HD Map Lane Detection Using Sensor Quality Weighting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing lane detection systems for autonomous vehicles face reliability issues due to GPS errors and erroneous lane recognition caused by the use of single cameras, which affects the accuracy of autonomous driving.
Innovation Solution
A system that utilizes surrounding information sensors, HD map data, and a processor to calculate matching results for candidate lanes by considering sensing quality and HD map data validity, incorporating multiple cameras for improved lane detection and reducing errors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If GPS information and single camera are used for lane detection, then the system complexity is low, but the measurement precision and reliability of lane detection deteriorate
Solution Approach 1:
The patent combines multiple sensors (front camera, front lateral cameras, HD map data, GPS) into an integrated lane detection system. The processor fuses data from these multiple sources to determine candidate lanes and calculate matching results, thereby improving measurement precision while managing system complexity through unified processing.
Solution Approach 2:
The system uses multiple image sensors that can serve both general surround-view functions and specific lane detection functions. The front camera and front lateral cameras not only capture overall vehicle surroundings but also specifically detect lane lines, enabling multi-functional use of the sensor system.
2Reliability
If multiple sensors and HD map data are integrated for lane detection, then the measurement precision and reliability improve, but the device complexity increases
Solution Approach 1:
The lane detection process is segmented into distinct functional modules: candidate lane determination based on GPS and map data, image acquisition from multiple cameras, matching result calculation between detected lines and map data, and final lane determination. This segmentation allows complex multi-sensor processing to be managed through structured, modular operations.
Solution Approach 2:
The system calculates matching results by comparing lane lines detected from multiple cameras with lane lines from HD map data for each candidate lane. This feedback mechanism validates candidate lanes against multiple data sources, improving reliability by cross-verifying detections before final lane determination.
3Measurement precision
If matching results are calculated considering sensing quality and HD map validity, then the measurement precision improves, but the calculation complexity increases
Solution Approach 1:
The system changes the parameters used in matching calculations by incorporating sensing quality metrics and HD map data validity indicators. The matching result calculation considers not only geometric alignment but also quality weights and validity flags, adjusting the parameter set to include reliability metrics that improve precision.
Solution Approach 2:
The system performs preliminary validation of HD map data and sensing quality assessment before final matching calculations. By pre-evaluating data quality and validity, the system prepares weighted parameters in advance, reducing the computational burden during the actual matching process while maintaining high precision.
Data Source
AI summary
A system for detecting a host vehicle driving lane includes a surrounding information sensor that obtains surrounding information of a host vehicle, a memory that stores HD map data, and a processor that detects the host vehicle driving lane, where the processor receives the surrounding information and the HD map data, outputs matching results based on matching between the surrounding information and the HD map data for respective candidate locations on a plurality of candidate lanes, outputs respective accumulative results for the respective candidate locations according to the matching results and quality information of the surrounding information sensor, and determines the host vehicle driving lane among the plurality of candidate lanes according to the accumulative results.