Road Edge Estimation Across Camera Blind Sections
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems fail to accurately estimate the position of a road edge when it is not visible from a vehicle-mounted camera, leading to potential issues in vehicle control and navigation.
Innovation Solution
An apparatus and method that utilize sensor signals and vehicle position data to estimate the road edge by interpolating based on detected edges before and after undetected sections, considering road structure and vehicle motion, and omitting estimation when no edge is present.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If road edge is estimated using only visible sections from camera images, then measurement precision is maintained for visible areas, but reliability deteriorates for undetected sections where road edge position cannot be determined
Solution Approach 1:
The system performs preliminary classification of undetected sections into types (edge-present vs. edge-absent) using trajectory information, vehicle state data, and map information before estimation. This preliminary action enables the system to prepare appropriate estimation strategies in advance, improving reliability for undetected sections without compromising precision through unnecessary estimations
Solution Approach 2:
The system introduces trajectory information, vehicle state information, and map information as intermediary data sources to bridge the gap between visible and undetected road sections. These intermediaries provide contextual clues that enable reliable estimation in undetected sections while maintaining precision through validated inference mechanisms
2Productivity
If interpolation is performed for all undetected sections, then productivity is improved by providing continuous road edge data, but manufacturing precision deteriorates due to erroneous estimations in sections without road edges
Solution Approach 1:
The system segments undetected sections into distinct categories based on the presence or absence of road edges, using trajectory analysis, vehicle state information, and map data. This segmentation enables selective processing where interpolation is applied only to appropriate sections, maintaining precision while improving overall productivity through efficient resource allocation
Solution Approach 2:
The system applies different estimation qualities to different undetected sections based on local characteristics. Sections classified as having road edges receive interpolation treatment with high confidence, while sections without edges are marked as uncertain. This local quality approach ensures precision is maintained where applicable while productivity is improved through selective processing
3Reliability
If multiple data sources (sensor signals, position data, map information) are integrated, then reliability of road edge estimation is improved, but device complexity increases
Solution Approach 1:
The system employs a unified processing framework that handles multiple data sources (sensor signals, position data, map information, trajectory information) through a single integrated algorithm. This multi-functional approach improves reliability by comprehensively utilizing available data while reducing device complexity by avoiding separate processing modules for each data source
Data Source
AI summary
An apparatus for estimating a road edge includes a processor configured to estimate a trajectory of a vehicle, estimate the position of an edge of a road from images generated during travel of the vehicle by a camera, identify an undetected section when an image from which the edge of the road is not detected was generated, determine whether the road in the undetected section has a structure with a road edge, estimate the position of the edge of the road in the undetected section, based on the positions of the edge of the road in the sections in front of and behind the undetected section, when the road in the undetected section has a structure with an edge, and omit to estimate the position of the edge of the road in the undetected section when the road in the undetected section has a structure without an edge.