a white exterior) or reflected light from a preceding vehicle, if the middle part of the image is masked, the brightness of the monitoring area increases due to the effect of the three-dimensional object. Can be avoided. In addition, by excluding the mask area from the calculation target, the amount of calculation required for fail determination can be reduced.

The luminance sum A is calculated for each monitoring area N set as described above. The luminance sum Ai in a certain monitoring region Ni can be calculated as a value obtained by extracting a plurality of pixels uniformly distributed in the region as a sample and adding their luminance values (or their average value). . For example, a sample may be extracted every four pixels in both the horizontal and vertical directions. If a sample that is uniformly dispersed in the monitoring area is used, the luminance sum A in which the characteristics of the luminance state in the monitoring area are appropriately reflected can be calculated with a small amount of calculation. If the calculation amount is not considered, it is naturally possible to calculate the luminance value A by adding the luminance values of all the pixels included in the monitoring area.

Next, in step 2, step 1
, A distribution characteristic of the calculated sum A is calculated, and a sum of peaks A exceeds a predetermined threshold value Ath.
It is searched whether there is a peak. A necessary condition for the luminance sum peak is that the luminance sum A is a local maximum value equal to or larger than the determination threshold value Ath. Here, the relationship between the backlight circle and the distribution characteristic of the luminance sum A will be described with reference to FIGS. FIG.
8 is a diagram showing characteristics of a luminance sum A in the image (normal image) shown in FIG. FIG. 4 shows a luminance sum A in an image (abnormal image) in which a backlight circle is generated as shown in FIG.
FIG. 4 is a diagram showing characteristics of the present invention. These characteristics may be said to be the characteristics of the luminance distribution in the horizontal direction of the image. As can be seen from these luminance distribution characteristics, one of the features of the abnormal image in which the backlight circle occurs is that the luminance sum A of the monitoring area where the backlight circle exists is much larger than that of the normal image. Can be Therefore, when a backlight circle having a large luminance occurs, an appropriate value such that the luminance sum A exceeds it is set as the determination threshold value Ath. as a result,
If a negative determination is made in this step (for example, in the case of the distribution characteristics in FIG. 5), the image is determined to be a normal image, and the fail flag NG is set to 0 (step 10). During the period when the fail flag NG is 0, the normal monitoring control is continued.

If an affirmative determination is made in step 2 (for example, in the case of the distribution characteristics in FIG. 4), it is determined that there is a possibility that a backlight circle exists in the image, and the process proceeds to step 3 in order to further verify this. move on. In this step, the sum SUM of the second derivative values near the peak value Apeak described above
(Hereinafter referred to as a differential sum) is calculated. FIG. 6 is a diagram for explaining a method of calculating the differential sum SUM. First, a difference ΔA2 between the luminance sum Apeak of the monitoring area n1 at which the luminance sum peaks and the luminance sum A of the monitoring area (n1-1) one to the left of the area n1 is determined. Next, the monitoring area (n1-1)
And a difference ΔA1 between the luminance sum A of the monitoring area (n1-2) and the luminance sum A of the monitoring area (n1-2) one side to the left of the area (n1-1).
Similarly, the difference ΔA3,
Find ΔA4. The differential sum SUM is calculated by adding these four values (absolute values). The gentleness of the change near the peak can be evaluated from the differential sum SUM calculated in this manner. That is, as a peak characteristic in a normal image, even if there is a vicinity of a peak exceeding the determination threshold value Ath, the luminance sum A in the vicinity thereof tends to change rapidly (that is, the peak becomes steep) ( (See FIG. 5). Therefore, for a normal image, the differential sum SUM has a relatively large value. On the other hand, the abnormal screen has a gentle peak at the position of the backlight circle, and has broad characteristics. Therefore, since the change of the luminance sum A near the peak becomes relatively small,
The differential sum SUM has a relatively small value.

