JP5484123B2 - Vehicle obstacle detection device - Google Patents

Description

  The present invention relates to a vehicle including two types of sensors, for example, an ultrasonic sensor and a camera, and a vehicle that automatically adjusts processing parameters used for the obstacle detection processing of each sensor based on the obstacle detection processing results of both sensors. The present invention relates to a peripheral obstacle detection device.

As means for automatically detecting obstacles around the vehicle, there is means for detecting obstacles using images taken by a camera. However, in the obstacle detection process using the camera, erroneous detection or omission of detection occurs due to the influence of backlight or nighttime. For this reason, a vehicle peripheral obstacle detection device that uses both obstacle detection processing using an ultrasonic sensor that is not affected by backlight or nighttime and obstacle detection processing using a camera has been reported (for example, Patent Document 1). reference).

  In the vehicle surrounding obstacle detection device disclosed in Patent Document 1, when the gear is put in the back to stop the vehicle in the parking lot, the rear camera senses the rear image simultaneously with the CCD camera that captures the rear image. An ultrasonic sensor is installed behind the vehicle.

  The video imaged by the CCD camera is sent to an image obstacle detection unit, and this video image is processed to detect an obstacle. The output result from the ultrasonic sensor is sent to the ultrasonic obstacle detection unit, and the obstacle is detected by processing the output result.

The detection result of the obstacle by the image obstacle detection unit and the ultrasonic obstacle detection unit is sent to the image processing / control unit. In this image processing / control unit, when an obstacle is detected by at least one of the means, a video is displayed in which a mark for identifying the location of the detected obstacle is superimposed on the video taken by the CCD camera. To do. The image created by the image processing / control unit is displayed on a monitor, and gives a warning about an obstacle behind the vehicle to the driver.
The image processing / control unit can also control the CCD camera to display the detected obstacle in an enlarged manner and display the enlarged image on the monitor. As a result, the driver can visually confirm the contents of the obstacle.

As described above, in the vehicle surrounding obstacle detection device disclosed in Patent Document 1, when an obstacle is detected by either the image obstacle detection unit or the ultrasonic obstacle detection unit, this is notified to the driver. Present. Therefore, it is possible to suppress an obstacle detection omission even in backlight or nighttime conditions where obstacle detection processing using a CCD camera is difficult.

JP 2002-29345 A

  In the vehicle surrounding obstacle detection device disclosed in Patent Document 1, in order to cope with adverse conditions such as backlight and nighttime where obstacle detection processing using a CCD camera is difficult, ultrasonic sensors having different characteristics are used in combination. Describes a method for relieving obstacle detection omissions. However, even in an ultrasonic sensor, there is an environment where obstacle detection processing is difficult.

  For example, an ultrasonic sensor makes it difficult to detect an obstacle because ultrasonic waves are diffusely reflected in the air during bad weather such as strong snow. In addition, it is known that detection omissions and detection errors occur in environments such as heavy rain / dense fog / high temperature / extreme cold / inclined road surface / fence / wire mesh.

For this reason, even if a CCD camera and an ultrasonic sensor are simply used together, there is a problem that there is a state where obstacle detection processing is difficult.
In addition, the obstacle detection process itself by each sensor is not particularly changed in processing content, and does not improve the accuracy of the obstacle detection process by adapting to an adverse condition, for example.

  The present invention has been made to solve the above-described problems. In each sensor that detects an obstacle, the sensor is adaptively adjusted to an environmental condition in which the obstacle detection process is difficult, and the stability is improved. It is an object of the present invention to provide a vehicle surrounding obstacle detection device capable of realizing the detection of a high obstacle.

