KR101263975B1 - Method and System for Recongnizing Target Position for Parallel Parking - Google Patents

Abstract

The present invention relates to a target parking position recognition method and a system for parallel parking.

According to an embodiment of the present invention, a main reference cluster detection unit which receives range data and extracts a main reference cluster which is a range data bundle of a vehicle in front of a parking space; A search line extracting unit extracting a search line which is a linear component in a longitudinal direction of the vehicle in front of the parking target space by using the main reference cluster; A sub reference cluster detection unit which receives the search line and the range data and extracts a sub reference cluster which is a range data bundle of an obstacle behind the parking space; And a parking position setting unit configured to set a target parking position to park using the main reference cluster, the sub reference cluster, and the search line.

Auto Parking, Target Parking Location, Search Line

Description

Recognition Target Position for Parallel Parking and its System {Method and System for Recongnizing Target Position for Parallel Parking}

Embodiments of the present invention relate to a method and a system for recognizing target parking positions for parallel parking. More specifically, by installing a device for detecting the distance and direction of obstacles in the rear and rear of the vehicle, by using the detected range data automatically recognizes the target parking position in the parallel parking setting to set the exact target parking position for the parking assistance system The present invention relates to a target parking position recognition method for parallel parking and a system thereof.

In general, women and novice drivers have considerable difficulties when parking in parking lots and in some cases cause contact accidents. Thus, parking problems can put pressure on many driving inexperienced people and even avoid driving.

One of the ways to solve this problem is to install the sensor on the vehicle and automatically recognize the parking position so as to provide the driver with the convenience of parking. When the sensor is mounted on the vehicle to automatically recognize the parking position, an empty parking space is detected by mounting a device on the vehicle, such as a scanning laser radar, which detects range data indicating the distance and direction of the obstacle.

1 is a view illustrating a method for detecting an empty parking space when parallel parking by a conventional method.

As shown in FIG. 1, when the own vehicle equipped with the scanning laser radar in the rear and rear of the vehicle recognizes an empty parking space between the front vehicle and the rear vehicle, 'L extracted from the bundle of range data detected from the front vehicle (that is, the main reference cluster) An empty parking space is recognized using the main reference corner forming the 'shape' and the sub reference corner forming the 'L' shape extracted from the range data bundle (sub reference cluster) detected from the rear vehicle.

FIG. 2 is a diagram illustrating range data of a front vehicle and a rear vehicle that are detected when the host vehicle passes a parking target space.

As shown in FIG. 2, when the vehicle tries to recognize the parking space after passing the parking target space, which is an empty space where parking is available, when the main reference cluster is detected by detecting the main reference cluster from the preceding vehicle, the main reference corner is' It doesn't have an L 'shape.

In addition, even when the sub-reference cluster is detected from the rear car and the sub-reference corner is extracted, the distance between the host car and the rear car is far from that of the sub-reference corner.

Accordingly, when the main reference corner and the sub reference corner extracted in the situation of FIG. 2 are not 'L' shaped, the location of the empty parking space detected by the host vehicle may not be properly detected or may be detected as an incorrect location.

An embodiment of the present invention is equipped with a device for detecting the distance and direction of obstacles in the rear and rear of the vehicle to automatically recognize the target parking position in parallel parking by using the detected range data to determine the correct target parking position for the parking assistance system. Set it.

According to an embodiment of the present invention, a target parking position recognition system for parallel parking, comprising: a main reference cluster detector for receiving a range data and extracting a main reference cluster which is a range data bundle of a vehicle in front of a parking target space; A search line extracting unit extracting a search line which is a linear component in a longitudinal direction of the vehicle in front of the parking target space by using the main reference cluster; A sub reference cluster detection unit which receives the search line and the range data and extracts a sub reference cluster which is a range data bundle of an obstacle behind the parking space; And a parking position setting unit configured to set a target parking position using the main reference cluster, the sub reference cluster, and the search line.

The main reference cluster may be a cluster of range data at a position closest to the host vehicle.

