JP2010097488A - Peripheral monitoring device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a peripheral monitoring device distinguishing and recognizing an equipment user moving by use of equipment and a pedestrian. <P>SOLUTION: This peripheral monitoring deviceing device 1 includes a laser radar sensor 2, a camera sensor 3, and an ECU (Electronic Control Unit) 4. The sensors 2, 3 recognizes a shape of an object 11 around a vehicle 10. The ECU 4 decides presence/absence of a person candidate based on a shape recognition result, calculates a motion of the person candidate, decides presence/absence of a play equipment use motion based on the calculated motion, and detects the person candidate as the play equipment user H2 moving by use of the play equipment when it is decided that the play equipment use motion is present. The peripheral monitoring deviceing device 1 can distinguish the equipment user H2 from the pedestrian H1, and detect and recognize the equipment user H2. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a periphery monitoring device for monitoring the periphery of a vehicle such as an automobile.

As a conventional periphery monitoring device, for example, a device described in Patent Document 1 is known. In this periphery monitoring device, an object around the vehicle is recognized by distinguishing whether it is a person or an automobile, and the degree of risk is calculated according to the type of the recognized object.
JP 2003-81039 A

  Here, a tool user who moves using a tool (hereinafter also simply referred to as a “tool user”) and a pedestrian have different characteristics when moving, even for the same person. For this reason, for example, when performing vehicle driving assistance or the like, it is desirable that these can be distinguished. Therefore, in the periphery monitoring device as described above, it is required that the tool user and the pedestrian can be distinguished and recognized.

  Then, this invention makes it a subject to provide the periphery monitoring apparatus which can distinguish and recognize the tool user and pedestrian who move using a tool.

  In order to solve the above-described problem, a periphery monitoring device according to the present invention is a periphery monitoring device for monitoring the periphery of a vehicle, and includes a person recognition unit that recognizes a person and a predetermined person about the person recognized by the person recognition unit. Motion determining means for determining presence / absence of motion, and detection means for detecting a person recognized by the human recognition means as a tool user who moves using the tool based on a determination result by the motion determining means. It is characterized by that.

  In the periphery monitoring device of the present invention, since the person recognized by the person recognition means is detected as a tool user based on the determination result by the motion determination means, the tool user is detected and recognized separately from the pedestrian. Can do. Here, “pedestrian” means a person who moves without using a tool, and includes a person who walks and runs.

  Here, as a configuration that preferably exhibits the above-described effects, specifically, a configuration in which the motion determination unit determines whether or not there is a change in posture and a change in movement direction as the presence or absence of the predetermined motion.

  Further, the apparatus further comprises movement state recognition means for recognizing the body direction and movement direction of the person recognized by the person recognition means, and the detection means detects the detected tool user based on the recognition result by the movement state recognition means. It is preferable to classify according to. Thereby, for example, a tool user is classified into a case where a tool (for example, roller skate) whose body direction and moving direction match is used, and a case where a tool that does not match (for example, skateboard) is used. Can do. Therefore, it becomes possible to subdivide and detect tool users according to the tools to be used.

  ADVANTAGE OF THE INVENTION According to this invention, it becomes possible to distinguish and recognize the tool user and pedestrian who move using a tool.

  Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. In the following description, the same or equivalent elements will be denoted by the same reference numerals, and redundant description will be omitted.

  First, a first embodiment of the present invention will be described. FIG. 1 is a schematic configuration diagram showing a periphery monitoring device according to the first embodiment of the present invention, and FIG. 2 is a flowchart showing a processing procedure of the periphery monitoring device of FIG. As shown in FIG. 1, the periphery monitoring device 1 of this embodiment is mounted on a vehicle 10 such as an automobile and is for monitoring the periphery of the vehicle 10. Here, the periphery monitoring device 1 monitors an object 11 around the vehicle 10 traveling on the roadway 12 as an obstacle for operation control such as automatic driving and cruise control (including the entire vehicle speed).

  The periphery monitoring device 1 includes an electronic control unit (ECU) mainly composed of a laser radar sensor 2 attached to a front portion of a vehicle 10, a camera sensor 3 that captures the periphery of the vehicle, and a computer including, for example, a CPU and a ROM. Unit) 4. In the periphery monitoring device 1, the shape of the object 11 is recognized by the sensors 2 and 3 (S1 in FIG. 2), and the ECU 4 executes the following processing based on the shape recognition result.

  That is, based on the shape recognition result of the object 11 by the sensors 2 and 3, the presence / absence of a candidate estimated as a person (hereinafter simply referred to as “person candidate”) is determined (S2). If there is no person candidate, the process is terminated as it is, while if there is a person candidate, the operation of the person candidate is calculated (S3). Here, the temporal change and the moving direction of the shape of the human candidate are calculated as the operation of the human candidate.