The vehicle surrounding obstacle detection device according to the present invention is described in the ultrasonic sensor parameter table, an ultrasonic sensor for sensing an obstacle with ultrasonic waves with reference to the processing parameters described in the ultrasonic sensor parameter table. An image of an obstacle by referring to the processing parameter described in the camera parameter table and an ultrasonic signal processing unit that detects information on the obstacle based on the sensor output value from the ultrasonic sensor. Referring to the processing parameters described in the camera parameter table, the image processing unit that detects information on the obstacle based on the image captured by the camera, and the ultrasonic signal processing unit and the information of the detected obstacle by information and video processor of the obstacle integration match that obstacle integrating detector, obstacle integration test A display device for displaying information integrated obstacle by parts, based on the information of the ultrasound signal processing section and the detected obstacle by the video processing unit, described in the ultrasonic sensor parameter table or camera parameter table A parameter adjustment unit that adjusts a processing parameter, and the parameter adjustment unit evaluates an overlapping degree between an obstacle area detected by the ultrasonic signal processing unit and an obstacle area detected by the video processing unit. The values are calculated, and the processing parameters described in the ultrasonic sensor parameter table and the camera parameter table are adjusted so that the evaluation value becomes large.
Another vehicle surrounding obstacle detection device according to the present invention refers to an ultrasonic sensor that senses an obstacle with an ultrasonic wave with reference to a processing parameter described in the ultrasonic sensor parameter table, and an ultrasonic sensor parameter table. An ultrasonic signal processing unit that detects information on an obstacle based on a sensor output value by an ultrasonic sensor with reference to the described processing parameter, and an obstacle with reference to the processing parameter described in the camera parameter table Detected by an ultrasonic signal processing unit and a video processing unit that detects information on an obstacle based on the video captured by the camera with reference to a processing parameter described in the camera parameter table An obstacle integrated detection unit for integrating the obstacle information detected and the obstacle information detected by the image processing unit; Displayed in the ultrasonic sensor parameter table or the camera parameter table based on the obstacle information detected by the display device displaying the obstacle information integrated by the detection unit and the ultrasonic signal processing unit and the video processing unit. A parameter adjustment unit that adjusts the processing parameter, and the obstacle integrated detection unit detects an ultrasonic signal when the video processing unit detects a specific pattern that makes it difficult for the ultrasonic signal processing unit to detect information on the obstacle. The display unit notifies that the operation of detecting information on the obstacle by the processing unit cannot be guaranteed.
Another vehicle surrounding obstacle detection device according to the present invention refers to an ultrasonic sensor that senses an obstacle with an ultrasonic wave with reference to a processing parameter described in the ultrasonic sensor parameter table, and an ultrasonic sensor parameter table. An ultrasonic signal processing unit that detects information on an obstacle based on a sensor output value by an ultrasonic sensor with reference to the described processing parameter, and an obstacle with reference to the processing parameter described in the camera parameter table Detected by an ultrasonic signal processing unit and a video processing unit that detects information on an obstacle based on the video captured by the camera with reference to a processing parameter described in the camera parameter table An obstacle integrated detection unit for integrating the obstacle information detected and the obstacle information detected by the image processing unit; Displayed in the ultrasonic sensor parameter table or the camera parameter table based on the obstacle information detected by the display device displaying the obstacle information integrated by the detection unit and the ultrasonic signal processing unit and the video processing unit. A parameter adjustment unit that adjusts the processing parameter, and the obstacle integrated detection unit detects an obstacle by the video processing unit when the video processing unit detects a specific pattern that makes it difficult for the video processing unit to detect obstacle information. The display device notifies that the information detection operation of the object cannot be guaranteed.

According to the vehicle periphery obstacle detection apparatus of the present invention, since the structure described above, based on the harmful substance detection result disabled using an obstacle detection result and camera using an ultrasonic sensor, and an ultrasonic sensor ultrasonic sonication parameter processing unit or by automatically adjusting the process parameters of the camera you good beauty Film image processing unit, in each sensor that detects an obstacle, adaptively against challenging environmental conditions obstacle detection Adjustment is performed, and highly stable obstacle detection can be realized.
According to another vehicle peripheral obstacle detection device of the present invention, when the video processing unit detects a specific pattern in which it is difficult to detect information on the obstacle by the ultrasonic signal processing unit, the obstacle by the ultrasonic signal processing unit It is configured to notify that the information detection operation cannot be guaranteed, so it is possible to call the driver's attention when a situation is found in which the operation of the obstacle detection processing by the ultrasonic signal processing unit is not guaranteed. It becomes.
According to another vehicle surrounding obstacle detection device of the present invention, when the video processing unit detects a specific pattern in which it is difficult for the video processing unit to detect obstacle information, the video processing unit detects obstacle information. Since it is configured to be notified that the operation cannot be guaranteed, it is possible to call the driver's attention when a situation is found in which the operation of the obstacle detection process by the video processing unit is not guaranteed.

It is a block diagram which shows the structure of the vehicle periphery obstruction detection apparatus which concerns on Embodiment 1 of this invention. It is a figure which shows the example of attachment to the vehicle of the ultrasonic sensor and camera in Embodiment 1 of this invention. It is a flowchart which shows operation | movement of the vehicle surrounding obstacle detection apparatus which concerns on Embodiment 1 of this invention. It is a figure which shows the example of an obstacle detection result on favorable conditions in Embodiment 1 of this invention. It is a figure which shows the example of an obstacle detection result before parameter adjustment in case of snowing in Embodiment 1 of this invention. It is a figure which shows the example of an obstacle detection result after parameter adjustment in the case of snowing in Embodiment 1 of this invention. It is a figure which shows the example of an obstacle detection result before parameter adjustment in case the illumination in Embodiment 1 of this invention is dark. It is a figure which shows the example of an obstacle detection result after parameter adjustment when the illumination in Embodiment 1 of this invention is dark. It is a figure which shows the example of an obstacle detection result before the parameter adjustment in Embodiment 2 of this invention. It is a figure which shows the example of an obstacle detection result after the parameter adjustment in Embodiment 2 of this invention. It is a figure which shows the example of an obstacle detection result of the difficult detection situation in Embodiment 3 of this invention.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
Embodiment 1 FIG.
1 is a block diagram showing the configuration of a vehicle peripheral obstacle detection device according to Embodiment 1 of the present invention.
As shown in FIG. 1, the vehicle surrounding obstacle detection device includes an ultrasonic sensor 101, an ultrasonic signal processing unit 102, a camera 103, a video processing unit 104, an obstacle integrated detection unit 105, a display device 106, and ultrasonic sensor parameters. It comprises a table 107, a camera parameter table 108, and a parameter adjustment unit 109.

