Mobile Robot Obstacle Detection Using 2D Point Cloud Density
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing obstacle detection systems for mobile robots inaccurately identify elongated objects due to erroneous removal of point group data in three-dimensional measurements, leading to missed obstacle detection.
Innovation Solution
An obstacle detection apparatus that plots point clouds from sensors on a plane, extracts point clouds with a predetermined density, and computes obstacle positions, concentrating potential noise components to prevent erroneous noise determination, allowing for accurate obstacle detection while improving computational speed by removing noise from a two-dimensional plane rather than three-dimensional space.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If point clouds are processed in three-dimensional space to remove noise components, then noise removal can be performed, but point clouds of elongated objects may be erroneously removed and detection accuracy decreases
Solution Approach 1:
The patent transforms point cloud data from three-dimensional space to two-dimensional plane representation. This dimensional reduction concentrates scattered point clouds of elongated objects into denser regions on the plane, preventing their erroneous removal as noise while maintaining the ability to filter actual noise components through density-based extraction
2Productivity
If point clouds are processed in three-dimensional space, then complete spatial information is preserved, but computational speed decreases
Solution Approach 1:
The patent projects three-dimensional point clouds onto a two-dimensional plane, reducing computational complexity and increasing processing speed. The projection maintains essential spatial relationships needed for obstacle detection while eliminating the computational burden of full three-dimensional processing
3Reliability
If point clouds with low density are removed as noise, then noise filtering is achieved, but point clouds of elongated objects may be erroneously removed
Solution Approach 1:
By transforming to two-dimensional plane representation, the patent concentrates point clouds of elongated objects into denser regions. This concentration effect ensures that even objects with inherently low point cloud density in 3D space appear as dense regions on the 2D plane, preventing their erroneous removal as noise while maintaining effective noise filtering capability
Data Source
AI summary
An obstacle detection apparatus configured to compute a position of an obstacle from point clouds acquired by one or more sensors, the obstacle detection apparatus performing the processing of: plotting point clouds acquired by one or more sensors on a plane; extracting, from the point clouds plotted on the plane, a point cloud whose degree of density is equal to or larger than a predetermined degree of density; and computing a position of an obstacle from the extracted point cloud.