Target detection system and method for a vehicle

Abstract

In a case where a person departs from the vicinity of a vehicle in a longitudinal direction, a condition for generating a warning is satisfied due to a change in speed in a lateral direction. Thus, the warning system generates a false warning. To solve this problem, a target detection system for a vehicle and a target detection method for a vehicle are proposed, both of which are capable of calculating a final danger level that takes into account not only a result of recalculation of a collision time and a collision point but also presence or absence of a target detected by a camera sensor. The collision time and the collision point are recalculated taking into account a change in speed in the lateral direction when the target moves in the longitudinal direction.

Description

[0001] Cross-references to related applications

[0002] This application claims priority to Korean Patent Application No. 10-2021-0139451, filed on October 19, 2021, the entire contents of which are incorporated herein by reference for all purposes. Technical Field

[0003] This disclosure relates to a target detection system and a target detection method for a vehicle, and more particularly to a target detection system and a target detection method for a vehicle, wherein the system and the method are capable of improving false warning situations caused by speed errors in the lateral direction of the vehicle. Background Technology

[0004] A vehicle system for target detection refers to a system that detects targets approaching the blind spot of a user's vehicle and provides advance notification of the target's approach. As the detection unit of a target detection system, radar sensors that utilize reflected waves can be used, among other things.

[0005] The operating principle of radar used in vehicles is to transmit electromagnetic wave signals from the user's vehicle, receive electromagnetic wave signals reflected from different vehicles or obstacles, and use the time difference between these two signals and the change in Doppler frequency to estimate the distance or relative speed between the radar and different vehicles or obstacles.

[0006] However, there may be issues that could lead to false warnings due to incorrect detection caused by factors such as the type of obstacle or the direction of movement.

[0007] For example, when a person walks, various micro-Doppler effects occur due to the movements of their head, shoulders, upper body, arms, and legs. That is, the micro-Doppler effect refers to the additional Doppler frequency shift generated by vibrations and rotations in different parts of the person, in addition to the platform. The micro-Doppler effect is used when classifying target types. However, the micro-Doppler effect introduces errors when estimating target motion information. This is because obtaining multiple detection data from a single target leads to differences in the accuracy of distance, velocity, etc., depending on which of the multiple detection data is used as the representative information. Furthermore, because targets identified as people have a smaller radar cross section (RCS) than vehicles, the signal received after being reflected by the target has a smaller amplitude. Therefore, this may further increase information errors.

[0008] Therefore, in a case where a person exits from a side of a vehicle in a longitudinal direction, since a distance in a lateral direction between the vehicle and the person is short, in some cases, a false warning is generated. More specifically, in a case where a person gets off a vehicle parked near a user's vehicle and moves away from the user's vehicle in a longitudinal direction, since a distance in a lateral direction between the vehicle and the person is short and a micro-Doppler effect occurs as the person moves, a speed error in the lateral direction can occur. In this case, since a condition for generating a false alarm is satisfied due to the speed error in the lateral direction, there is a problem that the warning system generates a false alarm.

[0009] Therefore, there is an increasing need for technical development to provide a method for improving a false warning situation that occurs due to a change in speed in a lateral direction when a target moves in a longitudinal direction.

[0010] The foregoing is intended merely to aid in understanding the background of the present disclosure and therefore should not be interpreted as acknowledging the related art that is known to those of ordinary skill in the art. SUMMARY

[0011] An object of the present disclosure proposed to solve the above-described problems is to provide a target detection system for a vehicle and a target detection method for a vehicle, both of which are capable of calculating a final danger level, considering not only a result of recalculation of a time-to-collision and a collision point but also presence or absence of a target detected by a camera sensor 300. The time-to-collision and the collision point are recalculated considering a change in speed in a lateral direction that occurs when a target moves in the lateral direction.

[0012] To achieve the above object, according to one aspect of the present disclosure, there is provided a target detection system for a vehicle, the system including a detection sensor that detects a target based on a reflected wave, a camera sensor that monitors the target based on an image, and a controller that determines a first danger level of the target based on data acquired through the detection sensor and a change in speed of the target in a lateral direction, determines a second danger level of the target based on data acquired through the camera sensor, and calculates a final danger level based on the first danger level and the second danger level.

