Height Map Modification for High Vertical Resolution in Mobile Robot Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing environment map building techniques for mobile robots face challenges in achieving high vertical resolution and robustness against observation noises, particularly for humanoid robots, as they require a large number of grids and significant computational resources, leading to inefficiencies and inaccuracies in detecting obstacles and changes in the environment.
Innovation Solution
The method involves building a three-dimensional environment map using a relative-to-base-plane height map, which modifies height information based on three-dimensional grid occupancy probabilities, allowing for reduced computational load and enhanced robustness against observation noises while maintaining high resolution.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If three-dimensional grids are used to express the surrounding environment with high vertical resolution, then measurement precision is improved, but device complexity increases due to the large number of grids required
Solution Approach 1:
The patent segments the three-dimensional space into a coarse grid structure for general environment representation, while maintaining finer resolution only where needed for specific features. This segmentation allows the system to achieve high vertical resolution for obstacle detection without requiring fine-grained three-dimensional grids throughout the entire space, thereby reducing the total number of grids required.
Solution Approach 2:
The patent transitions from three-dimensional grid representation to two-dimensional grid representation with added height information. By expressing the environment as a two-dimensional map with height values at each grid point rather than using full three-dimensional grids, the system achieves comparable vertical resolution while dramatically reducing the number of grids from millions to thousands.
2Measurement precision
If three-dimensional grids with high vertical resolution are used, then measurement precision is improved, but loss of energy increases due to significant computational resources required
Solution Approach 1:
The patent segments the computational domain by processing only the two-dimensional projection of the environment and calculating height information selectively. This approach divides the computational workload into manageable parts, avoiding the need to process all three-dimensional grid cells, thereby reducing energy consumption while maintaining the ability to detect vertical features with high precision.
Solution Approach 2:
The patent reduces computational complexity by changing from three-dimensional grid processing to two-dimensional grid processing with height attributes. This dimensionality reduction transforms an O(n³) computational problem into an O(n²) problem, significantly decreasing the computational resources and energy required while preserving vertical resolution through the height information stored at each two-dimensional grid point.
3Device complexity
If two-dimensional maps are used to express the surrounding environment, then device complexity is reduced, but measurement precision deteriorates due to loss of vertical information
Solution Approach 1:
The patent applies a two-dimensional grid structure enhanced with height information at each grid point. This approach maintains the simplicity of two-dimensional maps while incorporating vertical dimension data, allowing the system to represent three-dimensional environment features such as stairs and obstacles with high vertical resolution without adopting complex three-dimensional grid structures.
Solution Approach 2:
The patent creates a composite data structure that combines two-dimensional spatial coordinates with one-dimensional height information. This composite representation integrates the advantages of both two-dimensional maps (simplicity) and three-dimensional grids (vertical resolution), resulting in a hybrid structure that is computationally efficient while preserving essential vertical environmental information.
Data Source
AI summary
An environment map building method and an environment map building apparatus can express the surrounding embodiment with a high resolution in the height direction in a manner robust relative to observation noises. The present invention provides an environment map building apparatus for building a three-dimensional environment map to be used for three-dimensionally determining a region allowing a moving body to move therein according to external environment conditions, which comprises an external environment conditions detecting section that detects external environment conditions, a three-dimensional map building section that builds a three-dimensional map showing the state of occupancy of a three-dimensional grid on the basis of the external environment conditions detected by the external environment conditions detecting section, a relative-to-base-plane height map building section that builds a relative-to-base-plane height map showing heights from a two-dimensional base plane on the basis of the external environment conditions detected by the external environment conditions detecting section, and a height map information modifying section that modifies the information of the relative-to-base-plane height map on the basis of the information on the three-dimensional map.