EP-4737947-A1 - METHOD AND DEVICE FOR DETERMINING POSITION OF OBSTACLE IN SURROUNDING ENVIRONMENT OF VEHICLE

EP4737947A1EP 4737947 A1EP4737947 A1EP 4737947A1EP-4737947-A1

Abstract

The present application relates to a method for determining the position of an obstacle in the surroundings of a vehicle based on ultrasonic signals, comprising: clustering at least a portion of ultrasonic data samples in an ultrasonic data sample set to obtain one or more ultrasonic data sample clusters; determining one or more corresponding bounding boxes based on the one or more ultrasonic data sample clusters, wherein the one or more bounding boxes respectively comprise one or more ultrasonic data sample subsets in the ultrasonic data sample set; determining, for an ultrasonic data sample subset in the one or more ultrasonic data sample subsets, ultrasonic signal-related features of the ultrasonic data sample subset, and determining the position of an obstacle corresponding to the ultrasonic data sample subset based on the ultrasonic signal-related features of the ultrasonic data sample subset.

Inventors

LIU, SHU
ZHANG, Kaixuan
YANG, HAO
GU, CHENYI
CAO, Longxu

Assignees

Robert Bosch GmbH

Dates

Publication Date: 20260506
Application Date: 20230628

Claims (20)

