Pedestrian Behavior Classification for Collision Avoidance
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing pedestrian protection systems face challenges in accurately predicting pedestrian movements across roadways, leading to potential erroneous triggering of emergency braking or evasive maneuvers, which can result in either collisions or unnecessary evasive actions.
Innovation Solution
A method and device that classify pedestrian behavior by analyzing sensor signals and surroundings information using physical variables such as velocity, acceleration, and spatial relationships, allowing for a context-dependent prediction of pedestrian movement, and activating passenger protection devices only when necessary to avoid collisions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If emergency braking or evasive maneuvers are initiated based on pedestrian movement prediction, then collision avoidance is improved, but erroneous triggering increases leading to unnecessary evasive actions
Solution Approach 1:
The system performs preliminary classification of pedestrian behavior into categories (cooperative, uncertain, uncooperative) before initiating emergency maneuvers. This preliminary assessment uses multiple physical variables (lateral position, longitudinal velocity, acceleration) to predict future pedestrian trajectories, allowing the system to distinguish between pedestrians who will actually cross the vehicle's path and those who won't, thereby reducing erroneous triggering while maintaining collision avoidance capability
Solution Approach 2:
The system changes the parameters used for decision-making from simple distance-based thresholds to multiple physical variables including lateral distance to roadway edge, longitudinal velocity, acceleration, and predicted trajectory. By monitoring changes in these parameters over time and comparing them against classified behavior patterns, the system achieves more accurate prediction of pedestrian intent, resolving the contradiction between reliable collision avoidance and precise behavior classification
2Measurement precision
If conservative assumptions about pedestrian movement are used, then erroneous trigger rate decreases, but the system may miss some actual collision risks
Solution Approach 1:
The system dynamically adjusts its assessment based on the classified pedestrian behavior type. For cooperative pedestrians, more conservative assumptions are applied with higher thresholds for triggering emergency maneuvers. For uncooperative pedestrians, the system adopts more aggressive assumptions with lower thresholds, predicting that these pedestrians are more likely to continue their crossing path. This dynamic adaptation allows the system to maintain high classification accuracy while preserving collision detection reliability across different pedestrian types
Data Source
AI summary
A method for classifying a behavior, of a pedestrian when crossing a roadway of a vehicle, includes reading in a sensor signal to detect the pedestrian and at least one piece of surroundings information regarding surroundings of the pedestrian. The sensor signal represents here a signal of at least one sensor of the vehicle. The method further includes ascertaining at least one physical variable of a correlation between the pedestrian and the at least one piece of surroundings information. Finally, the method includes classifying the behavior of the pedestrian using the at least one physical variable.