Intersection Lane Connection Mapping Using Aerial and Telemetry Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems for detecting road and lane connections at intersections rely on on-vehicle modules, which limits their effectiveness and accuracy, especially in scenarios where such modules are unavailable or malfunctioning.
Innovation Solution
A method utilizing aerial imagery data and vehicle telemetry data, combined with deep learning models like GANs and machine learning classifiers, to detect road edges and classify maneuvers, constructing probabilistic finite state automata for lane pairing, and employing homotopy models for accurate lane edge determination, without relying on on-vehicle modules.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If on-vehicle modules are used for detecting road and lane connections, then the system can collect real-time data from vehicles, but the system becomes dependent on on-vehicle modules which may be unavailable or malfunctioning
Solution Approach 1:
The patent introduces an intermediary system consisting of aerial imagery data and server-based processing that mediates between vehicles and the road mapping system. Instead of directly relying on on-vehicle modules, the system uses aerial images as an intermediate data source that can be processed to extract road and lane connection information, thereby reducing dependency on vehicle-mounted sensors while maintaining detection capability
Solution Approach 2:
The patent creates a copy of the road and lane connection information by generating synthetic aerial imagery through rendering engines. This synthetic data serves as a backup or alternative source when real aerial imagery is unavailable, allowing the system to maintain operational reliability without depending on physical on-vehicle modules
2Measurement precision
If aerial imagery data and multiple data sources are processed using deep learning models, then detection accuracy improves, but computational complexity and processing time increase
Solution Approach 1:
The patent segments the complex detection task into multiple independent processing stages: aerial imagery acquisition, vehicle telemetry data collection, deep learning-based road edge detection, maneuver classification, and lane connection inference. Each stage processes specific data types independently, reducing overall system complexity while maintaining high detection accuracy through specialized processing at each step
Solution Approach 2:
The patent introduces an intermediary server that acts as a mediator between data collection and complex analysis. The server receives raw aerial imagery and telemetry data, performs preliminary processing and filtering, then passes refined data to deep learning models. This intermediary layer reduces the computational burden on individual components while enabling accurate detection through coordinated processing
3Measurement precision
If a hybrid approach combining aerial imagery and vehicle telemetry data is used, then lane connection detection accuracy improves, but data integration complexity increases
Solution Approach 1:
The patent merges aerial imagery data and vehicle telemetry data into a unified analysis framework. Aerial imagery provides top-down spatial information about road geometry, while vehicle telemetry provides ground-level validation of actual vehicle paths. By combining these complementary data sources through coordinated processing algorithms, the system achieves higher detection accuracy than either source alone while managing integration complexity through structured data fusion protocols
Data Source
AI summary
A method for detecting road edges at a predetermined intersection, comprising: receiving, by the controller, aerial imagery data about the predetermined intersection; receiving, by the controller, vehicle telemetry data from at least one vehicle passing through the predetermined intersection; detecting, using the aerial imagery data and generative adversarial networks (GANs) executed on the controller, road edges at the predetermined intersection; classifying, using the vehicle telemetry data and a random forest classifier (RFC) executed on the controller, each vehicle trajectory passing through the predetermined intersection with a label corresponding to a unique maneuver to create a maneuver labeling at the predetermined intersection; constructing, using the maneuver labeling determined by the RFC and the road edges, a probabilistic finite state automata (PFSA) to pair inbound lanes with outbound lanes at the predetermined intersection; and determining lane edges at the predetermined intersection using a homotopy model.