Traffic Disturbance Detection Using Vehicle Image Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional traffic information systems lack detailed information about roadwork locations, making it difficult for users to accurately identify the start and end points of traffic disturbances.
Innovation Solution
An information processing system that uses vehicles to capture images and detect predetermined objects, such as road cones, while traveling, and calculates the number of detections within a specific time frame to determine the position information of traffic disturbance sites, allowing for precise identification and output of start and end points.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional traffic information is used to display approximate roadwork locations, then the system is simple and easy to operate, but the information detail and accuracy are insufficient
Solution Approach 1:
The system performs preliminary detection by having vehicles capture images and detect objects (road cones, signs) before the traffic information is needed. Position information of traffic disturbance sites is determined in advance by analyzing captured images and comparing with map data, so detailed information is ready when users need it
Solution Approach 2:
The system introduces an intermediary processing layer between image capture and traffic information output. The server acts as an intermediary that receives captured images, performs object detection, determines position information by comparing with map data, and then outputs refined traffic information with precise start and end points
2Loss of information
If detailed traffic information with precise start and end points is provided, then user navigation accuracy is improved, but the processing complexity and time required increase
Solution Approach 1:
The system segments the traffic disturbance site into distinct start and end points by detecting objects at different locations. By dividing the detection process into identifying individual objects and their positions, the system can determine precise boundaries of traffic disturbances without requiring exhaustive analysis of the entire area
Solution Approach 2:
The system performs detection on a partial basis by focusing on key indicator objects (road cones, signs) rather than analyzing every element in the captured images. This selective detection approach provides sufficient information to determine traffic disturbance boundaries without excessive processing time
3Reliability
If multiple vehicles are used to detect and determine traffic disturbance positions, then the reliability and accuracy of traffic information are improved, but the system complexity and data processing requirements increase
Solution Approach 1:
The system merges detection results from multiple vehicles by having them all capture images of the same traffic disturbance site. The server combines these images and their associated position information to determine a unified, reliable position of the traffic disturbance, leveraging multiple observations to improve accuracy
Data Source
AI summary
An information processing system includes a vehicle and a server. The vehicle generates a plurality of captured still images in which scenery outside the vehicle is captured, and acquires position information of the vehicle. The vehicle or the server executes a detection process of detecting a predetermined object in a captured moving image that includes each of the plurality of captured still images as a frame, calculates the number of times that the predetermined object is detected of a section of a predetermined amount of time while changing a position of the section, and determines position information of a candidate start point of a traffic disturbance site on the road based on the position information of the vehicle at a time when at least one captured still image included in the section when the number of times that the predetermined object is detected exceeds a first reference number, is generated.