Lane Marker Localization for Precise Vehicle Lateral Positioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current GPS technology is inaccurate for determining a vehicle's lateral position within a roadway, experiencing significant drift, which is unacceptable for advanced autonomous driving systems, and even when augmented with inertial measurement units, the resolution remains too high for precise self-driving applications.
Innovation Solution
A system that approximates a vehicle's region using GPS or IMU devices, generates a response map from environmental images or sensor data, compares it to a region map, and predicts the vehicle's location based on differences, with the ability to output this information to advanced driver-assistance systems (ADAS) at high frequencies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GPS technology is used to determine vehicle lateral position, then the system is simple and widely available, but the measurement precision is insufficient with drift of 10 meters or more
Solution Approach 1:
The patent combines GPS/IMU positioning data with visual lane marker detection data to create a hybrid localization system. The GPS provides coarse position estimates while visual recognition provides fine-grained lateral position correction, merging two different measurement approaches to achieve both reasonable accuracy and system simplicity.
Solution Approach 2:
The patent introduces an intermediary processing system that takes GPS coordinates and IMU data, correlates them with pre-stored map data containing lane marker positions, and computes corrected lateral position. This intermediary layer transforms low-precision GPS data into high-precision lateral position information by using the map as a reference framework.
2Measurement precision
If IMU is added to GPS to improve positioning accuracy, then measurement precision improves somewhat, but drift remains too high at 1-2 meters for self-driving requirements
Solution Approach 1:
The patent uses pre-stored map data as an intermediary reference that allows the system to correct accumulated drift from IMU integration. By continuously comparing observed lane marker positions against expected positions from map data, the system can detect and correct drift, enabling long-term accurate positioning without requiring extremely precise hardware.
Solution Approach 2:
The patent implements a feedback mechanism where the system continuously compares visual lane marker detection results with predicted positions from GPS/IMU and map data. When discrepancies are detected indicating drift, the system adjusts its position estimates accordingly, creating a closed-loop system that actively corrects errors rather than allowing them to accumulate.
3Measurement precision
If visual lane marker detection is used to achieve 10 cm precision, then measurement precision is sufficient for autonomous driving, but the system becomes more complex requiring image processing infrastructure
Solution Approach 1:
The patent applies partial action by using visual detection only when needed for correction rather than continuously. The system relies primarily on GPS/IMU for position estimation and uses visual lane marker detection selectively to correct lateral position errors, reducing the overall computational burden while maintaining accuracy when precision is most critical.
Solution Approach 2:
The patent segments the localization problem into two distinct components: coarse positioning handled by GPS/IMU and fine-grained lateral correction handled by visual detection. This segmentation allows each subsystem to be optimized independently, with the simple GPS providing framework positioning and the more complex visual system providing only the specific lateral correction function.
4Productivity
If high-frequency position updates are provided to ADAS systems, then the productivity and responsiveness of autonomous driving functions is improved, but the computational load and energy consumption increase
Solution Approach 1:
The patent implements periodic action by updating visual lane marker detection at specific intervals rather than continuously. The system performs visual processing at predetermined frequencies to correct position estimates, allowing computational resources to be periodically engaged rather than continuously consumed, thus reducing overall energy usage while maintaining adequate update rates for ADAS responsiveness.
Data Source
AI summary
Various embodiments of the present disclosure provide a system and method for lane marking localization that may be utilized by autonomous or semi-autonomous vehicles traveling within the lane. In an embodiment, the system comprises a locating device adapted to determine the vehicle's geographic location; a database; a region map; a response map; a camera; and a computer connected to the locating device, database, and camera, wherein the computer is adapted to: receive the region map, wherein the region map corresponds to a specified geographic location; generate the response map by receiving information from the camera, the information relating to the environment in which the vehicle is located; identifying lane markers observed by the camera; and plotting identified lane markers on the response map; compare the response map to the region map; and generate a predicted vehicle location based on the comparison of the response map and the region map.