Laser Scanner Feature Point Extraction for Obstacle Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Laser scanner data alone is insufficient for accurately classifying obstacles like vehicles and pedestrians due to its limited provision of angle and distance information, making it difficult to distinguish between different types of obstacles.
Innovation Solution
A multi-layer laser scanner system that extracts feature points from laser scanner data by separating it into layers, restoring three-dimensional coordinates, calculating standard deviations, and determining gradients, allowing for more accurate classification of obstacles based on pre-stored feature points.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If laser scanner data is used for obstacle recognition, then distance and angle information can be obtained, but the data is insufficient to accurately classify different types of obstacles
Solution Approach 1:
The patent segments laser scanner data into multiple layers based on distance ranges, extracting feature points from each layer independently. This segmentation allows the system to analyze different spatial regions separately, improving obstacle classification by capturing layered structural information that would be lost in holistic analysis.
Solution Approach 2:
The patent transforms two-dimensional laser scanner measurements (angle and distance) into three-dimensional spatial understanding by organizing data across multiple distance layers. This dimensional transformation enables the system to infer obstacle type from spatial distribution patterns across layers, compensating for the inherent information limitations of laser scanner data.
2Measurement precision
If multi-layer laser scanner data is processed to extract feature points, then obstacle classification accuracy improves, but computational complexity increases
Solution Approach 1:
The patent extracts only the most relevant feature points from laser scanner data within each layer, rather than processing all measurement data. By identifying and extracting key feature points that characterize obstacle structures, the system reduces computational load while maintaining classification accuracy.
Solution Approach 2:
The patent performs preliminary organization of laser scanner data into distance-based layers before feature extraction. This pre-processing step structures the data in a way that facilitates efficient feature point identification and reduces the complexity of subsequent analysis by eliminating the need for complex spatial calculations.
Data Source
AI summary
An apparatus and method for extracting a feature point to recognize an obstacle using a laser scanner are provided. The apparatus includes a laser scanner that is installed at a front of a traveling vehicle and is configured to obtain laser scanner data having a plurality of layers in real time. In addition, a controller is configured to separate the laser scanner data obtained by the laser scanner into a plurality of layers to extract measurement data present in each layer and determine feature points of the measurement data to classify a type of obstacle based on a plurality of stored feature points.