A method for determining the position of an obstacle in the surroundings of a vehicle based on ultrasonic signals, comprising: clustering at least a portion of ultrasonic data samples in an ultrasonic data sample set to obtain one or more ultrasonic data sample clusters; determining one or more corresponding bounding boxes based on the one or more ultrasonic data sample clusters, wherein the one or more bounding boxes respectively comprise one or more ultrasonic data sample subsets in the ultrasonic data sample set; and determining, for an ultrasonic data sample subset in the one or more ultrasonic data sample subsets, ultrasonic signal-related features of the ultrasonic data sample subset, and determining the position of an obstacle corresponding to the ultrasonic data sample subset based on the ultrasonic signal-related features of the ultrasonic data sample subset.
The method according to claim 1, wherein the ultrasonic data sample set comprises a first type of ultrasonic data sample set and a second type of ultrasonic data sample set; wherein clustering at least a portion of the ultrasonic data samples in the ultrasonic data sample set comprises: clustering only the first type of ultrasonic data sample set to obtain the one or more ultrasonic data sample clusters; and wherein the one or more ultrasonic data sample subsets respectively comprised in the one or more bounding boxes comprise the first type of ultrasonic data samples and the second type of ultrasonic data samples respectively comprised in the one or more bounding boxes.
The method according to claim 2, wherein the first type of ultrasonic data samples comprise intersection point data samples, and the second type of ultrasonic data samples comprise echo data samples.
The method according to claim 3, wherein the echo data samples comprise at least a portion of the following: echo first coordinates, echo second coordinates, echo amplitude, echo significance, and echo distance of corresponding echo of each ultrasonic signal obtained based on a center line detection method; and/or wherein the intersection point data samples comprise at least a portion of the following: intersection point first coordinates, intersection point second coordinates, intersection point distance, intersection point neighbor distance, first sensor coordinates, second sensor coordinates, intersection point neighbor first coordinates, intersection point neighbor second coordinates, and angle with the sensor.
The method according to any one of claims 1 to 4, wherein clustering at least a portion of ultrasonic data samples in the ultrasonic data sample set comprises: clustering at least a portion of the ultrasonic data samples in the ultrasonic data sample set by Density-Based Spatial Clustering of Applications with Noise (DBSCAN).
The method according to any one of claims 1 to 4, wherein determining one or more bounding boxes based on the one or more ultrasonic data sample clusters comprises: determining a bounding box corresponding to each ultrasonic data sample cluster in the one or more ultrasonic data sample clusters based on a spatial dimension corresponding to the ultrasonic data sample cluster, wherein the bounding box contains the ultrasonic data sample cluster and the additional ultrasonic data samples.
The method according to claim 6, wherein determining a bounding box corresponding to each ultrasonic data sample cluster in the one or more ultrasonic data sample clusters based on a spatial dimension corresponding to the ultrasonic data sample cluster comprises: determining a first sample endpoint and a second sample endpoint in the vehicle's travel direction, and a third sample endpoint and a fourth sample endpoint in a direction perpendicular to the travel direction in the ultrasonic data sample cluster; and expanding a boundary defined by the first, second, third, and fourth sample endpoints in the travel direction and the perpendicular direction to obtain a bounding box for the ultrasonic data sample cluster.
The method according to claim 6, wherein determining a bounding box corresponding to each ultrasonic data sample cluster in the one or more ultrasonic data sample clusters based on a spatial dimension corresponding to the ultrasonic data sample cluster comprises: determining a first sample endpoint and a second sample endpoint in the vehicle's travel direction in the ultrasonic data sample cluster; determining a distance between the first and second sample endpoints in the travel direction; determining a bounding box dimension from a plurality of bounding box dimensions based on whether the distance falls within a distance interval of a plurality of distance intervals, wherein the plurality of distance intervals correspond to the plurality of bounding box dimensions; and determining a bounding box having the determined bounding box dimension for the ultrasonic data sample cluster.
The method according to any one of claims 1 to 4, wherein two of the one or more bounding boxes overlap, the method further comprising: determining whether to merge the two overlapping bounding boxes into one bounding box based on first features of ultrasonic data samples from two ultrasonic data sample subsets within the two bounding boxes; and/or determining whether to merge the two overlapping bounding boxes into one bounding box based on a comparison of the overlapping area of the two bounding boxes and the merged area of the two bounding boxes.
The method according to claim 1, wherein determining ultrasonic signal-related features of the ultrasonic data sample subset comprises: determining ultrasonic signal-related features of the ultrasonic data sample subset based on all ultrasonic data samples in the ultrasonic data sample subset; or determining ultrasonic signal-related features of the ultrasonic data sample subset based on a first portion of samples at an end of the ultrasonic data sample subset and a second portion of samples at another end of the ultrasonic data sample subset.
The method according to claim 1, wherein determining ultrasonic signal-related features of the ultrasonic data sample subset comprises: determining a proportion of a portion of data samples in the ultrasonic data sample subset based on the dimension of a bounding box corresponding to the ultrasonic data sample subset; determining a first portion of samples at an end and a second portion of samples at the other end of the ultrasonic data sample subset based on the proportion; determining the ultrasonic signal-related features of the ultrasonic data sample subset based on the first portion of samples and the second portion of samples.
The method according to claim 10 or 11, wherein determining ultrasonic signal-related features of the ultrasonic data sample subset comprises: determining a first set of features and a second set of features related to the ultrasonic signals of the ultrasonic data sample subset based on the first portion of samples, and determining a third set of features and a fourth set of features related to the ultrasonic signals of the ultrasonic data sample subset based on the second portion of samples.
The method according to claim 12, wherein determining ultrasonic signal-related features of the ultrasonic data sample subset comprises: determining a first set of features and a second set of features related to the ultrasonic signals of the ultrasonic data sample subset by respectively calculating a first portion of statistics and a second portion of statistics from a plurality of statistics for the first portion of features and the second portion of features of the ultrasonic data in the first portion of samples, and determining a third set of features and a fourth set of features related to the ultrasonic signals of the ultrasonic data sample subset by respectively calculating a third portion of statistics and a fourth portion of statistics from the plurality of statistics for the third portion of features and the fourth portion of features of the ultrasonic data in the second portion of samples.
The method according to claim 13, wherein determining the position of an obstacle corresponding to the ultrasonic data sample subset comprises: determining a first coordinate value and a second coordinate value of a first position of the obstacle corresponding to the ultrasonic data sample subset based on the first set of features and the second set of features related to the ultrasonic signals of the ultrasonic data sample subset, and determining a third coordinate value and a fourth coordinate value of a second position of the obstacle corresponding to the ultrasonic data sample subset based on the third set of features and the fourth set of features related to the ultrasound signals of the ultrasonic data sample subset.
The method according to claim 14, wherein determining the position of the obstacle corresponding to the ultrasonic data sample subset based on the ultrasonic signal-related feature of the ultrasonic data sample subset is performed by an xgboost model, wherein the xgboost model comprises a first xgboost submodel, a second xgboost submodel, a third xgboost submodel, and a fourth xgboost submodel, wherein the first xgboost submodel, the second xgboost submodel, the third xgboost submodel, and the fourth xgboost submodel respectively generate the first coordinate value, the second coordinate value, the third coordinate value, and the fourth coordinate value based on the first set of features, the second set of features, the third set of features, and the fourth set of features related to the ultrasound signals of the ultrasonic data sample subset.
A method for detecting an empty parking space, comprising: determining the position of an obstacle using the method according to any one of claims 1 to 15; and detecting an empty parking space based on the determined position of the obstacle.
A method for assisting parking, comprising: detecting an empty parking space using the method according to claim 16; planning a route for driving the vehicle into the empty parking space based on the detected empty parking space.
A control system for a vehicle, comprising: one or more processing units, the processing units, when executing program instructions, configured to perform the method according to one of claims 1 to 17.
A vehicle, comprising: an ultrasonic sensor for transmitting and receiving ultrasonic signals; the control system according to claim 18.
A machine-readable storage medium having executable instructions stored thereon, the instructions, when executed, causing one or more processors to perform the method according to any one of claims 1 to 17.