[0013] In the system, in a case where the target exists within a threshold distance in the lateral direction, the controller can determine the first danger level based on the change in speed of the target in the lateral direction.

[0014] In the system, the controller can calculate a collision time and a collision point as factors for determining the first danger level, and in the case where a change in velocity in the lateral direction occurs, the controller can recalculate the collision time and the collision point by adding the change in velocity in the lateral direction to the collision time and the collision point.

[0015] In the system, the controller determines the first danger level as indicating "danger" or "safe", and in the case where the recalculated collision time is shorter than a threshold value and the recalculated impact point falls within a threshold range, the controller can determine the first danger level as indicating "danger".

[0016] In the system, the controller can determine a second danger level as indicating "danger" or "safe", and can extract a left side and a right side of an image acquired by monitoring by the camera sensor as a warning area, and in the case where the target exists on the warning area, the controller can determine the second danger level as indicating "danger". In the system, the controller can calculate a value of an image change of the warning area, and when the value of the image change is equal to or higher than a threshold value, the controller can determine the second danger level as indicating "danger".

[0017] In the system, the controller can determine a first danger level and a second danger level as indicating "danger" or "safe", in the case where both the first danger level and the second danger level indicate danger, the controller can determine the final danger level as indicating "danger", and in the case where both the first danger level and the second danger level indicate "safe", the controller can determine the final danger level as indicating "safe".

[0018] In the system, in the case where only one of the first danger level and the second danger level indicates "danger", the controller can determine the final danger level as "pending", and in the case where the final danger level indicates "pending", the controller can delay generation of a warning, and when the generation of the warning is delayed for a predetermined time or more, can generate a warning.

[0019] According to another aspect of the present disclosure, there is provided a target detection method for a vehicle, the method including: detecting a target based on a reflected wave by a detection sensor; determining, by a controller, a first danger level of the target in consideration of data acquired by the detection sensor and a change in velocity of the target in a lateral direction; determining, by the controller, a second danger level of the target based on data acquired by a camera sensor; and calculating, by the controller, a final danger level of the target based on the first danger level and the second danger level.

[0020] According to the present disclosure, even when a detected target moves in a longitudinal direction, in the case where an error can occur in calculating a danger level due to a change in velocity in a lateral direction, not only the result of recalculating a collision time and a collision point is considered, but also the presence or absence of a target detected by a camera sensor can be considered, a more accurate final danger level can be calculated. The collision time and the collision point are recalculated considering the change in velocity in the lateral direction. BRIEF DESCRIPTION OF DRAWINGS

[0021] The above and other objects, features and other advantages of the present disclosure will be more clearly understood from the following detailed description taken in conjunction with the accompanying drawings, in which:

[0022] Figure 1 is a block diagram illustrating a target detection system for a vehicle according to a first embodiment of the present disclosure;

[0023] Figure 2 is a view illustrating a case where an error warning occurs due to a change in velocity of a target in a lateral direction;

[0024] Figure 3 is a graph illustrating information for tracking a target that moves away in a longitudinal direction, the information being acquired through measurement of a detection sensor;

[0025] Figure 4 is a view illustrating a case where a target is located at a short distance;

[0026] Figure 5 is a view illustrating a case where a target is located at a distance further than a threshold distance away from the short distance in the lateral direction;

[0027] Figure 6 is a view illustrating a warning region that can be detected by a camera sensor; and

[0028] Figure 7 is a flowchart illustrating a target detection method for a vehicle according to a second embodiment of the present disclosure. DETAILED DESCRIPTION

[0029] Figure 1 is a block diagram illustrating a target detection system for a vehicle according to a first embodiment of the present disclosure. Figure 2 is a view illustrating a case where an error warning occurs due to a change in velocity of a target in a lateral direction even when the target moves in a longitudinal direction. Figure 3 is a graph illustrating positions and velocities in a longitudinal and lateral directions and illustrating generation of an error warning, the positions and velocities in the longitudinal and lateral directions being measured by a detection sensor 100 and being information for tracking a target that moves away in the longitudinal direction. Figure 4is a view showing a case where the target is located at a short distance. Figure 5 is a view showing a case where the target is located at a short distance beyond a threshold distance in a lateral direction. Figure 6 is a view showing a case where a warning area appears on the left and right sides of the vehicle, respectively, which can be detected by the camera sensor 300. Figure 7 is a flowchart showing a target detection method for a vehicle according to a second embodiment of the present disclosure.

