JP2010079716A - Pedestrian detection device for vehicle - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a pedestrian detection device with high detection accuracy capable of detecting a pedestrian by an imaging means and detecting a shielding material existing between one's own vehicle and the pedestrian without increasing a control processing load. <P>SOLUTION: A characteristic point imaged by a camera for pedestrian detection is also used as a characteristic point of guardrail determination. Further, a corresponding point at which characteristic points imaged by respective cameras coincide is detected from the characteristic point, first and second search regions a and b are set in one portions of a first region A upper than the installation height of the guardrail and a lower second region B that can include the guardrail in an imaging region, and even though the guardrail is difficult to perform pattern matching to determine a difference in the number of corresponding points between the respective regions, it is possible to reduce a control processing load and a data amount to store. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a pedestrian detection device for a vehicle, and more particularly to pedestrian detection control for detecting a shield between a host vehicle and a pedestrian.

  Conventionally, in order to detect an obstacle such as a vehicle or a pedestrian existing around the host vehicle, for example, an imaging unit such as a CCD camera or a CMOS camera is provided, and the captured image is processed to be used for driving and running the vehicle. Techniques for obtaining are known.

  In Patent Document 1, an image around a vehicle imaged by an imaging unit mounted on a vehicle is input to a white line feature extraction unit, and a contour point of a guardrail parallel to the white line and the white line of a road is obtained. Detected as Further, a technique has been proposed in which a captured image is compressed after being input to an intermittent image acquisition unit, and feature data of a white line feature extraction unit and image data of the intermittent image acquisition unit are stored in time order.

  In Patent Document 1, since the contour point of the guardrail is detected as a polynomial, accurate road conditions can be recorded with a small capacity, and feature information of a moving object can be extracted from an image even when the background changes. Is possible.

  Patent Document 2 defines areas corresponding to a person's head-to-shoulder and shoulder-to-arm part from the image data around the host vehicle captured by the camera, and walks the object to be imaged in this area. A technique for recognizing a person as a person is proposed. In patent document 2, the detection precision of a pedestrian can be improved by detecting the area | region equivalent to the shoulder from the head where the feature of a pedestrian appears notably from the image of a target object.

JP-A-11-213137 JP 2004-303219 A

  Conventional pedestrian recognition, as in Patent Document 2, sets feature detection regions of the body such as the head and shoulders, and the degree of coincidence between the feature detected in each region and the feature for pedestrian recognition On the other hand, a stationary object such as a guardrail as a background is determined by using a standard template or by using another discriminant as in Patent Document 1. That is, since the pedestrian recognition and the stationary object such as the guard rail are separately performed for object recognition, the processing is complicated in terms of control.

  In particular, in pedestrian recognition of a vehicle, the degree of gazing at the pedestrian is completely different depending on whether the pedestrian is present in front of a shield such as a guardrail. For example, when there is a pedestrian in front of the host vehicle, if there is a guardrail between the host vehicle and the pedestrian, the degree of gazing at the occupant will be low, but if there is no shield, sufficient gazing at the occupant is necessary. is there.

  In other words, it is essential to identify and recognize a pedestrian and a stationary object in the background in order to perform pedestrian recognition. It is necessary to perform the pedestrian determination in a short time.

  An object of the present invention is to provide a pedestrian detection device with high detection accuracy without increasing the processing burden on control of detection of a pedestrian by an imaging means and detection of a shielding object between the host vehicle and the pedestrian. It is to be.

  The vehicle pedestrian detection device according to the present invention includes a plurality of imaging units mounted on the vehicle, and an image processing unit that performs image processing on a captured image captured by the imaging unit. The pedestrian is detected based on the above.

  According to the first aspect of the present invention, there is provided a pedestrian detection unit that detects a pedestrian based on imaging data obtained from an image captured by the first imaging unit, and imaging data of the pedestrian detected by the pedestrian detection unit. Search area setting means for setting an area for searching for a feature point characterizing a pedestrian, and a feature point characterizing a pedestrian in the pedestrian image data obtained from an image captured by a second imaging means, Based on the corresponding points detected by the corresponding point detecting means, the corresponding point detecting means for detecting the characteristic points matching the characteristic points of the pedestrian's imaging data in the area as the corresponding points in the search area. It has the shielding object determination means which determines the shielding object between a vehicle and the said pedestrian.