  Subsequently, a motion determination is executed based on the calculated motion, and based on the motion determination, a human candidate is detected and recognized as a pedestrian H1, a playground equipment user H2, or unknown.

  Specifically, the presence or absence of a walking motion is determined based on the calculated motion, and if it is determined that there is a walking motion, a human candidate is detected as a normal pedestrian H1 (S4 → S5). Examples of the walking motion include motions of walking feet and up and down swinging motions. As a result, the driving control parameters of the vehicle 10 are corrected so as to correspond to the movement of the pedestrian H1, and the brake actuator 17 is controlled in accordance with the movement of the pedestrian H1.

  Further, based on the calculated operation, the presence / absence of a playground equipment use operation (predetermined operation) that is an operation in the state of using the playground equipment is determined. It is detected as a playground equipment user H2 that moves using the playground equipment (S4 → S6). More specifically, as a playground equipment use operation, for example, it is determined whether or not a person candidate is moving in a substantially constant direction and in a substantially constant direction for a short time of about several seconds. It is detected as a playground equipment user (hereinafter referred to as “roller playground equipment user”) H2 using a roller skate, a skateboard, a kickboard, or the like. As a result, the operation control parameters of the vehicle 10 are corrected so as to correspond to the movement of the roller playground equipment user H2, and the brake actuator 17 is controlled in accordance with the movement of the roller playground equipment user H2.

  Furthermore, it is determined whether or not the camera is in a stationary state (substantially no movement) based on the calculated motion. If it is determined that the camera is in a stationary state, the candidate is detected as unknown, and the process ends as it is (S4 → S7). Here, even when it is determined that the walking motion and the playground equipment use motion are not present based on the calculated motion, the human candidate is detected as unknown.

  By the way, in the conventional peripheral monitoring device, since it is difficult to distinguish and recognize the roller playground equipment user H2 that is not different from the pedestrian in appearance, the roller playground equipment user H2 is erroneously recognized as a pedestrian, and driving control is performed. There is a risk that the movement speed, movement mode, and the like estimated at the time will be incorrect.

  On the other hand, in this embodiment, as described above, the motion of the recognized human candidate is calculated, and the human candidate is detected as the pedestrian H1 or the playground equipment user H2 based on this motion. Therefore, it becomes possible to recognize playground equipment user H2 separately from a pedestrian. As a result, it is possible to realize driving control in accordance with each of the movements of the pedestrian and the playground equipment user, and the contact avoidance effect between the vehicle 10 and the person can be further enhanced.

  In particular, when a planting 14 for separating them is disposed between the roadway 12 and the sidewalk 13 and the playground equipment user H2 is equipped with a roller playground equipment 15 such as roller skates (see FIG. 1), this implantation is performed. Therefore, the roller playground equipment user H2 is often recognized as the pedestrian H1 or is recognized as unknown. In this regard, in the present embodiment, as described above, in order to determine whether or not the robot is moving in a substantially constant direction and in a substantially constant direction for a short time, and detecting a human candidate as the roller playground equipment user H2, The pedestrian H1 and the roller playground equipment user H2 can be distinguished and recognized with high accuracy. This is because the roller playground equipment user H2 normally slides in a short period of time and thus has almost no posture change and moves in a certain direction.

  Next, a second embodiment of the present invention will be described. In the description of the present embodiment, differences from the first embodiment will be mainly described.

FIG. 3 is a flowchart showing a processing procedure of the periphery monitoring apparatus according to the second embodiment of the present invention. In the periphery monitoring device of the present embodiment, the ECU 4 executes the following process in addition to the process of the first embodiment.

That is, when the recognized human candidate is detected as the roller playground equipment user H2, the movement vector (movement direction) of the roller playground equipment user H2 is calculated and recognized based on the shape recognition result by the sensors 2 and 3 (FIG. 3). S11 → S12). Moreover, based on the shape recognition result by the sensors 2 and 3, the direction of the body of the roller playground equipment user H2 is calculated and recognized (S13). Then, the correlation between the movement vector and the body direction is determined, and the roller playground equipment user H2 is classified for each roller playground equipment 15 used based on the correlation.

  Specifically, when it is determined that the movement vector and the body direction are substantially orthogonal, the roller playground equipment user H2 is detected as a user such as a skateboard (S14 → S15). On the other hand, when it is determined that the movement vector and the body direction are substantially the same, the roller playground equipment user H2 is detected as a user such as roller skates (S14 → S16). On the other hand, when there is no correlation between the movement vector and the body direction (orthogonal and inconsistent), the roller playground equipment user H2 is detected as unknown (S14 → S17).