The main reference cluster and the subreference cluster may be extracted in a region of interest indicating a predetermined width and a predetermined length in the host vehicle.

The search line may be a straight line connecting the random point having the smallest linear fitting error value with the main reference cluster start point by linearly fitting a main reference cluster start point and an arbitrary point in the main reference cluster.

In addition, the search line defines a random point as a corner point when a linear fitting error value exceeds a limit by linearly fitting a main reference cluster start point and an arbitrary point in the main reference cluster, and the main reference cluster. Excluding a point after the corner point, a linear fitting of one of the main reference cluster start point and a point in the main reference cluster connects the one point having the smallest linear fitting error value and the main reference cluster start point. It may be a straight line.

The main reference cluster detector may exclude the range data that is separated from the search line by a predetermined width or more from the main reference cluster.

The sub reference cluster may be set only among the range data existing at a distance within a predetermined sub extraction distance from the search line.

The target parking position may be set to be parallel to the search line in the space between the closest point of the main reference cluster and the closest point of the sub reference cluster.

According to another exemplary embodiment of the present invention, there is provided a method for recognizing a target parking position for parallel parking, the method comprising: extracting a main reference cluster which is a range data bundle of a vehicle in front of a parking target space by receiving range data; Extracting a search line which is a linear component in a longitudinal direction of the vehicle in front of the parking target space by using the main reference cluster; Receiving the search line and the range data and extracting a sub reference cluster which is a range data bundle of a rear obstacle of the parking space; And setting a target parking position using the main reference cluster, the sub reference cluster, and the search line.

Hereinafter, some embodiments of the present invention will be described in detail with reference to exemplary drawings. In adding reference numerals to the components of each drawing, it should be noted that the same reference numerals are used to refer to the same components as much as possible even if displayed on different drawings. In the following description of the embodiments of the present invention, a detailed description of known functions and configurations incorporated herein will be omitted when it may make the subject matter of the present invention rather unclear.

In describing the components of the embodiment of the present invention, terms such as first, second, A, B, (a), and (b) may be used. These terms are intended to distinguish the constituent elements from other constituent elements, and the terms do not limit the nature, order or order of the constituent elements. When a component is described as being "connected", "coupled", or "connected" to another component, the component may be directly connected to or connected to the other component, It should be understood that an element may be "connected," "coupled," or "connected."

3 is a view showing a target parking position recognition system for parallel parking according to an embodiment of the present invention.

As shown in FIG. 3, the target parking position recognition system for parallel parking according to an embodiment of the present invention includes a main reference cluster detector 304, a search line extractor 306, a sub reference cluster detector 308, and the like. Parking position setting unit 310 is included.

The main reference cluster detector 304 receives the range data from the scanning laser radar 302 and extracts a main reference cluster which is a bundle of range data of a vehicle in front of the parking space (hereinafter, referred to as a front vehicle). .

The search line extractor 306 extracts a search line that is a linear component in the longitudinal direction of the vehicle in front of the parking space using the extracted reference cluster.

The sub-reference cluster detection unit 308 extracts a sub-reference cluster which is a bundle of range data of the rear obstacle of the parking space using the range data received from the scanning laser radar 302 and the extracted search line. Here, the rear obstacle may be not only a vehicle but also other structures. In the present embodiment, the rear obstacle is regarded as a vehicle. Therefore, in the following description, the rear obstacle is defined as a rear car and described.

The parking position setting unit 310 sets the target parking position using the extracted main reference cluster, sub reference cluster, and search line.

The target parking position recognition system for parallel parking according to an embodiment of the present invention recognizes the target parking position in four steps:

Step 1: extract the main reference cluster,

Step 2: extract search lines,

Step 3: extract the sub-reference cluster,

Step 4: Set the target parking location.

As shown in FIG. 2, the present embodiment will be described using a case where a host vehicle and a parked vehicle are located in a parking lot.

In this embodiment, it is assumed that the scanning laser radar 302 generates range data while rotating counterclockwise.

