Large-Model Road Map Construction for Accurate Lane Attributes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing road map construction methods for autonomous vehicles suffer from low lane accuracy and inefficient updates, making it difficult to accurately represent actual lane attributes and maintain a timely representation of road conditions.
Innovation Solution
A method and apparatus using a large model to process target prompt information and images collected by onboard sensors, enhancing lane detection accuracy by incorporating associated-region lane attributes and constructing regional road maps with improved detection and update efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional road map construction methods are used, then the system is simpler, but lane accuracy is low and update efficiency is poor
Solution Approach 1:
The patent introduces a large model as an intermediary component that processes prompt information and image data to generate accurate lane attributes. This intermediary system enables high-precision lane detection without requiring complex traditional processing pipelines, thereby improving measurement precision while managing system complexity through modular integration.
Solution Approach 2:
The patent replaces traditional mechanical/map-based lane detection systems with an AI-driven large model system. This substitution eliminates the need for complex traditional algorithms and hardware configurations, achieving superior lane accuracy through computational intelligence rather than mechanical processing systems.
2Productivity
If traditional road map construction methods are used, then the system is simpler, but update efficiency is low
Solution Approach 1:
The patent implements continuous map updating through the large model system that continuously processes new image data and prompt information. This continuous action enables real-time or near-real-time map updates, significantly improving productivity while the modular architecture keeps system complexity manageable through ongoing optimization rather than complete system redesign.
3Measurement precision
If associated-region lane attributes are incorporated, then lane detection accuracy improves, but processing time increases
Solution Approach 1:
The patent performs preliminary processing by pre-processing image data and prompt information before feeding them to the large model. This preliminary action includes preparing associated-region lane attributes in advance, which reduces the actual processing time when the large model executes, thereby achieving high accuracy while minimizing time loss through efficient data preparation.
Data Source
AI summary
A method of constructing a map based on a large model, a vehicle control method, an electronic device, and a storage medium are provided, which relate to a field of artificial intelligence technology, in particular to fields of computer vision, deep learning, large model and generative model technologies. The method includes: acquiring an associated-region lane attribute and an image to be detected, which is collected by an onboard sensor and represents a road region to be detected; constructing a target prompt information based on the associated-region lane attribute; and processing the target prompt information and the image to be detected by using the large model to obtain a regional road map for the road region to be detected. The associated-region lane attribute corresponds to an associated road region, and the associated road region and the road region to be detected meet a predetermined similarity condition.