[0030] Figure 1 is a view showing a target detection system for a vehicle according to a first embodiment of the present disclosure.

[0031] The target detection system for a vehicle according to the present disclosure is configured to include a detection sensor 100, a camera sensor 300, and a controller 500. The detection sensor 100 detects a target using a reflected wave. The camera sensor 300 detects a target using an image. The controller 500 calculates a danger level through data acquired by the detection sensor 100 and the camera sensor 300.

[0032] The controller 500 according to an exemplary embodiment of the present disclosure can be implemented by a non-volatile memory (not shown) and a processor (not shown). The non-volatile memory is configured to store data associated with an algorithm for controlling the operation of various components of the vehicle or data associated with data on a software command for executing the algorithm. The processor is configured to perform the following operations using the data stored in the non-volatile memory. Here, the memory and the processor can be implemented as separate chips, respectively. Alternatively, the memory and the processor can be implemented as a single integrated chip. The processor can be a single processor or a combination of two or more processors.

[0033] The detection sensor 100 is a sensor capable of detecting tracking information (distance, speed, azimuth, etc.) of a target based on a reflected wave. The detection sensor 100 can be installed at at least one of the front side, the rear side, and the side surface of the vehicle. As an implementation example, the detection sensor 100 can generate tracking information using the Doppler effect. In addition, the Micro-Doppler effect can be used. The detection sensor 100 can be a radar using a reflected wave, an ultrasonic sensor, a radar sensor, etc.

[0034] The camera sensor 300 can be installed on at least one of the front, rear, and side surfaces of the vehicle. The camera sensor 300 can acquire image data of the vicinity of the vehicle. As an implementation example, the camera sensor 300 can acquire an image including a warning area appearing on the right or left side of the vehicle by monitoring.

[0035] The controller 500 can determine the first danger level through data acquired by the detection sensor 100. Here, the data acquired by the detection sensor 100 can be tracking information of the target, such as distance, speed, and azimuth. The controller 500 can calculate a time-to-collision (TTC) and a collision point as factors for determining the first danger level. At this time, the controller 500 can calculate the TTC and the collision point using the position and speed in the longitudinal direction and the position and speed in the lateral direction. Also, even if the target moves in the longitudinal direction, in the case where the speed in the lateral direction changes, the controller 500 can determine the first danger level considering the change. In this case, the TTC and the collision point are recalculated by adding the change in the speed in the lateral direction to the values obtained by calculating the TTC and the collision point, respectively.

[0036] The controller 500 determines the first danger level as "danger" or "safe". At this time, in the case where the recalculated TTC is less than a threshold value of the TTC and the recalculated collision point falls within a threshold range of the collision point, the first danger level is determined to indicate "danger". In other cases, the first danger level is determined to indicate "safe". That is, even though the change in the speed in the lateral direction is considered, in the case where the TTC is short and the collision point is close to the vehicle, the first danger level is determined to indicate "danger" in a manner to generate a warning.

[0037] The camera sensor 300 can acquire an image through monitoring, and thus can detect a target. The camera sensor 300 can be a sensor capable of acquiring an RGB image. The controller 500 can determine a second danger level according to whether a target is detected from the image acquired by the monitoring of the camera sensor 300. Like the first danger level, the second danger level can be determined to indicate "danger" or "safe".

[0038] Also, the controller 500 can extract, as a warning area, portions of the left and right sides of the image acquired through the monitoring from the image. The target detected by the detection sensor 100 is likely to be extracted from the left and right sides of the image acquired by the camera sensor 300. Thus, the controller 500 can regard the portions of the left and right sides of the image as the warning area, and thus can determine the second danger level. Accordingly, in the case where the target exists in the portions extracted as the warning area, the controller 500 determines the second danger level as "danger". At this time, in the case where the image of the warning area changes, the controller 500 can determine that the target exists. More specifically, when a value of the image change of the warning area is calculated and the value of the image change is equal to or higher than a threshold value, it can be determined that the target exists, and thus the second danger level can be determined to indicate "danger".