The scanning laser radar 302 rotates horizontally on the ground, emits light of a predetermined wavelength, detects light reflected by an obstacle, and generates range data. The scanning laser radar 302 is generally mounted on the side rear of the vehicle and scans the laser light and acquires the distance and angle data pairs from the reflected light, and the distance and angle data pairs obtained at this time are called range data. .

The main reference cluster detector 304 extracts a main reference cluster from range data existing in a region of interest (ROI).

4 is a diagram illustrating an example of an ROI.

As shown in Fig. 4, the region of interest is from (−vehicle length) to (5 * vehicle length) in the longitudinal direction of the porcelain center around the scanning laser radar 302, and from 0 to (3 * vehicle width in the porcelain width direction. Means the area up to). That is, assuming that the position of the scanning laser radar is (0,0) coordinates, the forward direction of the host vehicle is negative in the longitudinal direction, the rear direction of the host vehicle is positive in the longitudinal direction, and the left side of the host vehicle forward direction is in the direction of the host vehicle width. If it is set to +, coordinates as shown in FIG. 4 may be generated. On the other hand, the size and coordinates of the region of interest may vary depending on the embodiment, the right side of the vehicle forward direction (meaning the left side on the screen) may also be the region of interest. Only the left side of the direction (the right side when viewed on the screen) will be described by way of example. In addition, unless otherwise indicated in the following descriptions, the expressions “left” and “right” refer to left and right sides as viewed based on the forward direction of the host vehicle.

In the present exemplary embodiment, the main reference cluster and the subreference cluster may be extracted in a region of interest indicating a predetermined width and a predetermined length in the host vehicle.

5 is a diagram illustrating a main reference cluster and a search line extracted for a vehicle in the ROI, and FIG. 6 is a diagram illustrating an example of extracting a data cluster.

The main reference cluster detector 304 detects the main reference cluster which is the range data bundle of the preceding vehicle from the range data existing in the ROI. The main reference cluster may be configured as a cluster of range data at the position closest to the host vehicle.

Here, the range data cluster refers to a bundle of range data collected while forming a predetermined length or more in the region of interest. As shown in FIG. 6, if the range data received from the scanning laser radar 302 differs from the previous received value by a predetermined size or more while generating the data cluster while receiving the range data, one cluster is finished. it means. Techniques for recognizing clusters from range data are described in detail in Korean Patent 10-2007-5554 and Korean Patent 10-2007-108342.

Subsequent range data with adjacent values is added to the currently recognized range data cluster. If range data having a value different from a previously received range data value by a predetermined size or more is received, the range data is different from the currently recognized cluster. Recognize it as a starting point for another data cluster, or as an isolated or invalid point that has nothing to do with the data cluster.

The data cluster closest to the host vehicle in the region of interest is extracted as the main reference cluster.

Since the scanning laser radar 302 rotates counterclockwise and the region of interest starts from the left side of the vehicle forward direction, the data cluster located on the left side of the vehicle forward direction becomes the main reference cluster.

When the main reference cluster is detected, the search line is extracted by the search line extractor 306. The search line is a side reference line for setting a target parking position of the own vehicle, and means a straight line component in the longitudinal direction of the vehicle in front of the parking space.

The search line extractor 306 may set a straight line having a minimum linear fitting error as a search line by performing a linear fitting on the main reference cluster when the main reference cluster is straight.

The search line is a straight line connecting the main reference cluster start point with the random point having the smallest linear fitting error value by linearly fitting the main reference cluster start point and an arbitrary point in the main reference cluster.

In the case of the straight fitting, an imaginary straight line connecting two points to an arbitrary point in the main reference cluster from the range data point at which the main reference cluster starts, and then an error value in which each point in the reference cluster is separated from the virtual straight line The sum of the sums is called a linear fitting error. For every arbitrary point in the main reference cluster, a linear fitting error value is obtained, and a straight line when the linear fitting error value is the smallest is set as a search line.