The ultrasonic sensor 101 performs obstacle sensing with ultrasonic waves while referring to the processing parameters described in the ultrasonic sensor parameter table 107. FIG. 2 shows a case where four ultrasonic sensors 101 are attached to the rear of the vehicle, and a sensing area 201 of each ultrasonic sensor 101 is shown. In this ultrasonic sensor 101, an ultrasonic wave is emitted and a reflected wave from an obstacle is incident.
Sensor output information indicating a sensor output value (intensity of reflected wave) by the ultrasonic sensor 101 is sent to the ultrasonic signal processing unit 102.

The ultrasonic signal processing unit 102 detects the presence or absence of an obstacle based on the sensor output value from the ultrasonic sensor 101 while referring to the processing parameters described in the ultrasonic sensor parameter table 107.
Further, when detecting the obstacle, the ultrasonic signal processing unit 102 specifies the area of the obstacle based on the sensing area 201 of the ultrasonic sensor 101 and the measured distance to the obstacle. Obstacle information indicating the area of the obstacle specified by the ultrasonic signal processing unit 102 is sent to the obstacle integrated detection unit 105.

The camera 103 captures an obstacle image while referring to the processing parameters described in the camera parameter table 108. FIG. 2 shows the case where one camera 103 is attached to the rear of the vehicle, and shows the field of view 202 of this camera.
The video imaged by the camera 103 is sent to the video processing unit 104 and the obstacle integrated detection unit 105.

The video processing unit 104 detects the presence or absence of an obstacle by analyzing the video from the camera 103 while referring to the processing parameters described in the camera parameter table 108.
In addition, when detecting an obstacle, the video processing unit 104 identifies an area of the obstacle. Obstacle information indicating the area of the obstacle specified by the video processing unit 104 is sent to the obstacle integrated detection unit 105.

The obstacle integration detection unit 105 integrates the obstacle information obtained by the ultrasonic signal processing unit 102 and the video processing unit 104 and displays a mark for identifying the obstacle region on the video image captured by the camera 103. It is to generate video. The video generated by the obstacle integrated detection unit 105 is sent to the display device 106. Also, the obstacle information sent to the obstacle integration detection unit 105 is sent to the parameter adjustment unit 109.
The display device 106 displays the video generated by the obstacle integrated detection unit 105.

  The ultrasonic sensor parameter table 107 holds processing parameters referred to when the ultrasonic sensor 101 or the ultrasonic signal processing unit 102 performs processing. In the ultrasonic sensor parameter table 107, for example, a determination threshold value used when performing an obstacle detection in the ultrasonic signal processing unit 102 and a value of a control parameter that determines the shape of the sensing area 201 of the ultrasonic sensor 101 are described. ing.

  The camera parameter table 108 holds processing parameters referred to when the camera 103 or the video processing unit 104 performs processing. The camera parameter table 108 includes, for example, the sensitivity of the camera 103 and parameters related to image processing (edge extraction processing parameters, binarization processing thresholds, etc.) used when the image processing unit 104 performs obstacle detection. Have been described.

  The parameter adjustment unit 109 adjusts the processing parameters described in the ultrasonic sensor parameter table 107 and the camera parameter table 108 based on the obstacle information from each sensor sent to the obstacle integrated detection unit 105.

Next, the operation of the vehicle surrounding obstacle detection device configured as described above will be described.
FIG. 3 is a flowchart showing the operation of the vehicle surrounding obstacle detection device according to Embodiment 1 of the present invention. FIG. 4 is a diagram showing an example of an obstacle detection result under favorable conditions in Embodiment 1 of the present invention.
In FIG. 4, reference numeral 401 is an image captured by the camera 103, reference numeral 402 is a sensor output value of the ultrasonic sensor 101, reference numeral 403 is a determination threshold, reference numeral 404 is an area of an obstacle by the ultrasonic sensor 101, and reference numeral 405 is a camera 103. Shows the area of obstacles.

  In the operation of the vehicle surrounding obstacle device, as shown in FIG. 3, when the vehicle backs and parks, the ultrasonic sensor 101 refers to the processing parameters described in the ultrasonic sensor parameter table 107, An obstacle behind the vehicle is sensed by ultrasonic waves (step ST31). Here, as shown in FIG. 4, when an obstacle exists in the sensing region 201 of the ultrasonic sensor 101, the sensor output value (intensity of the reflected wave) 402 becomes high. Sensor output information indicating a sensor output value from the ultrasonic sensor 101 is sent to the ultrasonic signal processing unit 102.

