Landmark Detection Using Laser Scanner Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting landmarks in traffic environments for precise vehicle localization are resource-intensive, slow, and lack robustness, making them unsuitable for efficient automatic driving functions.
Innovation Solution
A method using a laser scanner to detect data sets, segment them using Euclidean Cluster Extraction Algorithm, and apply principal component analysis to determine landmark parameters, allowing for rapid and efficient detection of landmarks such as lane markings and traffic signs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If comprehensive spatially resolved mapping is performed using point- or raster-based methods, then measurement precision of vehicle position is improved, but device complexity and computational requirements increase significantly
Solution Approach 1:
The patent segments the detection space into discrete detection sectors radiating from the mobile unit, and further segments detected objects into linear segments with specific geometric parameters. This segmentation allows processing of only relevant spatial regions and objects, rather than comprehensive mapping of entire space, reducing computational complexity while maintaining positioning precision through focused analysis of landmark characteristics in each sector.
Solution Approach 2:
The patent transitions from point- or raster-based 2D/3D spatial mapping to a parameter-based representation that describes objects in terms of geometric parameters (distance, angle, segment characteristics). This dimensional transformation from spatial grid to parameter space reduces data volume and computational requirements while preserving the information needed for precise vehicle localization.
2Measurement precision
If comprehensive spatial mapping is performed, then measurement precision is improved, but processing speed deteriorates due to high computational load
Solution Approach 1:
The patent extracts only the essential geometric parameters (distance, angle, segment characteristics) from detected objects, rather than processing complete spatial maps. By taking out only the relevant landmark parameters needed for positioning, the system achieves precise vehicle location without the computational burden of comprehensive spatial mapping, thereby improving processing speed while maintaining precision.
3Measurement precision
If detailed spatial mapping is performed, then measurement precision is improved, but memory requirements increase
Solution Approach 1:
The patent segments detected objects into linear segments characterized by a small set of parameters (start point, end point, distance, angle) rather than storing complete spatial maps. This segmentation reduces memory requirements from storing entire spatial grids to storing only the essential parameters of detected landmarks, while maintaining sufficient precision for vehicle localization through parameter-based representation.
4Reliability
If robust landmark detection is achieved through comprehensive analysis, then reliability is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary segmentation of detected objects into linear segments and pre-calculates their geometric parameters (distance, angle, segment characteristics) before using them for vehicle localization. This preliminary action prepares the data in advance in a parameterized form, ensuring reliable landmark detection while reducing processing time during actual positioning operations, as the heavy computational work is already completed in the segmentation phase.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables precise and rapid landmark detection with reduced computational and memory requirements, facilitating accurate vehicle localization in real-time.
Implementation Method 1
data sets are acquired by a laser scanner, wherein the data sets comprise data points
Data Source
Figure 1A~1B
Figure 2
Figure 3~4B
AI summary
The invention relates to a method for detecting landmarks (13, 14, 15, 16) in a traffic environment of a mobile unit (1), in which a laser scanner (3a) captures data records, the data records comprising data points. Data points of a particular number of data records are stored as output data, and the output data are used to determine segments by means of segmenting, the segments having respective associated data points. For each of the determined segments, a main component analysis is used to determine landmark parameters of the respective segment, and the segments are assigned an object class on the basis of the respective landmark parameters determined for them. Landmark observations are determined, each landmark observation having associated landmark parameters and an associated object class, and the landmark observations are output. The invention further relates to a system for detecting landmarks in a traffic environment of a mobile unit (1).