LIDAR Free Space Detection Using Half-Line Obstacle Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current free space detection systems for vehicles, particularly in autonomous driving and ADAS, face challenges in accurately detecting unoccupied spaces with high reliability due to the need for extensive calculations and the risk of misrecognition, especially when using vision sensors, which can misidentify obstacles and deteriorate in performance compared to LIDAR sensors.
Innovation Solution
A free space detection apparatus and method that utilizes a LIDAR sensor to derive obstacle points and half lines from point data, generating a map for space classification and determining cell states to accurately detect free spaces with reduced calculations and misrecognition, leveraging the higher performance of LIDAR sensors over vision sensors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If vision sensors are used for free space detection, then the system can process visual information, but the computational load increases and misrecognition risk increases
Solution Approach 1:
The patent replaces vision-based computational processing with LIDAR-based direct measurement. Instead of using cameras that require complex image processing, filtering, and depth map calculations, the system uses LIDAR sensors to directly obtain distance information and generate point cloud data, which is then processed through a simplified grid-based obstacle detection algorithm.
Solution Approach 2:
The patent changes the detection parameter from 2D image features to 3D spatial coordinates. By using LIDAR point cloud data with explicit distance measurements (x, y, z coordinates), the system directly represents obstacle positions in three-dimensional space without requiring complex visual feature extraction and depth estimation processes.
2Measurement precision
If vision sensors are used for free space detection, then visual information can be acquired, but misrecognition of obstacles as free space occurs
Solution Approach 1:
The patent transitions from 2D image processing to 3D spatial detection. By using LIDAR point cloud data that provides explicit three-dimensional coordinates (x, y, z), the system can accurately represent obstacle positions and dimensions in space, eliminating the ambiguity inherent in 2D image interpretation where objects difficult to discriminate from the road may be misrecognized as free space.
3Productivity
If radar is used for free space detection, then detection can be performed, but sensor performance and information resolution deteriorate
Solution Approach 1:
The patent uses LIDAR to create a detailed point cloud representation (copy) of the three-dimensional environment. This point cloud data accurately captures the spatial distribution of obstacles with high resolution, providing both the detection capability of radar and the measurement precision of optical sensors.
Data Source
AI summary
A free space detection apparatus for a vehicle is provided to detect a free space based on a lidar sensor and recognize a driving environment. The apparatus includes a point data processor that removes noise of point data acquired from the LIDAR sensor and derives a half line with respect to an obstacle point from which noise has been removed based on coordinates of the obstacle point. A map processor generates a map for space classification, checks whether the obstacle point or the half line with respect to the obstacle point is present in a cell of the map and determines cell states of the map. A free space detector detects a free space based on the cell states of the map.