  Next, the ultrasonic signal processing unit 102 detects the presence or absence of an obstacle based on the sensor output value from the ultrasonic sensor 101 while referring to the processing parameters described in the ultrasonic sensor parameter table 107 (step ST32). Here, as shown in FIG. 4, in the ultrasonic signal processing unit 102, when the sensor output value 402 by the ultrasonic sensor 101 exceeds the determination threshold value 403, an obstacle is present in the sensing direction of the ultrasonic sensor 101. It is determined that it exists.

If the ultrasonic signal processing unit 102 determines in step ST32 that an obstacle is present, the ultrasonic signal processing unit 102 then specifies the area of the obstacle (step ST33). Here, as shown in FIG. 4, for example, when the sensor output values 402 of the central two ultrasonic sensors 101 exceed the determination threshold value 403, it is understood that there are obstacles on these sensing areas 201. . Further, the ultrasonic sensor 101 can measure the distance to the obstacle. Therefore, the ultrasonic signal processing unit 102 can specify the obstacle region 404 based on the distance to the obstacle and the size of the sensing region 201 of the ultrasonic sensor 101 where the obstacle is detected.
Obstacle information indicating the obstacle region 404 identified by the ultrasonic signal processing unit 102 is sent to the obstacle integrated detection unit 105.

  On the other hand, the camera 103 captures an image of an obstacle behind the vehicle while referring to the processing parameters described in the camera parameter table 108 (step ST34). The video imaged by the camera 103 is sent to the video processing unit 104 and the obstacle integrated detection unit 105.

  Next, the video processing unit 104 analyzes the video from the camera 103 while referring to the processing parameters described in the camera parameter table 108, and detects the presence or absence of an obstacle (step ST35). Note that the video processing unit 104 detects an obstacle present in the video by using an existing method such as taking a difference between the background video and the current video.

  In step ST35, if the video processing unit 104 determines that an obstacle exists, the video processing unit 104 then specifies the area of the obstacle (step ST36). Here, as illustrated in FIG. 4, the video processing unit 104 specifies an obstacle region 405 by performing image processing on the video. Obstacle information indicating the obstacle area 405 identified by the video processing unit 104 is sent to the obstacle integrated detection unit 105.

Next, the obstacle integration detection unit 105 integrates the obstacle information obtained by the ultrasonic signal processing unit 102 and the image processing unit 104 and superimposes a mark for specifying the obstacle region on the image captured by the camera 103. The image is displayed and generated (step ST37). Here, as shown in FIG. 4, the obstacle integrated detection unit 105 surrounds the obstacle region 404 identified by the ultrasonic signal processing unit 102 on the image captured by the camera 103 with a solid line, and performs image processing. By surrounding the obstacle region 405 specified by the unit 104 with a dotted line, an image 401 in which the obstacle region is marked is generated.
The video 401 generated by the obstacle integrated detection unit 105 is sent to the display device 106 and displayed, thereby being presented to the driver.
Also, the obstacle information by the ultrasonic signal processing unit 102 and the video processing unit 104 sent to the obstacle integrated detection unit 105 is sent to the parameter adjustment unit 109.

Next, the parameter adjustment unit 109 adjusts the processing parameters described in the ultrasonic sensor parameter table 107 and the camera parameter table 108 based on the obstacle information from each sensor sent from the obstacle integrated detection unit 105 (step) ST38).
Hereinafter, parameter adjustment by the parameter adjustment unit 109 will be described.

First, parameter adjustment by the parameter adjustment unit 109 under conditions in which obstacle detection processing by the ultrasonic signal processing unit 102 is difficult will be described.
FIG. 5 is a diagram showing an example of an obstacle detection result before parameter adjustment when it is snowing in the first embodiment of the present invention. FIG. 6 is a diagram showing an example of an obstacle detection result after parameter adjustment when it is snowing in the first embodiment of the present invention.
In FIG. 5, reference numeral 501 denotes an image captured by the camera 103, reference numeral 502 denotes a sensor output value of the ultrasonic sensor 101, reference numeral 503 denotes a determination threshold value, and reference numeral 504 denotes an obstacle area by the video processing unit 104.
In FIG. 6, reference numeral 601 is an image captured by the camera 103, reference numeral 602 is a sensor output value of the ultrasonic sensor 101, reference numeral 603 is a determination threshold value, reference numeral 604 is an obstacle area by the ultrasonic signal processing unit 102, and reference numeral 605 is a reference numeral. An area of an obstacle by the video processing unit 104 is shown.

As shown in FIG. 5, in the case of bad weather with strong snow, the obstacle detection process by the video processing unit 104 is an image processing method if an obstacle is reflected in the video imaged by the camera 103. Thus, the obstacle area 504 can be specified.
On the other hand, in the obstacle detection processing by the ultrasonic signal processing unit 102, since the ultrasonic waves are irregularly reflected by snow, the intensity 502 of the reflected wave is weakened and does not exceed the determination threshold value 503. Therefore, the ultrasonic sensor 101 cannot detect an obstacle.