Therefore, an appropriate determination threshold value SU that can distinguish whether the change of the luminance sum A near the peak is smooth or not.
If Mth is set, the gentleness of the peak can be evaluated from the value of the differential sum SUM calculated in step 3 (step 4). If a negative determination is made in this step, that is, if the differential sum SUM is equal to or less than the threshold value SUMth, it is determined that the peak is steep and the peak is not due to the backlight circle, and the failure flag N
G is set to 0 (step 10). Conversely, if the differential sum SUM is smaller than the threshold value SUMth, it is determined that the peak is gentle and there is a possibility that a backlight circle exists, and the process proceeds to step 5 in order to examine the distribution characteristics of the luminance in more detail. When there are a plurality of peaks described above in the distribution characteristics, the differential sum SUM is calculated for each peak. Then, if there is at least one differential sum SUM larger than the determination threshold value SUMth, it is determined that there is a possibility that a retrograde circle exists, and the procedure from step 5 is performed.

In step 5 and subsequent steps, in order to further examine the presence or absence of a backlight circle, the variance VAR of the luminance distribution characteristic of the captured image, which indicates the degree of variation of the luminance sum A, is calculated based on the following equation. (Step 5).

VAR ² = ｛(A 1 −A _ave ) ² + (A 2 −A _ave )
² +... + (An−A _ave ) ² ｝ / n = Σ (Ai ² ) / n−
(ΣA _i / n) ² n: number of monitoring areas (= 32) A _ave : average value of luminance sum

Since the normal image tends to have a relatively small variation in the luminance sum A, the variance value VAR also becomes small (see FIG. 5). On the other hand, in an abnormal image in which a backlight circle exists, the dispersion tends to increase, and the variance VAR also increases. Therefore, if an appropriate threshold value VARth capable of judging whether or not the luminance sum A fluctuates due to the backlight circle is set, the abnormality of the image is judged from the value of the variance value VAR calculated in Step 5. (Step 6). If a negative determination is made in this step, that is, if the variance value VAR is equal to or smaller than the determination threshold value VARth, it is determined that there is no backlight circle, and a failure is determined in step 10 as in step 4. The flag NG is set to 0. On the other hand, if an affirmative determination is made in step 6, the luminance variance in the horizontal direction of the image is large, and there is a possibility that a backlight circle exists in the image.

Reaching step 7 through the above-described steps means that the possibility that a backlight circle exists in the image is considerably high. As described above, the fail-safe temporarily suspends monitoring outside the vehicle, and therefore it is preferable to accurately verify the existence of the backlight circle in order to avoid unnecessary operation of the fail-safe. Therefore, the determinations in steps 7 and 8 are performed.
First, in step 7, the maximum value A in the luminance distribution characteristics
max (sum of brightness A of monitoring area n1 in FIGS. 4 and 5)
It is determined whether or not the difference between Amin and Amin (the luminance sum A of the monitoring area n2) is greater than a threshold value ΔAth as a determination value. As a result, it is possible to prevent a normal image that can be captured when the entire area in front of the vehicle is relatively bright from being determined as an abnormal image. Further, in step 8, it is determined whether or not the left and right images, that is, the peak position n1 of the reference image and the peak position n1 of the comparison image are substantially at the same position. When using the stereo method, the distance of the backlight circle appears as infinity. Therefore, when a backlight circle occurs, the horizontal displacement (parallax) of the peak positions in both images is theoretically zero. Conversely, when a peak is generated due to the presence of a three-dimensional object with extremely high luminance, parallax is generated according to the distance from the vehicle to the three-dimensional object. Therefore, an appropriate threshold value is provided, and it is determined that a backlight circle has occurred only when the difference between the peak position shift amount calculated from the left and right images and the threshold value is small.

If the determinations in steps 2, 4, 6, 7, and 8 are all affirmative, the failure flag NG
Is set to 1, fail-safe is performed, and normal outside monitoring is temporarily stopped.

From the viewpoint of ensuring both the reliability of the monitoring outside the vehicle and the stability of the control, the switching of the NG flag is performed when the above-mentioned determination of normality / abnormality is continued for each predetermined cycle. Is preferred. For example, when setting the fail flag NG from 0 to 1,
Change when the abnormality (failure) judgment continues for 5 cycles (corresponding to 0.5 seconds). When the fail flag NG is set from 1 to 0, the normal judgment is made for 20 cycles (2.0 cycles).
Second).