7 is a diagram illustrating an example of extracting a search line after finding a main reference corner.

When the main reference cluster is a 'L' shape having a corner that is not a straight line, the linear fitting error value exceeds the set limit by linearly fitting the start point of the main reference cluster and an arbitrary point in the main reference cluster. The point is defined as a corner point, and any one of the above points having the smallest linear fitting error value by linearly fitting any one of the start point of the main reference cluster and the point in the main reference cluster except the point after the corner point in the main reference cluster. The search line can be set to a straight line connecting the main reference cluster and the starting point.

If the main reference cluster has an 'L' shape having a corner that is not a straight line, the corner of the main reference cluster (that is, the main reference corner) is first found. In this case, the straight line of the main reference cluster is from the starting point of the main reference cluster to the corner point. The search line can be set by performing a linear fitting on the linear portion of the main reference cluster.

When the corner point of the main reference cluster is found, the corner point is defined as the point where the linear fitting error value falls outside the set error range while linearly fitting the main reference cluster to a point in an arbitrary main reference cluster.

Meanwhile, the process of finding the main reference cluster, the process of finding the corner points of the main reference cluster, and the process of setting the search line may be simultaneously performed.

At this time, the main reference cluster detector 304 excludes the range data that is separated by a predetermined width or more from the search line in the main reference cluster.

During the process of searching for the main reference cluster and setting up the search line, data that exceeds a certain distance (for example, 0.75 * self-width) from the search line among the range data is set to be excluded from the main reference cluster. In setting up the reference cluster, it can be set up at a higher speed.

When the main reference cluster is detected, the corner of the main reference cluster may be extracted by first checking whether the shape is 'L' shape through the corner fitting. For the 'L'-shaped test, the corner error of the main reference cluster is obtained by assuming that a random point in the main reference cluster is a corner. This is called corner fitting.

Assume that any point in the main reference cluster is a corner point, and each point in the main reference cluster assumes an 'L' shape with the line connecting the starting point of the main reference cluster and the end point of the main reference cluster around this corner point. Find the corner fitting error by adding up the sum of the distances away from the 'L' straight line.

If the corner fitting error is less than or equal to the set value, the cluster is regarded as a main reference cluster having an 'L' shape, and if it is greater than or equal to the set value, the cluster is regarded as a main reference cluster having a straight shape.

If the main reference cluster is 'L' shaped, the search line may be set from the corner to the longitudinal linear axis. If the main reference cluster is straight, the search line may be set to the search line.

If the vehicle has a rounded corner instead of an 'L' shaped corner, the corner point may be obtained by performing a corner fitting corresponding thereto.

As described above, the technique for obtaining the corner point from the data cluster is described in detail in Korean Patent 10-2007-108342, and thus, further description thereof will be omitted.

8 is a diagram illustrating an example of extracting a sub-reference cluster.

After obtaining the main reference cluster and the search line, we obtain the data cluster for the rear car, that is, the sub reference cluster.

The sub reference cluster provides information about the car behind. The sub reference cluster considers the first data cluster found after the main reference cluster extraction as the sub reference cluster. However, it is set only among the range data existing within the search line and a distance within a predetermined sub extraction distance (for example, 0.5 * own vehicle width). That is, the sub-reference cluster can be extracted without being affected by the adjacent side vehicle existing between the front vehicle and the rear vehicle by ignoring the data cluster outside the preset sub-extraction distance.

As shown in Fig. 8, the first cluster detected after the main reference cluster in the region of interest is a cluster by an adjacent vehicle (assuming that the scanning laser radar rotates counterclockwise). However, even if there is a cluster that detects an adjacent vehicle, it cannot be a sub-reference cluster because it is out of the sub-extraction distance from the search line.

Therefore, the cluster within the sub extraction distance from the search line becomes the sub reference cluster among the clusters detected by the rear vehicle.

In this way, the range data between the main reference cluster and the sub reference cluster can be ignored. This allows the extracted data cluster to extract the sub reference cluster effectively even in a complicated situation due to the large number of adjacent vehicles.