  In this case, it is possible to warn the driver if an obstacle can be detected by either means. However, if the vehicle moves under bad snowy weather and enters a dark place, the camera 103 can no longer detect the obstacle, which may cause no obstacle at all.

  As described above, when an obstacle is detected from the video processing unit 104 but no obstacle is detected from the ultrasonic signal processing unit 102, the parameter adjustment unit 109 updates the contents of the ultrasonic sensor parameter table 107. . As a result, the obstacle detection processing by the ultrasonic signal processing unit 102 is automatically adapted so that it can be performed even under adverse conditions.

  For example, as shown in FIG. 6, the parameter adjustment unit 109 lowers the value of the determination threshold value 603 in the ultrasonic sensor parameter table 107. Thereby, the sensor output value 602 by the ultrasonic sensor 101 can exceed the determination threshold value 603. Therefore, the ultrasonic signal processing unit 102 can identify the obstacle region 604.

As a result, as shown in FIG. 6, an image in which the obstacle region 604 by the ultrasonic signal processing unit 102 and the obstacle region 605 by the image processing unit 104 are highlighted on the image 601 captured by the camera 103 is displayed. Can be presented to the driver.
During the subsequent operation, by using the value of this processing parameter, an obstacle can be detected even when the vehicle moves under bad weather with snow and enters a dark place.

Next, parameter adjustment by the parameter adjustment unit 109 under conditions where obstacle detection by the video processing unit 104 is difficult will be described.
FIG. 7 is a diagram showing an example of an obstacle detection result before parameter adjustment when the illumination is dark according to Embodiment 1 of the present invention. FIG. 8 is a diagram showing an example of an obstacle detection result after parameter adjustment when the illumination is dark according to Embodiment 1 of the present invention.
In FIG. 7, reference numeral 701 denotes an image captured by the camera 103, reference numeral 702 denotes a sensor output value of the ultrasonic sensor 101, reference numeral 703 denotes a determination threshold value, and reference numeral 704 denotes an obstacle region by the ultrasonic signal processing unit 102. .
In FIG. 8, reference numeral 801 denotes an image captured by the camera 103, reference numeral 802 denotes an obstacle area by the ultrasonic signal processing unit 102, and reference numeral 803 denotes an obstacle area by the video processing unit 104.

As shown in FIG. 7, in the dark environment at dusk, since the ultrasonic wave is not affected in the obstacle detection processing by the ultrasonic signal processing unit 102, the ultrasonic wave is present when there is an obstacle. The sensor output value 702 from the sensor 101 exceeds the determination threshold value 703. Accordingly, the ultrasonic signal processing unit 102 can identify the obstacle region 704.
On the other hand, in the obstacle detection process by the video processing unit 104, the video captured by the camera 103 is dark and it is difficult to separate the obstacle and the background, so that the obstacle cannot be detected.

  In this case, it is possible to warn the driver if an obstacle can be detected by either means. However, when driving in this dark environment and encountering a situation where the ultrasonic sensor 101 does not operate correctly (such as a nearby parking lot or another ultrasonic sensor at the railroad crossing), the ultrasonic sensor 101 also fails. Since an object cannot be detected, an obstacle may not be detected at all.

  As described above, when an obstacle is detected from the ultrasonic signal processing unit 102 but no obstacle is detected from the video processing unit 104, the parameter adjustment unit 109 updates the contents of the camera parameter table 108. As a result, the obstacle detection process by the video processing unit 104 is automatically adapted so that it can be performed even under adverse conditions.

  For example, as shown in FIG. 8, the parameter adjustment unit 109 modifies parameters such as edge extraction processing so that the sensitivity of the camera 103 in the camera parameter table 108 is increased and an obstacle is easily detected. As a result, the video processing unit 104 can specify the obstacle region 803 from the video imaged by the camera 103.

As a result, as shown in FIG. 8, an image in which an obstacle region 802 by the ultrasonic signal processing unit 102 and an obstacle region 803 by the image processing unit 104 are highlighted on an image 801 photographed by the camera 103. Can be presented to the driver.
During the subsequent operation, the value of this parameter can be used to detect an obstacle even if it moves in a dark environment at dusk and snow falls.

  As described above, according to the first embodiment, the ultrasonic sensor 101, the camera 103, and the ultrasonic signal processing unit 102 and the video are based on the obstacle information by the ultrasonic signal processing unit 102 and the video processing unit 104. Since the parameter value of the processing unit 104 is configured to be automatically adjusted, the processing parameter adjustment is adaptively performed even in a condition where obstacle detection processing by the ultrasonic signal processing unit 102 or the video processing unit 104 is difficult. Therefore, it is possible to realize highly stable obstacle detection.

  In the parameter adjustment unit 109 according to the first embodiment, the case where the determination threshold is changed in the parameter adjustment in the ultrasonic sensor parameter table 107 has been described. However, the present invention is not limited to this, and the output of the ultrasonic sensor 101 is not limited thereto. You may perform operations such as raising the direction or changing the directivity.