  In the first aspect of the invention, in order to determine the shielding object between the host vehicle and the pedestrian based on the corresponding points for detecting the pedestrian, it is necessary to separately extract a special feature point for determining the shielding object. There is no need to detect the distance of the shield alone.

  According to a second aspect of the present invention, in the first aspect of the invention, the pedestrian detection means divides the detected pedestrian imaging data into a head region, a torso region, and a leg region. Detection is performed by pattern matching of regions.

  According to a third aspect of the present invention, in the first or second aspect of the invention, the search area setting means is set based on the detected pedestrian imaging data, and the first area above a predetermined height and below the predetermined height. The first and second search areas are set in a part of each of the second areas.

  According to a fourth aspect of the present invention, in the third aspect of the present invention, the shielding object determining means detects a shielding object between the host vehicle and the pedestrian by relative comparison of the number of corresponding points detected in the first and second search areas. It is characterized by determining.

  According to a fifth aspect of the present invention, in the third aspect of the present invention, the shielding object judging means detects the shielding object between the host vehicle and the pedestrian by comparing the distances of corresponding points in the first and second search areas. It is characterized by determining.

  According to a sixth aspect of the present invention, in the first or second aspect of the invention, the search area setting means sets the search area corresponding to an installation height of a guardrail, and the shielding object determining means is the correspondence inspection. When the number of departures is less than or equal to a predetermined value, the shield between the host vehicle and the pedestrian is determined as a guardrail.

  According to a seventh aspect of the present invention, in the sixth aspect of the present invention, the shielding object judging means is configured to compare the own vehicle and the pedestrian by relative comparison of the detected number of corresponding points in the search area higher than the guardrail installation height and the search area including the guardrail. It is characterized by determining a guardrail between.

  The invention of claim 8 is characterized in that, in the invention of claim 1 or 2, when the pedestrian detection means detects a pedestrian in front of the host vehicle, the pedestrian is continuously imaged.

  According to a ninth aspect of the present invention, in the eighth aspect of the invention, the shielding object judging unit is configured to compare the vehicle and the walking with a relative comparison between a corresponding point detection rate at a predetermined time and a corresponding point detection rate after a predetermined time has elapsed from the predetermined time. It is characterized by determining a blocking object between persons.

  According to the first aspect of the present invention, detection of a pedestrian by an imaging unit and detection of a shielding object between the host vehicle and the pedestrian can be executed without increasing the control processing load, and high detection accuracy is obtained. be able to. That is, since the feature point for detecting the pedestrian is also used as the feature point for the obstacle determination, the processing time for the obstacle determination can be shortened. Moreover, since the relative position between the pedestrian and the shielding object is detected using the corresponding points detected by the first and second imaging means for detecting the pedestrian position, the distance of the shielding object alone is not detected. The presence or absence of an obstacle between the host vehicle and the pedestrian can be determined.

  According to the invention of claim 2, in order to categorize the detected pedestrian imaging data into a head region, a torso region, and a leg region, and to detect a pedestrian by pattern matching of each segmented region, Pedestrians can be detected with high accuracy and ease.

  According to the third aspect of the present invention, the first and second portions of the first region above the predetermined height and the second region below the predetermined height set based on the detected pedestrian image data are first, Since the second search area is set, the control processing load and the amount of data to be accumulated can be reduced.

  According to invention of Claim 4, in order to determine the obstruction | occlusion between the own vehicle and the said pedestrian by the relative comparison of the corresponding point detection number in the 1st and 2nd search area | region, the processing load on control and the data to accumulate | store The relative positional relationship between the pedestrian and the shielding object can be detected while reducing the amount.

  According to the fifth aspect of the present invention, the control processing burden and the accumulated data are determined in order to determine the shielding object between the own vehicle and the pedestrian by the relative comparison of the distance between the corresponding points in the first and second search areas. The relative position between the pedestrian and the shielding object can be accurately detected while reducing the amount.

  According to invention of Claim 6, while setting said search area | region according to the installation height of a guardrail, the said obstruction determination means is the said vehicle and the said walk, when the said corresponding point detection number is below a predetermined value. Since the obstacle between the pedestrians is determined to be a guard rail, the relative positional relationship between the pedestrian and the guard rail can be detected while reducing the control processing load and the amount of data to be accumulated even if the guard rail is difficult to pattern match.

  According to the invention of claim 7, since the guard rail between the host vehicle and the pedestrian is determined by relative comparison of the number of corresponding points detected in the search area higher than the guard rail installation height and the search area including the guard rail, Can be detected with high accuracy.

