Point Cloud Obstacle Detection for Non-Planar Travel Surfaces
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous mobile bodies face challenges in detecting non-planar environments, such as slopes and depressions, using conventional two-dimensional sensors, which can lead to obstacles being misidentified and difficulties in navigating complex terrain.
Innovation Solution
The use of a combination of three-dimensional and two-dimensional point clouds to detect obstacle locations, where the three-dimensional point cloud determines the traveling surface and the two-dimensional point cloud provides information corresponding to the traveling direction, allowing for accurate detection of obstacles and terrain features.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a three-dimensional sensor is used to detect non-planar environments, then the detection accuracy of terrain features is improved, but the data amount and calculation cost increase
Solution Approach 1:
The patent segments the three-dimensional point cloud data into multiple two-dimensional cross-sectional views along the traveling direction. By dividing the 3D space into sequential 2D slices, the system processes smaller data portions independently, reducing overall data volume while preserving terrain feature detection capability across the entire 3D space.
Solution Approach 2:
The patent transforms three-dimensional point cloud data into two-dimensional cross-sectional representations by projecting points onto 2D planes perpendicular to the traveling direction. This dimensionality reduction converts complex 3D spatial data into simpler 2D profiles that are easier and less costly to process while maintaining essential terrain information.
2Ease of manufacture
If a two-dimensional sensor is used for detection, then the data processing cost is reduced, but the ability to detect non-planar environments and terrain features deteriorates
Solution Approach 1:
The patent introduces a virtual third dimension by creating multiple 2D cross-sectional views along the traveling direction. Instead of using a complex 3D sensor, the system generates sequential 2D slices from the environment, effectively adding dimensional information through computational processing rather than hardware complexity.
Solution Approach 2:
The patent performs preliminary processing of point cloud data by pre-segmenting the 3D space into 2D cross-sections before obstacle detection. This preprocessing step organizes the spatial data into a structured format that enables accurate terrain feature detection using simpler 2D analysis methods.
3Reliability
If three-dimensional information is accumulated and analyzed, then the detection of non-planar environments is improved, but the calculation complexity and cost increase
Solution Approach 1:
The patent divides the complex task of 3D terrain analysis into multiple independent 2D cross-sectional analyses. Each cross-section is processed separately to identify local terrain features, and results are integrated to form a comprehensive understanding of the non-planar environment, reducing overall computational complexity.
Solution Approach 2:
The patent reduces calculation complexity by transforming 3D point cloud processing into a series of 2D cross-sectional processing tasks. This dimensionality reduction simplifies the computational algorithms required for terrain analysis while maintaining the ability to detect non-planar features through the accumulation of 2D slice information.
Data Source
AI summary
To provide an information processing apparatus, an information processing method, and an information processing program that enable detection of a non-planar environment at low cost. The information processing apparatus according to an embodiment includes: a detection unit that detects an obstacle location that becomes obstruction for traveling in a traveling direction on the basis of a three-dimensional point cloud determined to be a traveling surface and a two-dimensional point cloud corresponding to the traveling direction.