  And the subsequent moving direction of the roller playground equipment user H2 is estimated according to the classified roller playground equipment 15. Thereby, based on this prediction result, the parameter of the driving control of the vehicle 10 is corrected and the brake actuator 17 is controlled.

  As described above, according to the present embodiment, the roller playground equipment user H2 is classified according to whether the roller skate is used or the skateboard is used, and the roller playground equipment user H2 is subdivided according to the roller playground equipment 15 used. Can be detected. That is, the roller playground equipment 15 can be recognized individually without being recognized all at once, and the roller playground equipment user H2 can be identified according to the type. Here, it is possible to distinguish between a skateboard or the like and a roller skate or the like among those who are equipped with the roller play equipment 15. Thereby, it becomes possible to predict the operation of the playground equipment user H2 with higher accuracy.

  Here, as shown in FIG. 4, when the roller playground equipment 15 used by the roller playground equipment user H2 is the skateboard 16, the direction of the person is on the side of the roadway 12 or on the opposite side of the roadway 12, so that general peripheral monitoring is performed. According to the apparatus, there is a possibility that the roller playground equipment user H2 erroneously estimates that the user is traveling on the roadway. On the other hand, in this embodiment, since it is possible to distinguish between a skateboard or the like and a roller skate or the like, even if the roller playground equipment user H2 is facing the roadway 12 side, it is possible to appropriately estimate that it does not proceed to the roadway 12 side. In addition, operation control can be performed in consideration of unforeseen circumstances such as falling down.

  In addition, “skateboard etc.” intends a roller playground equipment in which the movement vector and the body direction at the time of use are almost orthogonal, such as a skateboard, and “roller skate etc.” In this way, the roller playground equipment in which the movement vector and the body direction at the time of use almost coincide is intended.

  The preferred embodiment of the present invention has been described above, but the present invention is not limited to the above embodiment.

  For example, in the above embodiment, the playground equipment user H2 that uses the playground equipment is detected, but the use of a tool that uses a tool (for example, a wheelchair) that is not a playground equipment but is used for movement other than the vehicle. A person may be detected. Note that the play equipment to which the present invention is applicable is not limited to the roller play equipment 15 described above, and may be applied to various kinds of play equipment.

  In the above embodiment, the presence / absence of a playground equipment use operation, which is a predetermined operation, is determined by determining whether or not the vehicle is moving in a substantially constant direction for a short time, but the present invention is not limited to this. What is necessary is just to determine the presence or absence of the predetermined operation | movement according to the tool to be used.

  By the way, in the above, the sensors 2 and 3 and the ECU 4 for determining the presence / absence of a human candidate constitute a person recognition means, and the ECU 4 for judging the presence / absence of a playground equipment use action constitutes an action determination means and detects the tool user H2. The ECU 4 that performs the detection constitutes detection means, and the sensors 2 and 3 and the ECU 4 that calculates the movement vector and the body orientation constitute movement state recognition means.

It is a schematic block diagram which shows the periphery monitoring apparatus which concerns on 1st Embodiment of this invention. It is a flowchart which shows the process sequence of the periphery monitoring apparatus of FIG. It is a flowchart which shows the process sequence of the periphery monitoring apparatus which concerns on 2nd Embodiment of this invention. It is a figure for demonstrating the case where the roller playground equipment user who utilizes a skateboard is recognized in the periphery monitoring apparatus which concerns on 2nd Embodiment of this invention.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Perimeter monitoring apparatus, 2 ... Laser radar sensor (person recognition means), 3 ... Camera sensor (person recognition means), 4 ... ECU (person recognition means, operation | movement determination means, detection means, movement state recognition means), 10 ... Vehicle, H2 ... Tool user.

Claims (3)

A periphery monitoring device for monitoring a vehicle periphery,
Human recognition means for recognizing people,
Action determining means for determining the presence or absence of a predetermined action for the person recognized by the person recognition means;
Peripheral monitoring device, comprising: detecting means for detecting the person recognized by the person recognition means as a tool user who moves using a tool based on a determination result by the motion determination means .   The periphery monitoring device according to claim 1, wherein the motion determination unit determines whether or not there is a change in posture and a change in movement direction as the presence or absence of the predetermined motion. A moving state recognizing means for recognizing the body direction and moving direction of the person recognized by the person recognizing means;
The periphery monitoring device according to claim 1, wherein the detection unit classifies the detected tool user according to the tool based on a recognition result by the movement state recognition unit.

Images