  According to invention of Claim 8, when a pedestrian is detected ahead of the own vehicle, since the said pedestrian is imaged continuously, the presence or absence of the obstruction | occlusion between the own vehicle and a pedestrian can be determined accurately.

  According to the ninth aspect of the present invention, the shield between the host vehicle and the pedestrian is determined by a relative comparison between the corresponding point detection rate at a predetermined time and the corresponding point detection rate after a predetermined time has elapsed from the predetermined time. Can be determined, and the shielding object can be determined with high accuracy.

Hereinafter, the best mode for carrying out the present invention will be described with reference to the drawings.
The vehicle pedestrian detection device according to the present invention includes a plurality of imaging units mounted on the vehicle, and an image processing unit that performs image processing on a captured image captured by the imaging unit. The pedestrian is detected based on the above.

  A first embodiment will be described with reference to FIGS. FIG. 1 is a layout diagram of each component of the pedestrian detection device 1 according to the present embodiment. As shown in FIG. 1, the pedestrian detection device 1 includes a first CCD camera 2 as a first imaging unit that images a front space around the host vehicle V, and a second imaging unit that images a front space around the host vehicle V. As a second CCD camera 3.

  The first CCD camera 2 is arranged on the right side of the room mirror in the front part of the vehicle compartment, and the second CCD camera 3 is arranged on the left side of the room mirror in the front part of the vehicle compartment. These CCD cameras 2 and 3 image the surroundings of the host vehicle V, and extract and specify obstacles such as pedestrians that should avoid contact. The first and second imaging means may be image sensors such as CMOS cameras, and are not limited to CCD cameras.

  The pedestrian detection device 1 has an alarm device 4 arranged on an instrument panel in front of the driver's seat. This alarm device 4 displays and alerts obstacles such as pedestrians around the host vehicle V detected by the CCD cameras 2 and 3 and visually or audibly informs the occupant to avoid contact with the obstacles. Raise an alarm. Specifically, equipment such as a speaker, a lamp, a buzzer, and a display corresponds to the alarm device 4.

  FIG. 2 shows a control system diagram centering on the control unit 5 of the pedestrian detection apparatus 1. The control unit 5 includes an image processing unit 6 (image processing unit), a pedestrian detection unit 7 (pedestrian detection unit), a pedestrian detection matching database 8, a pedestrian region setting unit 9, an intermediate region setting unit 10, and a search region. The setting unit 11 (search area setting unit), the corresponding point detection unit 12 (corresponding point detection unit), and the shielding object determination unit 13 (shielding object determination unit).

  The image processing unit 6 performs image processing based on the captured image signals from the CCD cameras 2 and 3 and extracts image data related to the contour shape, the feature portion, and the like of the imaging target. Image data picked up by the CCD cameras 2 and 3 is stored in a first storage unit 6a and a second storage unit 6b provided for each camera after image processing.

  The pedestrian detection unit 7 compares the image data of the first CCD camera 2 obtained from the first storage unit 6a with the pedestrian detection data stored in the pedestrian detection matching database 8 to determine whether the imaging target is a pedestrian. Determine whether or not. Specifically, using template data that divides the image data into a head region corresponding to the pedestrian's head position, a torso region corresponding to the torso position, and a leg region corresponding to the leg position, Pattern matching of each divided area is performed in order from the head, and when the matching rate is a predetermined value or more, it is recognized that the imaging target is a pedestrian. In addition, when the pedestrian detection unit 7 recognizes a pedestrian, the captured image of the pedestrian is continuously processed until the pedestrian deviates from the front imaging area, and the subsequent determination is performed. Yes.

  The pedestrian detection matching database 8 includes template data for pattern matching divided into a pedestrian's head region, torso region, and leg region. The database 8 includes head template data having head characteristics for each age and gender, torso template data in which a degree of association with each head template data is set, and to each torso template data. Leg template data to which the degree of association is set is stored as three levels of hierarchical data. When predetermined head template data is selected, individual trunk template data is selected and collated in descending order of the degree of association. Leg template data is similarly selected and pedestrian detection is performed.

  As shown in FIGS. 3a and 3b, the pedestrian region setting unit 9 is a substantially rectangular shape including a pedestrian's head region, torso region, and leg region detected by the pedestrian detection unit 7 on the image data. A pedestrian area X having a shape is set.