Embodiment 2. FIG.
In the first embodiment, the parameter adjustment in the case where an obstacle is detected only in one of the obstacle detection processes has been described. However, in the second embodiment, the automatic adjustment is performed using information such as the detected area and number. A method for adjusting the processing parameters will be described.
The configuration of the vehicle surrounding obstacle detection device according to the second embodiment is the same as the configuration of the vehicle surrounding obstacle detection device according to the first embodiment shown in FIG. Give an explanation.

  The operation of the vehicle surrounding obstacle detection device according to the second embodiment is the same as the operation of the vehicle surrounding obstacle detection device according to the first embodiment shown in FIG. 3 except for the internal processing of the parameter adjustment unit 109. . Hereinafter, parameter adjustment by the parameter adjustment unit 109 will be described.

Here, the parameter adjustment unit 109 calculates an evaluation value D indicating the degree of overlap between the obstacle region by the ultrasonic signal processing unit 102 and the obstacle region by the video processing unit 104, and sets the parameter so that the evaluation value D increases. Automatically adjust the value. For example, the following formula (1) is used as the evaluation value D.
Evaluation value D = (Area (ultrasound) ∩Area (camera)) / (Area (ultrasound) ∪Area (camera)) + f (number of obstacle areas detected from the camera, obstacle areas detected by the ultrasound) Number) (1)

  In Expression (1), Area (X) represents a region on the image of the obstacle region specified by the obstacle detection processing unit X (the ultrasonic signal processing unit 102 or the video processing unit 104). In addition, (Area (ultrasonic wave) （Area (camera)) indicates an area of a portion where two regions overlap. (Area (ultrasonic wave) ∪Area (camera)) indicates an area of a region covered by two regions (a portion that does not overlap + a portion that overlaps). F is an evaluation function. For example, when there is a difference between the number of obstacle areas detected by the video processing unit 104 and the number of obstacle areas detected by the ultrasonic signal processing unit 102, the value is a small value. In some cases, a large value is indicated. Thus, the lower the coincidence of the obstacle detection results when the ultrasonic sensor 101 and the camera 103 are used, the smaller the evaluation value D is.

Next, parameter adjustment by the parameter adjustment unit 109 will be described with reference to specific drawings.
FIG. 9 is a diagram showing an example of an obstacle detection result before parameter adjustment according to Embodiment 2 of the present invention. FIG. 10 is a diagram showing an example of an obstacle detection result after parameter adjustment in the second embodiment of the present invention.
In FIG. 9, reference numeral 901 is an image captured by the camera 103, reference numeral 902 is a sensor output value of the ultrasonic sensor 101, reference numeral 903 is a determination threshold, reference numeral 904 is an area of an obstacle by the ultrasonic signal processing unit 102, and reference numeral 905 is An area of an obstacle by the video processing unit 104, reference numeral 906 indicates an overlapping area of the obstacle area by both means.
In FIG. 10, reference numeral 1001 is an image captured by the camera 103, reference numeral 1002 is a sensor output value of the ultrasonic sensor 101, reference numeral 1003 is a determination threshold, reference numeral 1004 is an area of an obstacle by the ultrasonic signal processing unit 102, and reference numeral 1005 is An obstacle area by the video processing unit 104, reference numeral 1006 indicates an overlapping area of the obstacle areas by both means.

In the case shown in FIG. 9, the obstacle detection processing by the ultrasonic signal processing unit 102 can detect only a part of obstacles (humans) due to the influence of the irregular reflection of ultrasonic waves by snow. Area 904 is specified smaller than the actual obstacle area.
On the other hand, in the obstacle detection by the camera 103, only a part of the obstacles (upper part of the obstacle) can be detected due to the influence of the dark illumination, and the obstacle area 905 is smaller than the actual obstacle area. Identified.

  In this case, since the two obstacle regions by the ultrasonic signal processing unit 102 and the video processing unit 104 are displaced, the area of the overlapping region 906 becomes small. Therefore, the evaluation value D calculated by the parameter adjustment unit 109 is small.

  Thus, when the evaluation value D is small, the parameter adjustment unit 109 updates the contents of the ultrasonic sensor parameter table 107 and the camera parameter table 108 so that the evaluation value D becomes large.

  For example, as illustrated in FIG. 10, the parameter adjustment unit 109 lowers the determination threshold 1003 in the ultrasonic sensor parameter table 107. Thereby, since the sensor output value 1002 by the ultrasonic sensor 101 can exceed the determination threshold value 1003, the ultrasonic signal processing unit 102 can specify the region 1004 that appropriately includes the obstacle.

  The parameter adjustment unit 109 corrects parameters such as edge extraction processing so that the sensitivity of the camera 103 in the camera parameter table 108 is increased and an obstacle is easily detected. Thereby, the image processing unit 104 can specify the region 1005 that appropriately includes the obstacle.

