Vehicle Warning Control Using Pedestrian Trajectory Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing pedestrian detection systems often transmit unnecessary warnings to vehicles by incorrectly identifying individuals outside their path, such as those in gardens or houses, due to limitations in tracking movement trajectories and distinguishing between relevant and irrelevant targets.
Innovation Solution
A warning system that utilizes a vehicle-mounted camera and MEC server to process image recognition data, differentiate between pedestrians on and off the road, and selectively transmit alerts based on trajectory information from nearby terminals, ensuring accurate warning information is sent only to relevant targets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If pedestrian detection systems transmit warnings to vehicles based on detected pedestrians, then safety warnings are provided to vehicles, but unnecessary warnings are transmitted to vehicles when pedestrians are outside the vehicle's path
Solution Approach 1:
The system performs preliminary actions by storing movement trajectories of pedestrians in advance and predicting future locations before actual pedestrian-vehicle interaction occurs. This allows the system to pre-determine which pedestrians are relevant to the vehicle's path, enabling selective warning transmission only to vehicles that may actually be affected, thereby reducing false alarms while maintaining safety warnings.
2Loss of information
If the system tracks all detected pedestrians, then complete pedestrian information is available, but computational resources and data processing requirements increase
Solution Approach 1:
The system extracts only the essential and relevant information from the complete pedestrian dataset by identifying and processing only those pedestrians whose trajectories indicate potential interaction with the vehicle. By separating relevant pedestrians (those on or near the vehicle's path) from irrelevant ones (those clearly outside the path), the system reduces computational complexity while maintaining tracking completeness for safety-critical cases.
Data Source
AI summary
A control device includes: an acquisition unit configured to acquire information indicating a location of a warning target, and the number of the warning targets, the warning target being recognized from an image captured by an image capture device mounted on a moving body; a reception control unit configured to perform a control to receive, from a plurality of external terminals existing near the location of the warning target acquired by the acquisition unit, trajectory information indicating a past movement trajectory of each of the plurality of external terminals; a selection unit configured to select, from among the plurality of external terminals, one or more external terminals that are transmission targets of warning information based on the past movement trajectory; and a transmission control unit configured to perform a control to transmit the warning information to the one or more external terminals selected by the selection unit.