[0039] The controller 500 determines a final danger level in consideration of the first danger level and the second danger level. In a case where both the first danger level and the second danger level indicate "danger", the final danger level is determined to indicate "danger". In a case where both the first danger level and the second danger level indicate "safe", the final danger level is determined to indicate "safe". In a case where the final danger level is "danger", a warning is generated. In a case where the final danger level is "safe", no warning is generated.

[0040] In a case where one of the first danger level and the second danger level indicates "danger", the final danger level is determined to indicate "pending". As an implementation example, in a case where the first danger level indicates "danger" and the second danger level indicates "safe", or in a case where the first danger level indicates "safe" and the second danger level indicates "danger", the final danger level can be determined to indicate "pending".

[0041] In a case where the final danger level indicates "pending", generation of a warning is delayed. At this time, when the generation of the warning is delayed for a predetermined time or more, the warning can be generated. The reason for this is to warn the driver of a situation in which a collision will occur only in a case where one of the first danger level and the second danger level. Furthermore, as an implementation example, even after the final danger level is determined to indicate "pending", because the first and second danger levels continue to be determined, the final danger level can change to "danger". In this case, when the generation of the warning is delayed for less than the predetermined time, the warning can be generated.

[0042] As an implementation example, the final danger level can be determined to indicate "high", "medium", or "low". In a case where both the first danger level and the second danger level indicate "danger", the final danger level is determined to be "high". In a case where one of the first danger level and the second danger level indicates "danger", the final danger level is determined to indicate "medium". In a case where both the first danger level and the second danger level indicate "safe", the final danger level is determined to be "low". At this time, in a case where the final danger level is determined to be "medium", generation of a warning is delayed.

[0043] Figure 2is a view showing a case where an error warning occurs due to a change in the speed and direction detected as a result of a change in the speed in the lateral direction Y even if the target T moves in the longitudinal direction X at the actual speed of the target T. Even if the direction in which the target T actually moves is the longitudinal direction X in which the user's vehicle moves, the target T can experience a change in the speed in the lateral direction Y. As an implementation example, in the case where the target T is a person moving in the longitudinal direction X, a micro-Doppler effect occurs in each part of his / her body, thereby causing a change in the speed in the lateral direction Y. This change causes a change in the speed and direction detected by the sensor of the vehicle. Thus, it is erroneously determined that the point P at which the target T approaches the vehicle moves. Thus, an error warning is caused to be generated.

[0044] Figure 3 is a graph showing information for tracking the target T in a case where the target T is moving away in the longitudinal direction X as detected by the detection sensor 100. Specifically, FIG. A shows a change in the moving distance of the target T in the longitudinal direction X over time, and a indicates an increase in the moving distance in the longitudinal direction X. FIG. B shows a change in the moving distance of the target T in the lateral direction Y over time, and b indicates that the moving distance of the target T in the lateral direction Y alternately increases and decreases when the target T moves away. FIG. C shows a change in the speed in the longitudinal direction X over time. FIG. D shows a change in the speed in the lateral direction Y over time. d indicates that the speed of the target T in the lateral direction Y changes even if the target T moves in the longitudinal direction X. FIG. E shows that a Rear Cross Traffic Alert (RCTA) generates an error warning e due to a change in the speed in the lateral direction Y even if the target T moves away in the longitudinal direction X.

[0045] Figure 4 is a view showing a case where the target T is located at a short distance. More specifically, in a case where the target T exists within a short distance range N away from the user's vehicle, it is determined that the target T is at a short distance. In a case where the target T exists within the short distance range N, the controller 500 can start determining the first danger level. At this time, the first danger level and the second danger level can be determined simultaneously. As an implementation example, the short distance range N can be a range of a point of impact, and specifically, the short distance can fall within a threshold range for the point of impact.

