Rear Vehicle Detection with Overlap Rate for Lane-Change Warnings

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional vehicle detection systems, such as ADAS, face challenges in accurately determining the trajectory of a target vehicle approaching from behind, leading to unnecessary alarm warnings when the vehicle changes lanes, as the estimation of crossing the driving lane can be difficult and inaccurate, especially at large distances.

Innovation Solution

A vehicle detection system that uses judging means to assess the overlap change rate of a target vehicle's lateral position over time, determining if it is an alert object by measuring the difference in lateral overlap divided by a cycle time, and setting a warning threshold to suppress alarms if the vehicle is not deemed an alert object, thereby improving the accuracy of predicting the vehicle's trajectory during lane changes.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional radar-based detection systems are used to track vehicles approaching from behind, then the system can detect target vehicles in adjacent lanes, but the estimation of crossing the driving lane becomes difficult and inaccurate at large distances, leading to erroneous merge-in detection and unnecessary warning signals

Engineering Contradiction:
Improvetrajectory estimation accuracyVSAvoidloss of trajectory information
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The patent introduces a new parameter dimension - the temporal dimension of lateral overlap measurement. By continuously measuring lateral overlap at multiple time points and calculating the rate of change over time, the system transforms a single static measurement problem into a dynamic multi-dimensional analysis, enabling accurate trajectory prediction even at large distances where conventional single-point estimation fails

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Solution Approach 2:

The system implements feedback by continuously monitoring the overlap change rate and comparing it against threshold values. This feedback mechanism allows the system to adaptively determine whether a target vehicle is performing a lane change or merge-in maneuver, correcting erroneous detections by using the temporal evolution of lateral overlap information to refine trajectory estimation

Inventive Principle:
Principle #23Feedback

2Reliability

If the system triggers warning signals for all detected target vehicles in the driving lane, then safety coverage is maximized, but unnecessary alarm warnings are issued when the target vehicle is actually changing lanes diagonally across the driving lane

Engineering Contradiction:
Improvewarning system reliabilityVSAvoidfalse alarm warnings
Core Design Contradiction:
ReliabilityVSObject-generated harmful factors

Solution Approach 1:

The patent applies dynamics by transitioning from static threshold-based detection to dynamic rate-of-change analysis. Instead of merely detecting the presence of a target vehicle in the driving lane, the system dynamically calculates the overlap change rate and uses this temporal derivative to distinguish between different vehicle behaviors, thereby reducing false alarms while maintaining reliable warning coverage

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system changes the detection parameter from simple presence/position detection to rate-of-change detection. By measuring how the lateral overlap changes over time rather than just the absolute position, the system can differentiate between vehicles maintaining lane position versus those executing lane changes, improving warning reliability while eliminating false alarms

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP4357807A1Vehicle detection system and method for detecting a target vehicle in a detection area located behind a subject vehicle
Publication Date: 2024.04.24 AUMOVIO AUTONOMOUS MOBILITY GERMANY GMBH
  • EP4357807A1 patent drawingFigure 1
  • EP4357807A1 patent drawingFigure 2
  • EP4357807A1 patent drawing

AI summary

The invention relates to a vehicle detection system (10) for detecting a target vehicle (60) in a detection area (DA) located behind a subject vehicle (50), the vehicle detection system (10) comprising judging means (20) configured to judge whether a target vehicle (60) which has been detected to approach the subject vehicle (50) occupying a driving lane (DL) from behind in a first adjacent lane (ALF) which is on one side of the driving lane (DL) and which target vehicle (60) has been further detected to change from the first adjacent lane (ALF) by diagonally passing the driving lane (DL) on to a second adjacent lane (ALS) on the other side of the driving lane (DL), is an alert object, to output a result of judgement to warning means (30) which warns a driver of the subject vehicle (50) that a target vehicle (60) is present behind the subject vehicle (50) based on the judging means (20) judging the target vehicle (60) to be an alert object, and to determine, on the basis of a change of a lateral overlap (65) of the target vehicle (60) in the detection area (DA), an overlap change rate which represents the change in size of the lateral overlap (65) over a specified time, wherein the judging means (20) is configured to judge the target vehicle (60) to be an alert object or to be not an alert object on the basis of the overlap change rate. Moreover, the invention relates to a corresponding method for detecting a target vehicle in a detection area located behind a subject vehicle.