In this way, (0.75 * own vehicle width) set as the end point of the main reference cluster set to reduce the influence of noise caused by other adjacent vehicles or obstacles, and (0.5 * own vehicle width) which is the sub extraction distance set as the start point of the sub reference cluster. The criteria may be set differently for each embodiment.

FIG. 9 is a diagram illustrating an example of extracting a main reference cluster, a search line, and a sub reference cluster from a data cluster extracted in an actual situation.

As shown in FIG. 9, the sub-reference cluster can be recognized successfully despite the data cluster by other adjacent vehicles.

FIG. 10 is a diagram illustrating an end point of a main reference cluster and a nearest point of a sub reference cluster and a free space set therefrom.

When the main reference cluster and the sub reference cluster are extracted, the space between the end point of the main reference cluster (when the main reference cluster is linear) and the nearest point of the sub reference cluster is recognized as free space. Here, the closest point of the sub-reference cluster refers to range data that is projected close to the most main reference cluster when all range data in the sub-reference cluster is projected on the search line.

If the main reference cluster is 'L' shaped, the free space is recognized based on the closest point of the main reference cluster, and the closest point of the main reference cluster is when all range data in the main reference cluster is projected on the search line. It means the range data projected closest to the sub reference cluster.

As shown in Fig. 10, the space between the end point of the main reference cluster and the nearest point of the sub reference cluster is recognized as free space on the right side of the end point of the main reference cluster and the search line therefrom.

11 is a diagram illustrating an example of setting a free space and a target parking position extracted in an actual situation.

The parking position setting unit 310 calculates and sets the target parking position such that the center of the boundary line between the main reference cluster and the sub reference cluster (the center of the longitudinal direction of the free space of the free space) becomes the center of the target parking position and is parallel to the search line. do.

When setting the target parking position, the free space must be larger than the length of the own vehicle and allow enough space for parking.

As shown in Fig. 11, the target parking position is set so that the center thereof is the center of the free space, parallel to the search line, and the parking is possible before and after the target parking.

FIG. 12A illustrates a data cluster extracted when a vehicle is not properly parked and parked, especially when a rear car parks crookedly without aligning the parking in a line, and FIG. 12B illustrates a rear view of the vehicle in the situation of FIG. 12A. In the case where the extracted data cluster is not 'L', the target parking position is set.

As shown in FIGS. 12A and 12B, it is difficult to extract the 'L' data cluster of the rear car from the own car when the rear car parks the car diagonally without aligning the parking in a row, and in this case, the present invention is 'L' Because it does not recognize the target parking position by recognizing the ruler, the target parking position can be recognized successfully.

FIG. 13A is a diagram illustrating a data cluster extracted when an obstacle is complicated by an adjacent vehicle between a front vehicle and a rear vehicle, and FIG. 13B is a view illustrating a target parking position set in the situation of FIG. 13A. .

As shown in FIGS. 13A and 13B, when obstacles are complicated by adjacent vehicles between the front vehicle and the rear vehicle, the target parking position is likely to be set incorrectly due to the adjacent vehicle. In this case, the target parking position can be recognized successfully because the present invention ignores obstacles more than a certain distance from the longitudinal boundary of the front vehicle (= search line).

FIG. 14A is a photograph showing a situation in which a rear vehicle is parked obliquely, and FIG. 14B is a view illustrating a target parking position set in the situation of FIG. 14A.

As shown in Figs. 14A and 14B, when the rear car is parked at an angle, the conventional method uses the obliquely parked rear car to recognize the letter 'L' at an angle, so that the target parking position may be set incorrectly. In this case, the present invention can successfully recognize the target parking position because the present invention does not recognize the target parking position by recognizing the letter 'L'.

15 is a flowchart illustrating a method for recognizing a target parking position for parallel parking according to an embodiment of the present invention.