[0046] Figure 5is a view illustrating a case where the target T is located away from the short distance beyond a threshold distance in the lateral direction Y. As an implementation example, although the target T initially exists within the short distance range T, in a case where the target T moves beyond the short distance range N thereafter, the controller 500 does not determine the danger level. The reason for this is that if the user's vehicle is no longer at the short distance, the likelihood of a collision danger occurring is low. At this time, the moving distance away from the short distance range N in the lateral direction Y can be the threshold distance in the lateral direction Y. Thus, although the target T is initially located within the short distance range N, in a case where the target T moves the threshold distance or more in the lateral direction Y thereafter, the danger level is not determined.

[0047] Figure 6 is a view illustrating a warning area that can be detected by the camera sensor 300. The camera sensor 300 detects a target by monitoring an acquired image. The camera sensor 300 can be installed on at least one of the front, rear, and side of the vehicle. The controller 500 can extract a partial area appearing on the right side R and the left side L of the vehicle, respectively, as a warning area from the image acquired by monitoring. The controller 500 can determine whether an object T exists based on a value of an image change of the extracted warning area. More specifically, when the value of the image change of the warning area is calculated and the value of the image change is equal to or higher than a threshold value, it can be determined that the target exists, and thus the second danger level can be determined as indicating "danger". As an implementation example, in a case where the camera sensor 300 is installed on the rear of the vehicle, the controller 500 can extract a warning area including a parking guide line, and in this case, the right side R and the left side L of the parking guide line can be the warning area.

[0048] Figure 7 is a flowchart illustrating a target detection method for a vehicle according to a second embodiment of the disclosure.

[0049] Referring to Figure 7 The method includes a step of detecting a target by the detection sensor 100 using a reflected wave (S100), a step of determining a first danger level of the target by the controller 500 considering a speed change in a lateral direction of the target and data acquired by the detection sensor 100, a step of determining a second danger level of the target by the controller 500 through data acquired by the camera sensor 300, and a step of calculating a final danger level of the target by the controller 500 considering the first danger level and the second danger level.

[0050] More specifically, in a case where the detection sensor 100 detects a target, it is determined whether the target is initially located within a short distance from the vehicle (S200). In a case where the target is not located within the short distance, the target can be continuously detected without determining the danger level. Although the target is located within the short distance from the vehicle, in a case where the vehicle is subsequently outside the short distance range N, the danger level is not determined. That is, although the target is initially located at the short distance, in a case where the vehicle is subsequently moved by a threshold distance or more in the lateral direction (S300), the danger level is not determined.

[0051] In a case where the target is present within the close distance range N, the controller 500 calculates the TTC and the impact point based on the data acquired through the detection. At this time, in a case where the target experiences a change in speed in the lateral direction, the controller 500 re-calculates the TTC and the impact point considering the change in speed in the lateral direction (S400). That is, the TTC and the impact point are re-calculated by adding the change in speed in the lateral direction to the data of the calculated TTC and impact point. Then, the controller 500 determines the first danger level based on the re-calculated TTC and impact point (S500). At this time, the camera sensor 300 can monitor whether the target is present (S420), and the controller 500 can determine the second danger level using the data acquired through the camera sensor 300. As an implementation example, in a case where the first danger level indicates "danger", the second danger level can be determined (S520), and the first danger level and the second danger level can be simultaneously determined (520).

[0052] The controller 500 calculates the final danger level considering the first danger level and the second danger level. As an implementation example, in a case where both the first danger level and the second danger level indicate "danger", the final danger level is determined to indicate "danger", and a warning is generated (S600). Also, in a case where both the first danger level and the second danger level indicate "safe", the final danger level is determined to indicate "safe". Thus, a warning is not generated (S640). In a case where only one of the first danger level and the second danger level indicates "danger", the final danger level is determined to indicate "pending", and the generation of a warning is delayed. That is, in a case where the final danger level indicates "pending", even if any one of the first danger level and the second danger level indicates "danger", the generation of a warning is preferentially blocked (S620). When the generation of a warning is delayed for a predetermined time or more, the warning can be generated.

[0053] The above describes specific embodiments of the present disclosure, each feature of which is illustrated in the drawings, and it will be obvious to those of ordinary skill in the art that various modifications and changes can be made to the present disclosure without departing from the technical idea of the present disclosure.