  In this case, since the displacement between the two obstacle regions by the ultrasonic signal processing unit 102 and the video processing unit 104 is reduced, the area of the overlapping region 1005 is increased. Therefore, the evaluation value D calculated by the parameter adjustment unit 109 increases.

  As described above, according to the second embodiment, the evaluation value D is defined using information such as the detected area and number, and the processing parameters are automatically adjusted so that this value increases. Since it is configured, it is possible to appropriately adjust parameters even under various adverse conditions.

Embodiment 3 FIG.
In the first and second embodiments, the processing parameters are automatically adjusted based on the obstacle information so that the obstacle can be detected as much as possible. However, in the third embodiment, the obstacle cannot be correctly detected ( When it is determined that the situation is difficult to detect), this is shown for what is presented to the driver.
The configuration of the vehicle surrounding obstacle detection device according to the third embodiment is the same as the configuration of the vehicle surrounding obstacle detection device according to the first embodiment shown in FIG. Give an explanation.

FIG. 11 is a diagram showing an example of an obstacle detection result in a difficult detection state according to Embodiment 3 of the present invention.
In FIG. 11, reference numeral 1101 denotes an image captured by the camera 103, reference numeral 1102 denotes a sensor output value of the ultrasonic sensor 101, reference numeral 1103 denotes a determination threshold, and reference numeral 1104 denotes a fence.

  For example, as shown in FIG. 11, when a fence 1104 exists as an obstacle when the vehicle is parked in the back, the emitted wave from the ultrasonic sensor 101 passes through the fence 1104, and therefore the intensity of the reflected wave is reduced. The sensor output values 1102 shown are all low values. For this reason, the sensor output value 1102 does not exceed the determination threshold 1103, and no obstacle is detected by the ultrasonic signal processing unit 102.

  Therefore, in addition to the obstacle detection processing function described in the first and second embodiments, the video processing unit 104 is difficult to perform obstacle detection processing by the ultrasonic signal processing unit 102 from the video captured by the camera 103. It has a function of detecting a specific pattern (fence or wire mesh) in image processing. A geometric continuous pattern composed of straight lines such as the fence 1104 can be easily detected by image processing, and can be realized relatively easily by examining whether the straight lines are arranged at equal intervals. When a specific pattern is detected in the video by the video processing unit 104, the fact is sent to the obstacle integrated detection unit 105.

  In addition to the image generation processing function described in the first and second embodiments, the obstacle integrated detection unit 105 detects a specific pattern that is difficult for the obstacle detection processing by the ultrasonic signal processing unit 102 from the image processing unit 104. When the notification is received, the driver is notified by a picture, character, voice, or the like via the display device 106 that the operation of the obstacle detection processing by the ultrasonic signal processing unit 102 is not compensated. It has a function.

  Similarly, the video processing unit 104 selects a specific pattern (with a small luminance difference in the image and an average luminance value within the standard range) that is difficult for the video processing unit 104 to perform obstacle detection processing from the video captured by the camera 103. ) Is detected. When a specific pattern is detected by the video processing unit 104, the fact is sent to the obstacle integrated detection unit 105.

  The obstacle integrated detection unit 105 performs an obstacle detection operation by the video processing unit 104 when the video processing unit 104 is notified that a specific pattern that is difficult to detect by the video processing unit 104 is detected. The driver is notified that the situation is not compensated by a picture, a character, a voice or the like via the display device 106.

  As described above, according to the third embodiment, when a situation is determined in which it is determined that the obstacle detection processing by the ultrasonic signal processing unit 102 or the video processing unit 104 does not operate properly, this is indicated to the driver. Therefore, in an environment where the operation of the obstacle detection process is not guaranteed, the driver can be alerted and parking can be safely supported.

Embodiment 4 FIG.
In the third embodiment, when a situation is determined in which obstacle detection processing by the ultrasonic signal processing unit 102 or the video processing unit 104 is determined not to operate properly, the driver is warned to that effect. In the fourth embodiment, the evaluation value D shown in the second embodiment is used to determine whether the obstacle detection process is operating properly.
The configuration of the vehicle peripheral obstacle detection device according to the fourth embodiment is the same as the configuration of the vehicle peripheral obstacle detection device according to the first embodiment shown in FIG. Give an explanation.

  The evaluation value D calculated by the parameter adjustment unit 109 is small when the area of the obstacle by the ultrasonic signal processing unit 102 or the video processing unit 104 does not match. Therefore, when the evaluation value D is small, there is a possibility that the accuracy of the obstacle detection process is lowered due to some environmental factor.

Therefore, in addition to the video generation processing function described in Embodiment 1-3, the parameter adjustment unit 109 uses the ultrasonic signal processing unit 102 and the video processing unit 104 when the evaluation value D is smaller than a certain threshold value. Of the obstacle detection processing, it has a function of determining that at least one means does not operate properly and notifying the obstacle integrated detection unit 105 to that effect.
The obstacle integrated detection unit 105 that has received the notification from the parameter adjustment unit 109 informs the driver that the operation of the obstacle detection process is not guaranteed with a picture, text, voice, or the like via the display device. Notice.