Description

Technical Field The present application relates to automatic vehicle control, and more particularly, to a method and an apparatus for determining the position of an obstacle in the surroundings of the vehicle based on ultrasonic signals. Background Parking space detection is a crucial step in implementing vehicle-assisted parking. This can be accomplished by utilizing sensor signals to identify obstacles in the vehicle's surroundings. By enhancing the accuracy of obstacle position detection, parking space availability can be more precisely determined, thereby improving the performance of vehicle-assisted parking. Ultrasonic sensors (USSs) offer a cost-effective method for obstacle detection. However, traditional detection methods that rely on ultrasonic signals are usually rule-based and may not be effectively applied in complex scenarios. Additionally, the accuracy of obstacle position detection using ultrasonic signals requires improvement to enhance the performance of vehicle-assisted parking. Summary of the Invention The following introduction is provided in order to introduce selected concepts in a simple manner, and these concepts will be further described in the detailed description below. The introduction is not intended to highlight the key or necessary features of the claimed subject matter, nor is it intended to limit the scope of the claimed subject matter. According to one aspect of the present application, a method for determining the position of an obstacle in the surroundings of a vehicle based on ultrasonic signals is provided, wherein the vehicle has an ultrasonic sensor for transmitting and receiving ultrasonic signals to monitor the vehicle's surroundings. The method comprises: clustering at least a portion of ultrasonic data samples in an ultrasonic data sample set to obtain one or more ultrasonic data sample clusters; determining one or more corresponding bounding boxes based on the one or more ultrasonic data sample clusters, wherein the one or more bounding boxes respectively comprise one or more ultrasonic data sample subsets in the ultrasonic data sample set; and determining, for the ultrasonic data sample subset in the one or more ultrasonic data sample subsets, ultrasonic signal-related features of the ultrasonic data sample subset, and determining the position of an obstacle corresponding to the ultrasonic data sample subset based on the ultrasonic signal-related features of the ultrasonic data sample subset. According to one aspect of the present application, a method for detecting an empty parking space is provided, comprising: determining the position of an obstacle in the surroundings of a vehicle based on ultrasonic signals by performing the method described herein; and detecting an empty parking space based on the determined position of the obstacle. According to one aspect of the present application, a method for assisted parking is provided, comprising: determining the position of an obstacle in the surroundings of a vehicle based on ultrasonic signals by performing the method described herein; detecting an empty parking space based on the determined position of the obstacle; and planning a route for driving the vehicle into the empty parking space based on the detected empty parking space. According to one aspect of the present application, a control system for a vehicle is provided, comprising: one or more processing units, which, when executing program instructions, are configured to perform the method described herein for determining the position of an obstacle in the surroundings of the vehicle based on ultrasonic signals. According to one aspect of the present application, a machine-readable storage medium is provided, which stores executable instructions that, when executed, cause one or more processors to perform the method described herein for determining the position of an obstacle in the surroundings of a vehicle based on ultrasonic signals. By utilizing the method of determining the position of an obstacle in the surroundings of a vehicle based on ultrasonic signals according to the present application, it is possible to accurately detect the position of the obstacle based on ultrasonic signals in various complex scenarios. This capability enables precise detection of an empty parking space, thereby enhancing the performance of the assisted parking function. Brief Description of the Drawings The nature and advantages of the present disclosure may be further implemented by referring to the following accompanying drawings. In the drawings, similar components or features may have the same reference signs. FIG. 1 is a schematic diagram of a vehicle according to one example.FIG. 2 is a schematic diagram of a control system in a vehicle according to one example.FIG. 3 is a schematic diagram of ultrasonic data according to one example.FIG. 4 is a schematic diagram of a method and module for determining the position of an obstacle around a vehicle according to one ex