Lane-Traffic Light Mapping at Intersections With Low Memory Use
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing road information generation systems require storing captured images to associate lanes with traffic lights, leading to significant memory usage and potential errors in traffic light recognition, especially at intersections with multiple lanes and oblique traffic lights.
Innovation Solution
A road information generation apparatus that includes a camera and a microprocessor to recognize the position of the vehicle and traffic lights before entering an intersection, using detection data to generate road information by associating the traffic light with the travel lane, thereby reducing memory usage and improving recognition accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If captured images are accumulated to generate road information, then road information can be generated, but memory capacity is greatly used
Solution Approach 1:
The patent extracts only the essential information elements (traffic light position, lane information, intersection data) from captured images at the moment of passage, rather than storing the entire images. This extraction approach generates necessary road information while significantly reducing memory capacity requirements.
Solution Approach 2:
The system performs preliminary recognition and extraction of road information elements when the vehicle passes through intersections. By capturing and processing only the critical data at these specific moments rather than continuously storing images, the system generates comprehensive road information with minimal memory usage.
2Loss of information
If captured images are used to recognize traffic lights, then traffic light information can be obtained, but recognition accuracy decreases at intersections with multiple lanes and oblique traffic lights
Solution Approach 1:
The patent introduces a coordinate transformation mechanism that converts oblique traffic light positions into standardized coordinate systems relative to the vehicle's travel direction. This dimensional transformation allows accurate recognition of traffic lights regardless of their angular position at intersections with multiple lanes.
Solution Approach 2:
The system uses lane information and intersection geometry as intermediary data to bridge the gap between captured traffic light images and their actual positions. By introducing these intermediate reference frames, the system accurately determines traffic light locations even in complex multi-lane intersection scenarios.
Data Source
AI summary
A road information generation apparatus includes: an in-vehicle detection unit configured to detects a situation around a subject vehicle; and a microprocessor and a memory connected to the microprocessor. The microprocessor is configured to perform: recognizing a position of the subject vehicle; recognizing a travel lane of the subject vehicle and a traffic light corresponding to the travel lane of the subject vehicle, installed at an intersection, before the subject vehicle enters the intersection, based on a detection data detected by the in-vehicle detection unit and the position of the subject vehicle recognized in the recognizing, when the intersection is recognized ahead in a traveling direction of the subject vehicle, based on the detection data detected by the in-vehicle detection unit; and generating a road information associating the traffic light recognized in the recognizing the traffic light with the travel lane of the subject vehicle.