As described above, in the weak backlight detection method according to the present embodiment, a plurality of rectangular monitoring areas extending in the vertical direction of the image are set by dividing the image, and the characteristic of the luminance sum of each monitoring area is set. That is, the luminance distribution characteristics in the horizontal direction (lateral direction) of the captured image are obtained. When a backlight circle is generated in an image due to weak backlight, a gradation occurs in which the luminance changes relatively slowly in the horizontal direction of the image (see FIG. 4). Based on such knowledge, a characteristic change (distribution characteristic) is extracted from a lateral change pattern generated by the backlight circle, and the extracted characteristic is compared with the luminance distribution characteristic of the captured image to be detected. Thus, the presence or absence of the backlight circle is determined. The following is a summary of the distribution characteristics that are effective for determining the presence or absence of a backlight circle.

(1) The variance of the luminance sum A is large. (2) The peak of the luminance sum A is large. (3) The change near the peak of the luminance sum A is gentle.

Basically, if the items (1) to (3) are verified, it can be determined whether or not the image is an abnormal image. However, if an attempt is made to judge only one item, an erroneous judgment may occur. Therefore, it is preferable to use these in combination. In addition, in order to further increase the detection accuracy of the weak backlight situation, it is preferable to consider the following items.

(4) The difference between the maximum value and the minimum value of the luminance sum A is large. (5) In the stereo system, the peak positions of the reference image and the comparison image are almost the same.

The items (1) to (4) are not specific to the stereo camera. Therefore, as long as the weak backlight situation is detected by appropriately combining them, the present invention can be applied to a vehicle exterior monitoring device using a monocular camera or the like.

Further, although not determined in the flowchart of FIG. 2, it is also effective to refer to the following items.

(6) Average luminance value (or luminance saturation rate) in the upper area of the monitoring area (upper part of the mask area)
Has a larger average luminance value (or luminance saturation rate) than the lower area (the lower part of the mask area).

Here, the luminance saturation rate refers to a saturation level (for example, a luminance value of 2) for the entire pixel in a predetermined area.
It means the ratio occupied by pixels that have reached 240 or more in 56 tones.

As is clear from the above description, the main feature of the present invention is that the horizontal luminance distribution characteristics of an image are determined in the horizontal direction of an abnormal image to be judged as a failure (for example, an image having a backlight circle). The feature is to detect whether or not the image is an abnormal image by comparing with a backlight characteristic set in advance as a feature of the luminance distribution appearing in the direction. As in the above-described embodiment, a method of setting a determination condition (a magnitude relationship with a determination value) indicating a feature of an abnormal image is the most preferable method from the viewpoint of reducing the amount of calculation. However, the present invention is not limited to this, and can be naturally applied to other methods. For example, a plurality of patterns of the luminance dispersion characteristic in weak backlight are prepared as a template, and when a captured image matches any of the patterns, it is determined that a backlight circle exists in the image, and fail-safe is performed. You may do so.

In this embodiment, an example in which a backlight circle is present in a part of the image due to a weak backlight condition has been described as an example of an abnormal image to be failed. An image in which a backlight circle is present is a typical example that can be most accurately detected by the method according to the present invention. However, the present invention can also be applied to other optically affected images (for example, generation of smear).

[0040]

As described above, according to the present invention, even if an abnormality occurs in a captured image due to an optical influence or the like, such an image abnormality can be accurately detected. Therefore, even if such an abnormality occurs in a captured image, fail-safe can be performed in response to the abnormality, so that the safety of the vehicle exterior monitoring device can be ensured at a higher level.

[Brief description of the drawings]

FIG. 1 is a block diagram of a vehicle exterior monitoring device according to an embodiment;

FIG. 2 is a flowchart illustrating a procedure for determining a failure due to weak backlight.

FIG. 3 is a diagram for explaining a monitoring area set in an image area;

FIG. 4 is a diagram showing characteristics of a luminance sum A in an abnormal image in which a backlight circle has occurred;

FIG. 5 is a diagram showing a characteristic of a luminance sum A in a normal image.

FIG. 6 is a diagram for explaining a method of calculating a sum SUM of second derivative values;

FIG. 7 shows an example of a normal image.

FIG. 8 is a diagram showing an example of an abnormal image in which a backlight circle has occurred.