  The middle region setting unit 10 divides the pedestrian region X set by the pedestrian region setting unit 9 into a first region A corresponding to the upper body and a second region B corresponding to the lower body. In the present embodiment, the boundary between the first area A and the second area B is set at a position near the upper end of the guard rail G, for example, a height of 1 m from the road surface, and when the guard rail G is present, It is set so that the guardrail G occupies approximately half of the area.

  The search area setting unit 11 sets a first search area a having a predetermined area in the first area A and a second search area b having the same area as the first search area a in the second area B. The first search area a is configured to correspond to the chest area below the pedestrian's neck, and the second search area b corresponds to the front area corresponding to the side surface of the guardrail G when the guardrail G is present.

  The corresponding point detection unit 12 is a feature point in the first and second search areas a and b obtained from the image taken by the first CCD camera 2 and a feature point obtained from the image taken by the second CCD camera 3. Thus, feature points that coincide with the feature points of the first CCD camera 2 in the search area are detected as corresponding points in the first and second search areas a and b.

Based on FIG. 4, the process in the corresponding point detection part 12 is demonstrated. The upper left figure is the first search area a (aR) imaged by the first CCD camera 2, the upper right figure is the first search area a (aL) imaged by the second CCD camera 3, and the lower left figure is the first CCD camera 2. The second search area b (bR) imaged in FIG. 4B, and the lower right figure shows the second search area b (bL) imaged by the second CCD camera 3.
In the first search area a corresponding to the upper body, a feature point Pa that characterizes a pedestrian among pedestrians captured by the first CCD camera 2, for example, a heel of clothes is recognized. Note that TaR is an area where the first CCD camera 2 can process an image at a time, for example, a 4-bit image processing area.

  As shown in the upper right diagram, the second CCD camera 3 scans the processing area TaL, moves until it captures the feature point Pa, and detects the corresponding point (feature point) Pa when captured. The CCD cameras 2 and 3 are configured so as to be automatically focused, and since the installation positions of both the CCD cameras 2 and 3 are set in advance, up to the corresponding point Pa in addition to the vertical and horizontal positions on the image data. Can be calculated. Similarly, in the second search region b, the feature point Pb is captured by scanning the feature region Pb of the pedestrian captured by the first CCD camera 2 in the processing region TbL of the second CCD camera 3, and the corresponding point Pb is obtained. To detect.

  The shielding object determination unit 13 calculates the number of feature points Pa detected in the first search region a and the number of feature points Pb detected in the second search region b, and the difference between the feature point Pa and the feature point Pb. A relative comparison is performed. When this difference is larger than a predetermined value, it is determined that the guardrail G exists between the pedestrian and the host vehicle V, and no operation command is issued to the warning means 4. On the other hand, when the difference is equal to or smaller than the predetermined value, it is determined that the guardrail G is not present between the pedestrian and the host vehicle V, and an operation command is issued to the warning means 4.

  The chest region below the pedestrian's neck that is the first search area a has physical features such as the neck and shoulders, and clothing patterns and folds. As a result, a large number of corresponding points are detected. On the other hand, in the second search area b, similarly, there are legs, clothing patterns, wrinkles, etc., and if there is no shielding object, a large number of feature points Pb are detected. However, when there is a shield, in particular, the guardrail G, the surface is simple in structure, and since there is no pattern, the feature point Pb is not detected, and as a result, the corresponding point Pb is not detected. . The guardrail G is detected using the detection tendency of corresponding points in the guardrail G.

Next, control of the pedestrian detection apparatus 1 will be described based on the flowchart of FIG. Si (i = 1, 2,...) Indicates each processing step.
First, an image around the host vehicle V is picked up by the first and second CCD cameras 2 and 3, stored in the first and second storage units 6a and 6b, and image processing is performed (S1). The storage units 6a and 6b store image data for a predetermined time, for example, about 10 seconds. Next, the pattern matching template and the image data of the first CCD camera 2 are collated to calculate the matching rate of the characteristic portion (S2).

  Based on the calculation of S2, if it is determined that there is a pedestrian as a result of the determination of S3, the setting of the pedestrian area X (S4), the setting of the first area A and the second area B (S5), the first The search area a and the second search area b are set (S6), and the process proceeds to S7. In S7, the number of corresponding points in each of the first search area a and the second search area b is calculated, and the process proceeds to the next. If the result of determination in S3 is No, the process returns. If it is determined in S3 that there is a pedestrian, the pedestrian is continuously captured until the pedestrian moves out of the imaging area of the CCD cameras 2 and 3.