2 to 15 together, the target parking position recognition method for parallel parking according to an embodiment of the present invention includes a main reference cluster extraction step 1502, a search line extraction step 1504, and a sub reference cluster extraction step ( 1506 and a target parking position setting step 1508.

In the main reference cluster extraction step 1502, the range data is received to extract the main reference cluster which is the range data bundle of the vehicle in front of the parking space.

In the search line extraction step 1504, the search line, which is a straight line component in the longitudinal direction of the vehicle in front of the parking space, is extracted using the main reference cluster.

In the sub-reference cluster extraction step 1506, the search line and the range data are received to extract a sub-reference cluster which is a rear obstacle of the parking space, that is, the range data bundle of the rear vehicle.

In the target parking position setting step 1508, the target parking position is set using the main reference cluster, the sub reference cluster, and the search line.

As described above, according to an exemplary embodiment of the present invention, a parking assist system is provided by automatically detecting a target parking position in parallel parking by using a range data sensed by mounting a device detecting a distance and a direction of an obstacle on the rear and rear of a vehicle. It is effective to set the exact target parking position for the.

In particular, when the vehicle tries to recognize the parking space after passing the parking target space, which is an empty space where parking is available, the main reference corner has an 'L' shape when the main reference cluster is detected by detecting the main reference cluster from the front vehicle. Even if it does not appear, there is an effect of recognizing the target parking position.

In addition, when a predetermined sub-extraction distance is set from the search line, there is an effect of recognizing the target parking position without being affected even when there is an adjacent vehicle between the front car and the rear car of the parking target space.

In addition, even when the parking target space vehicles are aligned in a row and parked at an angle without parking, there is no need to recognize the 'L' data cluster, so that the target parking position can be recognized.

While the present invention has been described in connection with what is presently considered to be the most practical and preferred embodiments, it is to be understood that the invention is not limited to the disclosed embodiments. That is, all of the components may operate selectively in combination with one or more of them. In addition, although all of the components may be implemented in one independent hardware, each or all of the components may be selectively combined to perform some or all functions combined in one or a plurality of hardware. It may be implemented as a computer program having a. Codes and code segments constituting the computer program may be easily inferred by those skilled in the art. Such a computer program may be stored in a computer readable storage medium and read and executed by a computer, thereby implementing embodiments of the present invention. As the storage medium of the computer program, a magnetic recording medium, an optical recording medium, a carrier wave medium, or the like may be included.

In addition, the terms "comprise", "comprise", or "having" described above mean that the corresponding component may be embedded unless otherwise stated, and thus, other components. It should be construed that it may further include other components rather than to exclude them. All terms, including technical and scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs, unless otherwise defined. Commonly used terms, such as predefined terms, should be interpreted to be consistent with the contextual meanings of the related art, and are not to be construed as ideal or overly formal, unless expressly defined to the contrary.

The description above is merely illustrative of the technical idea of the present invention and various changes and modifications may be made by those skilled in the art without departing from the essential characteristics of the present invention. Therefore, the embodiments disclosed in the present invention are intended to illustrate rather than limit the scope of the present invention, and the scope of the technical idea of the present invention is not limited by these embodiments. The protection scope of the present invention should be interpreted by the following claims, and all technical ideas within the equivalent scope should be interpreted as being included in the scope of the present invention.

1 is a view illustrating a method for detecting an empty parking space when parallel parking by a conventional method.

FIG. 2 is a diagram illustrating range data of a front vehicle and a rear vehicle that are detected when the host vehicle passes a parking target space.

3 is a view showing a target parking position recognition system for parallel parking according to an embodiment of the present invention.

4 is a diagram illustrating an example of an ROI.

5 is a diagram illustrating a main reference cluster and a search line extracted for a vehicle in a region of interest;

6 is a diagram illustrating an example of extracting a data cluster.

7 is a diagram illustrating an example of extracting a search line after finding a main reference corner.

8 is a diagram illustrating an example of extracting a sub-reference cluster.

FIG. 9 is a diagram illustrating an example of extracting a main reference cluster, a search line, and a sub reference cluster from a data cluster extracted in an actual situation.

