Route Deviation Recognition Using Road-Segment Trees

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current route-deviation recognition methods in navigation systems are slow in scenarios with parallel roads and have a high error rate in complex intersections, as they rely on tracing back to an ingress node to determine route deviations.

Innovation Solution

A route-deviation recognition method that defines a candidate region around the current position to generate multiple road-segment tree structures, eliminating the need to trace back to an ingress node, thereby reducing response time and error rates in parallel road scenarios and avoiding incorrect recognition in complex intersections.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the current method of tracing back to an ingress node is used to determine route deviations, then the navigation system can identify route deviations, but the response time is slow in parallel road scenarios and the error rate is high in complex intersections

Engineering Contradiction:
Improveroute-deviation recognition accuracyVSAvoidroute-deviation recognition response time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent segments the road network into hierarchical levels (first-level road segments directly connected to current position, second-level road segments connected to first-level segments, etc.) and processes them in order of proximity. This segmentation allows the system to focus computation on relevant local areas rather than tracing back to distant ingress nodes, reducing response time while maintaining accuracy in parallel road and complex intersection scenarios

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent pre-establishes the hierarchical road segment structure and connection relationships in advance, so that during real-time navigation, the system can directly query and compare pre-processed road segment data without performing complex tracing operations. This preliminary structuring enables fast route-deviation detection by allowing direct comparison between current position and preorganized road segment hierarch

Inventive Principle:
Principle #10Preliminary action

2Productivity

If the ingress node tracing method is used, then route deviations can be detected, but the error rate increases in complex intersections and parallel road scenarios

Engineering Contradiction:
Improveroute-deviation detection efficiencyVSAvoidroute-deviation recognition accuracy
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent applies different processing strategies to different local road configurations. For parallel roads, it uses the hierarchical road segment structure to identify and select the correct road segment based on spatial proximity and connectivity. For complex intersections, it leverages the pre-established hierarchical relationships to systematically evaluate multiple possible paths. This localized adaptation to different road scenarios improves both accuracy and efficiency

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The system continuously compares the current position with the planned route using the hierarchical road segment structure, providing real-time feedback on route adherence. When a deviation is detected, the system can immediately identify the specific road segment where deviation occurred and recalculate the route, creating a closed-loop feedback mechanism that improves reliability through continuous verification and correction

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP3435035B1Route-deviation recognition method, terminal and storage medium
Publication Date: 2023.08.16 TENCENT TECHNOLOGY (SHENZHEN) CO LTD
  • EP3435035B1 patent drawingFigure 1~2
  • EP3435035B1 patent drawingFigure 3~4
  • EP3435035B1 patent drawingFigure 5~6

AI summary

Route-deviation recognition method, terminal, and storage medium are provided. The method includes: obtaining a current positioned-location; determining a candidate region covering the positioned-location; obtaining road network data corresponding to the candidate region to generate a plurality of road-segment tree structures according to the obtained road network data; obtaining a matching weight between the positioned-location and each road-segment branch in each of the road-segment tree structures; determining, according to the matching weight, a road-segment branch matching the positioned-location as a matched road-segment branch; and when the matched road-segment branch does not belong to a planned navigation route, determining that route-deviation occurs.