  If the difference between the number of corresponding points detected in the first search area a and the number of corresponding points detected in the second search area b is greater than a predetermined value as a result of the determination in S8, the distance between the pedestrian and the host vehicle V It determines with the guardrail G existing (S9), and returns. If the difference between the number of corresponding points detected in the first search area a and the number of corresponding points detected in the second search area b is equal to or smaller than a predetermined value as a result of the determination in S8, a guardrail is provided between the pedestrian and the host vehicle V. It determines with G not existing (S10), operates the alarm device 4 (S11), and returns.

  In this embodiment, the determination result at the present time and the determination result before a predetermined time are used for the determination of the guardrail G in S8. In other words, when the shielding object is recognized by the determination before the predetermined time, and when the shielding object is recognized even at the present time, it is determined that the shielding object extends in the horizontal direction. One.

Next, functions and effects of the first embodiment will be described.
Since the feature points for detecting pedestrians are also used as the feature points for determining the guardrail G (shielding object), the processing time for determining the shielding object can be shortened. Moreover, it is set based on the detected imaging data of the pedestrian, and is included in each of a first area A above the guardrail installation height and a second area B below that can include the guardrail in the imaging area. Since the first and second search areas a and b are set and the difference in the number of corresponding points in each area is determined, even if the guard rail G is difficult to match the pattern, the control processing load and the amount of data to be accumulated can be reduced. Can be reduced.

  In addition, the detected pedestrian imaging data is divided into a head region, a torso region, and a leg region, and pedestrians are detected by pattern matching of the respective divided regions. Can be detected. Furthermore, when a pedestrian is detected in front of the host vehicle V, since this pedestrian is continuously imaged, the presence or absence of the guard rail G between the host vehicle V and the pedestrian can be accurately determined.

A second embodiment will be described based on the flowchart of FIG.
The difference from the first embodiment is that, in the first embodiment, the presence / absence of the guardrail G as a shield is determined by the difference between the number of corresponding points detected in the first search area a and the number of corresponding points detected in the second search area b. In contrast, in the second embodiment, the presence / absence of the guardrail G as a shield is determined based on the difference between the distance to the first search area a and the distance to the second search area b.

  First, images around the host vehicle V are picked up by the first and second CCD cameras 2 and 3, stored in the first and second storage units 6 a and 6 b, and image processing is performed (S 21). Next, the pattern matching template and the image data of the first CCD camera 2 are collated to calculate the matching rate of the characteristic portion (S22).

  Based on the calculation of S22, when it is determined that there is a pedestrian as a result of the determination of S23, the setting of the pedestrian area X (S24), the setting of the first area A and the second area B (S25), and the first The search area a and the second search area b are set (S26), and the process proceeds to S27. In S27, the number of corresponding points in each of the first search area a and the second search area b is calculated, and the process proceeds to the next. If the result of determination in S23 is No, the process returns.

  The shielding object determination unit 13 calculates the distance to each corresponding point detected in the first search area a, and calculates the average value of the distance to each corresponding point to obtain the distance to the first search area a. Calculate (S28). Similarly, the distance to the second search area b is calculated (S29). If the difference between the distance to the first search area a and the distance to the second search area b is greater than a predetermined value as a result of the determination in S30, it is determined that the guardrail G exists between the pedestrian and the host vehicle V. (S31) and return.

  If the difference between the distance to the first search area a and the distance to the second search area b is equal to or smaller than a predetermined value as a result of the determination in S30, it is determined that there is no guardrail G between the pedestrian and the host vehicle V. Then, the alarm device 4 is actuated (S33), and the process returns. As in the first embodiment, the determination result at the present time and the determination result before a predetermined time are used for the determination of the guardrail G in S30.

Next, functions and effects of the second embodiment will be described.
Basically, the feature point for detecting the pedestrian is also used as the feature point for determining the guardrail G (shielding object), so that the processing time for determining the guardrail G can be shortened. Moreover, since the guard rail G between the host vehicle and the pedestrian is determined by relative comparison of the distances between corresponding points in the first and second search areas a and b, the control processing load and the amount of accumulated data are reduced. The relative position between the pedestrian and the guardrail G can be detected with high accuracy.

Next, a modification in which the above embodiment is partially changed will be described.
1) In the above-described embodiment, an example has been described in which the determination of the guardrail is based on the number of corresponding points at the present time, or the determination result of the distance and the determination result before a predetermined time, but only the current determination result is used. It is also possible to determine the guardrail.

