Driver Assistance System Pedestrian Detection Occlusion Handling

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current driver assistance systems face challenges in accurately detecting and tracking pedestrians using a combination of camera and radar data, particularly when pedestrians are occluded by stationary obstacles, leading to potential delays in detection and increased risk of collisions.

Innovation Solution

A driver assistance system that integrates an image sensor and radar to detect stationary obstacles, sets a range of interest (ROI) around them, and uses radar data to identify pedestrian candidates, with a camera confirming the detection based on speed and image data, allowing for timely and accurate pedestrian detection even when occluded.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If camera and radar data are fused to detect pedestrians, then detection reliability is improved, but detection accuracy deteriorates when pedestrians are occluded by stationary obstacles

Engineering Contradiction:
Improvepedestrian detection reliabilityVSAvoidpedestrian detection accuracy
Core Design Contradiction:
ReliabilityVSMeasurement precision

Solution Approach 1:

The system performs preliminary detection using radar to identify pedestrian candidates before camera confirmation, allowing early identification of potential pedestrians even when occluded. The radar continuously monitors the ROI and detects objects entering the monitoring range, enabling proactive detection before visual confirmation is possible.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system introduces an intermediate monitoring range around stationary obstacles as a buffer zone. This intermediary spatial region allows radar to detect objects that may emerge from behind obstacles, serving as a transition zone between occluded and visible states, and enabling earlier detection before complete occlusion occurs.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Reliability

If the entire field of view is monitored by the camera, then pedestrian detection coverage is improved, but computational time increases

Engineering Contradiction:
Improvepedestrian detection coverageVSAvoidcomputational delay
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The system segments the field of view into multiple regions of interest centered around detected stationary obstacles. Instead of processing the entire camera field of view, the system divides attention into discrete ROI zones where pedestrians are most likely to appear, reducing the computational burden while maintaining comprehensive coverage of critical areas.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system applies enhanced monitoring quality specifically to regions around stationary obstacles where pedestrian detection is most critical. The camera and radar resources are concentrated in these local zones with higher pedestrian likelihood, while other areas receive standard monitoring, optimizing the balance between coverage and computational efficiency.

Inventive Principle:
Principle #3Local quality

3Measurement precision

If radar monitoring range is expanded to cover all areas, then pedestrian candidate identification is improved, but system complexity increases

Engineering Contradiction:
Improvepedestrian candidate identification accuracyVSAvoidmonitoring system complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system performs preliminary identification of stationary obstacles using radar before establishing monitoring ranges. This preliminary detection step allows the system to focus subsequent monitoring efforts only on relevant areas around obstacles, rather than uniformly monitoring the entire space, thereby reducing system complexity while maintaining detection accuracy.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The monitoring range is dynamically adjusted based on the position and characteristics of detected stationary obstacles. The ROI expands and contracts, moves and repositions, according to the real-time locations of obstacles and pedestrians, allowing the system to adapt its complexity to the actual environmental conditions rather than maintaining fixed high-complexity monitoring everywhere.

Inventive Principle:
Principle #15Dynamics

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

Enhances pedestrian detection accuracy and efficiency by pre-selecting candidates using radar data and confirming with camera images, reducing computational delays and improving reliability in occluded scenarios.

Implementation Method 1

a radar mounted to the vehicle to have a field of sensing toward the outside of the vehicle and configured to acquire radar data

Methodology Applied
Scientific EffectRadar: Radar

Implementation Method 2

an image sensor mounted to a vehicle to have a field of view forward of the vehicle and configured to acquire image data

Methodology Applied
Scientific EffectLight reflection: Reflection

Data Source

PatentEP3644090B1Driver assistance system and control method thereof
Publication Date: 2022.04.13 HL KLEMOVE CORP
  • EP3644090B1 patent drawingFigure 1
  • EP3644090B1 patent drawingFigure 2
  • EP3644090B1 patent drawingFigure 3

AI summary

A driver assistance system includes an image sensor and a radar mounted to a vehicle and each having a sensing field oriented toward the outside of the vehicle. A controller processes the image data acquired by the camera and the radar data acquired by the radar, detects a stationary obstacle ahead of the vehicle on the basis of the image data or radar data, determines a monitoring range around a location of the stationary obstacle on the basis of the radar data, identifies an object present within the monitoring range on the basis of the radar data, and determines the object as a pedestrian on the basis of a speed of the object toward a road along which the vehicle travels.