  As described above, according to the fourth embodiment, when the value of the evaluation value D is smaller than the threshold value, at least one of the obstacle detection processes by the ultrasonic signal processing unit 102 and the video processing unit 104 is performed. It is judged that the situation is not working properly, and the driver is warned to that effect, so in an environment where the operation of the obstacle detection process is not guaranteed, the driver's attention is urged to be safe. It is possible to assist parking.

  DESCRIPTION OF SYMBOLS 101 Ultrasonic sensor, 102 Ultrasonic signal processing part, 103 Camera, 104 Image processing part, 105 Obstacle integrated detection part, 106 Display apparatus, 107 Ultrasonic sensor parameter table, 108 Camera parameter table, 109 Parameter adjustment part

Claims (4)

With reference to the processing parameters described in the ultrasonic sensor parameter table, an ultrasonic sensor for sensing an obstacle with ultrasonic waves,
With reference to the processing parameters described in the ultrasonic sensor parameter table, an ultrasonic signal processing unit that detects information on an obstacle based on a sensor output value by the ultrasonic sensor;
With reference to the processing parameters described in the camera parameter table, a camera that captures images of obstacles,
With reference to the processing parameters described in the camera parameter table, a video processing unit that detects information on an obstacle based on video captured by the camera;
Wherein the obstacle integrating detector that match integrate the information of the detected obstacle by the information and the image processing section of the detected obstacle by ultrasonic signal processing unit,
A display device for displaying information of obstacles integrated by the obstacle integration detection unit;
On the basis of the information of the detected obstacle by ultrasonic signal processing unit and the video processing unit, and a parameter adjuster for adjusting the process parameters described in the ultrasonic sensor parameter table or the camera parameter table ,
The parameter adjustment unit calculates an evaluation value indicating a degree of overlap between the obstacle region detected by the ultrasonic signal processing unit and the obstacle region detected by the video processing unit, and sets the evaluation value. Based on this, the processing parameters described in the ultrasonic sensor parameter table and the camera parameter table are adjusted so that the degree of overlap increases.
A vehicle surrounding obstacle detection device.   With reference to the processing parameters described in the ultrasonic sensor parameter table, an ultrasonic sensor for sensing an obstacle with ultrasonic waves,
  With reference to the processing parameters described in the ultrasonic sensor parameter table, an ultrasonic signal processing unit that detects information on an obstacle based on a sensor output value by the ultrasonic sensor;
  With reference to the processing parameters described in the camera parameter table, a camera that captures images of obstacles,
  With reference to the processing parameters described in the camera parameter table, a video processing unit that detects information on an obstacle based on video captured by the camera;
  An obstacle integrated detection unit that integrates the obstacle information detected by the ultrasonic signal processing unit and the obstacle information detected by the video processing unit;
  A display device for displaying information of obstacles integrated by the obstacle integration detection unit;
  A parameter adjustment unit that adjusts a processing parameter described in the ultrasonic sensor parameter table or the camera parameter table based on information on an obstacle detected by the ultrasonic signal processing unit and the video processing unit. ,
  The obstacle integrated detection unit detects obstacle information by the ultrasonic signal processing unit when the video processing unit detects a specific pattern in which it is difficult to detect obstacle information by the ultrasonic signal processing unit. A vehicle surrounding obstacle detection device that notifies that the operation cannot be guaranteed through the display device.   With reference to the processing parameters described in the ultrasonic sensor parameter table, an ultrasonic sensor for sensing an obstacle with ultrasonic waves,
  With reference to the processing parameters described in the ultrasonic sensor parameter table, an ultrasonic signal processing unit that detects information on an obstacle based on a sensor output value by the ultrasonic sensor;
  With reference to the processing parameters described in the camera parameter table, a camera that captures images of obstacles,
  With reference to the processing parameters described in the camera parameter table, a video processing unit that detects information on an obstacle based on video captured by the camera;
  An obstacle integrated detection unit that integrates the obstacle information detected by the ultrasonic signal processing unit and the obstacle information detected by the video processing unit;
  A display device for displaying information of obstacles integrated by the obstacle integration detection unit;
  A parameter adjustment unit that adjusts a processing parameter described in the ultrasonic sensor parameter table or the camera parameter table based on information on an obstacle detected by the ultrasonic signal processing unit and the video processing unit. ,
  The obstacle integrated detection unit cannot guarantee an operation of obstacle information detection by the video processing unit when the video processing unit detects a specific pattern that makes it difficult for the video processing unit to detect obstacle information. A vehicle surrounding obstacle detection device characterized by notifying through the display device. The obstacle integrated detection unit detects an obstacle by the ultrasonic signal processing unit and the video processing unit when the degree of overlap is lower than a predetermined threshold based on the evaluation value calculated by the parameter adjustment unit. vehicle near obstacle detection apparatus to the effect that operation can not be guaranteed according to claim 1, wherein the notifying via the display device.

Images