2] In the above embodiment, an example has been described in which an alarm is activated when the guardrail G is not present between the pedestrian and the host vehicle V. However, the present invention is not limited to the alarm, and may be applied to other driving support control. Is possible.

  In addition, those skilled in the art can implement the present invention by adding various modifications without departing from the spirit of the present invention, and the present invention includes such modifications.

It is a layout figure of each component of the pedestrian detection device concerning Example 1 of the present invention. It is a control system figure of the pedestrian detection device concerning this example 1. It is explanatory drawing of the pedestrian recognition which concerns on the present Example 1, Comprising: The case where a guardrail exists in front of a pedestrian is shown. It is explanatory drawing of the pedestrian recognition which concerns on the present Example 1, Comprising: The case where a guardrail is not in front of a pedestrian is shown. It is explanatory drawing of the corresponding point search which concerns on the present Example 1. FIG. 3 is a flowchart of pedestrian detection processing according to the first embodiment. 12 is a flowchart of pedestrian detection processing according to the second embodiment.

Explanation of symbols

V vehicle (own vehicle)
DESCRIPTION OF SYMBOLS 1 Pedestrian detection apparatus 2 1st CCD camera 3 2nd CCD camera 4 Alarm apparatus 5 Control unit 6 Image processing part 7 Pedestrian detection part 8 Pedestrian detection matching database 9 Pedestrian area setting part 10 Intermediate area setting part 11 Search area setting Part 12 Corresponding point detection part 13 Shielding object determination part A, B 1st, 2nd area | region a, b 1st, 2nd search area Pa, Pb Characteristic point (corresponding point)

Claims (9)

A pedestrian of a vehicle having a plurality of imaging means mounted on the vehicle and an image processing means for performing image processing on a captured image captured by the imaging means, and detecting a pedestrian based on the image-processed imaging data In the detection device,
Pedestrian detection means for detecting a pedestrian based on imaging data obtained from an image captured by the first imaging means;
A search area setting means for setting an area for searching for a feature point characterizing a pedestrian in the image data of the pedestrian detected by the pedestrian detection means;
A feature point characterizing a pedestrian in the pedestrian imaging data obtained from an image captured by the second imaging means, wherein the feature point coincides with a feature point of the pedestrian imaging data in the search region. Corresponding point detecting means for detecting as corresponding points in the search area;
A pedestrian detection apparatus for a vehicle, comprising: an obstacle determination unit that determines an obstacle between the host vehicle and the pedestrian based on the corresponding point detected by the corresponding point detection unit.   The pedestrian detection means divides the detected pedestrian imaging data into a head region, a torso region, and a leg region, and detects them by pattern matching of the respective divided regions. The pedestrian detection device for a vehicle according to 1.   The search area setting means includes first and second parts of a first area above a predetermined height and a second area below a predetermined height set based on the detected pedestrian imaging data. The pedestrian detection device for a vehicle according to claim 1, wherein a search area is set.   4. The vehicle according to claim 3, wherein the shielding object determination unit determines the shielding object between the host vehicle and the pedestrian by a relative comparison of the number of detected corresponding points in the first and second search areas. Pedestrian detection device.   4. The vehicle according to claim 3, wherein the shielding object determination unit determines an obstacle between the own vehicle and the pedestrian by relative comparison of distances between corresponding points in the first and second search areas. Pedestrian detection device.   The search area setting means sets the search area corresponding to the installation height of the guardrail, and the shielding object determination means determines whether the corresponding point detection number is less than or equal to a predetermined value between the host vehicle and the pedestrian. The pedestrian detection device for a vehicle according to claim 1, wherein the shield is determined as a guardrail.   The said shielding object determination means determines the guardrail between the own vehicle and the said pedestrian by the relative comparison of the corresponding point detection number in the search area higher than the installation height of a guardrail and the search area including a guardrail. The pedestrian detection device for a vehicle according to claim 1.   3. The vehicle pedestrian detection device according to claim 1, wherein when the pedestrian detection unit detects a pedestrian in front of the host vehicle, the pedestrian detection unit continuously images the pedestrian. 4. The shielding object determining means determines a shielding object between the host vehicle and the pedestrian by relative comparison between a corresponding point detection rate at a predetermined time point and a corresponding point detection rate after a predetermined time has elapsed from the predetermined time point. Item 9. The vehicle pedestrian detection device according to Item 8.