FIG. 9 is a diagram showing a distance image including erroneous distance information due to a backlight circle;

[Explanation of symbols]

1 main camera, 2 sub camera, 3,
4 A / D converter, 5 image correction unit, 6 stereo image processing unit, 7 distance data memory, 8 source image memory, 9 microcomputer, 10 road recognition unit, 11 three-dimensional object recognition unit, 12 failure determination unit, 13 processing unit, 14 AT control unit,
15 ABS control unit, 16 TCS control unit, 17 torque balance control unit, 18 engine control unit, 19 alarm device

Continued on front page F-term (reference) 2F065 AA04 AA06 BB13 CC40 DD00 DD04 DD12 EE08 FF05 FF61 JJ03 JJ26 NN17 NN18 NN20 PP01 QQ00 QQ08 QQ13 QQ23 QQ24 QQ25 QQ27 QQ29 QQ37 QQ41 QQ42 RR00ACB CB08 CB08 CH08 DA15 DB03 DB09 DC19 DC25

Claims (10)

[Claims] An outside monitoring apparatus that performs fail-safe when it is determined that a failure has occurred, an imaging unit that captures a scene outside the vehicle, and a calculation that calculates a luminance distribution characteristic in a horizontal direction of a captured image obtained by the imaging unit. Means, and a determination means for determining a failure when the luminance distribution characteristic calculated by the calculation means has a characteristic of a luminance distribution appearing in the horizontal direction of the abnormal image to be determined as a failure. Outside monitoring device with safe function. 2. An outside-vehicle monitoring device that performs fail-safe when it is determined that a failure has occurred, wherein: an imaging unit that captures a scene outside the vehicle; and a captured image obtained by the imaging unit is divided into a plurality of images. Calculating means for setting a plurality of monitoring areas extending in the vertical direction of the image, and calculating, for each of the monitoring areas, the magnitude of the overall luminance of the monitoring area; and When the luminance distribution characteristic in the horizontal direction of the captured image obtained from the relationship with the size has the characteristic of the luminance distribution appearing in the horizontal direction of the abnormal image to be determined as a failure, it has a determination unit for determining a failure. An external monitoring device with a fail-safe function. 3. The outside monitoring apparatus having a fail-safe function according to claim 1, wherein the abnormal image is an image in which a backlight circle exists in the image due to a weak backlight situation. . 4. The method according to claim 1, wherein the determining unit determines that a variance of the luminance variance characteristic of the captured image is larger than a criterion value.
The outside monitoring device having a fail function according to any one of claims 1 to 3, wherein it is determined that the vehicle has failed. 5. The method according to claim 1, wherein the determination unit determines a failure when the sum of the secondary differential values near the peak of the luminance dispersion characteristic of the captured image is smaller than a determination value. 4. An external monitoring device having a fail function according to any one of the above items 4 to 4. 6. The image processing apparatus according to claim 1, wherein the determination unit determines a failure when a peak value of the luminance dispersion characteristic of the captured image is larger than a determination value. 2. An external monitoring device having a fail function according to claim 1. 7. The method according to claim 1, wherein the determination unit determines a failure when a difference between a maximum value and a minimum value of the luminance dispersion characteristic of the captured image is larger than a determination value. A vehicle exterior monitoring device having the fail function according to claim 6. 8. A mask area is set in an intermediate portion of the monitor area, and the calculating means calculates the luminance sum of the monitor area based on an upper area and a lower area excluding the mask area. The outside monitoring device having a fail-safe function according to any one of claims 1 to 7, wherein: 9. The method according to claim 1, wherein the determining unit determines a failure when a luminance saturation rate of a pixel in the upper area is larger than a luminance saturation rate of a pixel in the lower area. 8. A vehicle exterior monitoring device having a fail function described in 8. 10. The stereo camera comprising a plurality of cameras, wherein the determination unit is configured to determine a peak position of the luminance dispersion characteristic in one captured image of the camera and the other of the cameras. 10. A failure function according to claim 1, wherein a failure is determined when a difference between a peak position of the luminance dispersion characteristic and a peak position of the captured image is within a threshold value. Outside monitoring device.