Claims

1. A target detection system for a vehicle, the system comprising: A detection sensor that detects the target based on waves reflected from it; A camera sensor that monitors the target based on an image including the target; and The controller determines a first danger level of the target based on data acquired by the detection sensor and the target's velocity change in the lateral direction, determines a second danger level of the target based on information acquired by the camera sensor, and calculates a final danger level based on the first danger level and the second danger level. The controller determines the first hazard level and the second hazard level as indicating danger or safety. Where both the first hazard level and the second hazard level indicate hazard, the controller determines the final hazard level as indicating hazard. Where both the first danger level and the second danger level indicate safety, the controller determines the final danger level to indicate safety. Where only one of the first and second hazard levels indicates a hazard, the controller determines the final hazard level as pending. Specifically, if the final danger level indication is pending, the controller delays the generation of a warning, and generates the warning only after a predetermined time or longer has elapsed. Wherein, if the target exists within a threshold distance in the lateral direction, the controller determines the first danger level based on the target's velocity change in the lateral direction. Specifically, the controller extracts the left and right portions of the image as warning areas, calculates the image change value of the warning areas, and determines the second danger level based on the image change value.

2. The system according to claim 1, wherein, The controller calculates the collision time and collision point as factors for determining the first danger level, and In the event of a change in velocity in the lateral direction, the controller recalculates the collision time and the collision point by adding the change in velocity in the lateral direction to the collision time and the collision point.

3. The system according to claim 2, wherein, The controller determines the first hazard level as indicating danger or safety, and Specifically, if the recalculated collision time is shorter than a first threshold and the recalculated collision point falls within a second threshold, the controller determines the first danger level as an indicated danger.

4. The system according to claim 1, wherein, The controller determines the second danger level as indicating danger or safety, and extracts the left and right sides of the image acquired by the camera sensor through monitoring as warning areas. Wherein, if the target is present in the warning area, the controller determines the second danger level as an indicated danger.

5. The system according to claim 4, wherein, The controller calculates the value of the image change in the warning area, and Specifically, when the value of the image change is equal to or higher than a threshold, the controller determines the second danger level as an indicated danger.

6. A target detection method for a vehicle, the method comprising: Targets are detected using sensors based on reflected waves; The controller determines the first danger level of the target based on the data acquired by the detection sensor and the target's velocity change in the lateral direction. The controller determines a second level of danger for the target based on images acquired by the camera sensor. and The controller calculates the final danger level of the target based on the first danger level and the second danger level. In calculating the final hazard level, the first hazard level and the second hazard level are determined to be either indicating hazard or safety. Where both the first hazard level and the second hazard level indicate hazard, the final hazard level is determined to be hazard-indicating. Where both the first hazard level and the second hazard level indicate safety, the final hazard level is determined to indicate safety. Specifically, when calculating the final hazard level, if only one of the first hazard level and the second hazard level is an indicated hazard, the final hazard level is determined to be undetermined. Specifically, if the final danger level indication is pending, the generation of the warning is delayed; and if the generation of the warning is delayed by a predetermined time or longer, the warning is generated. Specifically, when determining the first danger level, if the target moves within a threshold distance in the lateral direction, the first danger level is determined based on the change in the target's velocity in the lateral direction. Specifically, when determining the second danger level, the controller extracts the left and right portions of the image as warning areas, calculates the image change value of the warning areas, and determines the second danger level based on the image change value.

7. The method according to claim 6, wherein, In determining the first hazard level, the collision time and collision point are calculated as factors used to determine the first hazard level. In the event of a change in velocity in the lateral direction, the collision time and the collision point are recalculated by adding the change in velocity in the lateral direction to the collision time and the collision point.

8. The method according to claim 7, wherein, In determining the first hazard level, the first hazard level is determined to be either indicative of danger or safety, and Specifically, if the recalculated collision time is shorter than the first threshold and the recalculated collision point falls within the range of the second threshold, the first danger level is determined as an indicated danger.

9. The method according to claim 6, wherein, When determining the second hazard level, the second hazard level is determined to indicate danger or safety, and the left and right sides of the image acquired by the camera sensor through monitoring are extracted as warning regions, and the value of the image change in the warning regions is calculated. Wherein, if the value of the image change in the warning area is equal to or higher than the threshold, the second danger level is determined to indicate danger.

Images