FIG. 10 is a diagram illustrating an end point of a main reference cluster and a nearest point of a sub reference cluster and a free space set therefrom.

11 is a diagram illustrating an example of setting a free space and a target parking position extracted in an actual situation.

FIG. 12A illustrates a data cluster extracted when a vehicle is not properly parked and parked, especially when a rear car parks crookedly without aligning the parking in a line, and FIG. 12B illustrates a rear view of the vehicle in the situation of FIG. 12A. In the case where the extracted data cluster is not 'L', the target parking position is set.

FIG. 13A is a diagram illustrating a data cluster extracted when an obstacle is complicated by an adjacent vehicle between a front vehicle and a rear vehicle, and FIG. 13B is a view illustrating a target parking position set in the situation of FIG. 13A. .

FIG. 14A is a photograph showing a situation in which a rear vehicle is parked obliquely, and FIG. 14B is a view illustrating a target parking position set in the situation of FIG. 14A.

15 is a flowchart illustrating a method for recognizing a target parking position for parallel parking according to an embodiment of the present invention.

Description of the Related Art

302: scanning laser radar 304: main reference cluster detection unit

306: search line extraction unit 308: sub-reference cluster detection unit

310: parking position setting unit

Claims (9)

In the target parking position recognition system for parallel parking, A main reference cluster detector for receiving the range data and extracting a main reference cluster which is a range data bundle of the vehicle in front of the parking space; A search line extracting unit extracting a search line which is a linear component in a longitudinal direction of the vehicle in front of the parking target space by using the main reference cluster; A sub reference cluster detection unit which receives the search line and the range data and extracts a sub reference cluster which is a range data bundle of an obstacle behind the parking space; And Parking position setting unit for setting a target parking position using the main reference cluster, the sub reference cluster and the search line , ≪ / RTI & The sub reference cluster, A target parking position recognition system for parallel parking, characterized in that only the range data existing within a distance within a predetermined sub- extraction distance from the search line. The method of claim 1, The main reference cluster, A target parking position recognition system for parallel parking, characterized in that it is a cluster of range data in the position closest to the own vehicle. The method of claim 1, The main reference cluster and the subreference cluster, Target parking position recognition system for parallel parking, characterized in that extracted in the region of interest showing a certain width and a predetermined length in the host vehicle. The method of claim 1, The search line, Target parking position recognition for parallel parking, characterized in that a straight line connecting the starting point and the main reference cluster starting point and the arbitrary reference point having the smallest linear fitting error value by the linear fitting between the main reference cluster starting point system. The method of claim 1, The search line, When a linear fitting error value exceeds a set limit by linearly fitting a main reference cluster starting point and an arbitrary point in the main reference cluster, the random point is defined as a corner point and a point after the corner point in the main reference cluster. Is a straight line connecting one of the points having the smallest linear fitting error value and the main reference cluster start point by linearly fitting one of the main reference cluster start point and a point in the main reference cluster except Target parking location recognition system for parking. The method of claim 1, The main reference cluster detection unit, And the range data spaced apart from the search line by a predetermined width is excluded from the main reference cluster. delete The method of claim 1, The target parking position is, And a target parking position recognition system for parallel parking, wherein the space between the closest point of the main reference cluster and the closest point of the sub-reference cluster is parallel to the search line. In the target parking position recognition method for parallel parking, Receiving the range data and extracting a main reference cluster which is a range data bundle of the vehicle in front of the parking space; Extracting a search line which is a linear component in a longitudinal direction of the vehicle in front of the parking target space by using the main reference cluster; Receiving the search line and the range data and extracting a sub-reference cluster which is a range data bundle of an obstacle behind the parking space; And And setting a target parking position using the main reference cluster, the sub reference cluster, and the search line. Extracting the sub reference cluster, The target parking position recognition method for parallel parking, characterized in that it is set only from the range data existing within the distance within the predetermined sub-extraction distance to the search line.

Images