Curb Edge Detection Using Optical Lines and Altitude Jumps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Automated systems, such as self-driving vehicles and robots, face challenges in safely navigating environments with obstacles like curbs that can damage wheels, requiring effective methods to identify curb edges for safe operation.
Innovation Solution
A method involving optical imaging and topographic profiling using calibrated cameras and grids to detect curb edges by analyzing altitude jumps and applying edge filter kernels, allowing for efficient curb edge identification without line triangulation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional methods (e.g., line triangulation) are used to detect curb edges, then measurement precision may be adequate, but device complexity and computational burden increase
Solution Approach 1:
The patent extracts and utilizes only the essential features for curb detection: linear structures from optical images and altitude information from topographic profiles. By projecting linear structures into the topographic profile and examining altitude jumps only at these projected locations, the method eliminates the need for complex full-scene analysis while maintaining detection accuracy.
Solution Approach 2:
The detection process is segmented into distinct steps: (1) detecting linear structures in optical images, (2) projecting these structures into the topographic profile, (3) examining altitude jumps at projected locations, and (4) connecting verified altitude jumps to identify curb edges. This segmentation reduces computational complexity by focusing processing only on relevant regions.
2Reliability
If comprehensive environmental scanning is performed to ensure safety, then detection reliability improves, but processing time increases
Solution Approach 1:
The method performs preliminary detection of linear structures in the optical image before analyzing the topographic profile. This preliminary action allows the system to pre-identify potential curb locations, so that subsequent altitude jump examination is restricted only to these relevant regions rather than scanning the entire environment.
Solution Approach 2:
The patent combines two-dimensional linear structure detection from optical images with one-dimensional altitude analysis from topographic profiles. By projecting the 2D linear structures into the 1D topographic profile and examining altitude jumps at these projected locations, the method efficiently verifies curb edges without requiring comprehensive 3D environmental scanning.
Data Source
AI summary
A method for identifying edges of curbs in an environment of a mobile platform. The method includes: providing at least one optical image of the environment of the mobile platform; determining linear structures in the at least one optical image; providing a topographic profile of a ground surface of the environment of the mobile platform; projecting the linear structures into the topographic profile of the ground surface; verifying the edges of curbs by examining, on both sides of the projected linear structures, the topographic profile for altitude jumps corresponding to a curb structure; connecting the verified altitude jumps in order to identify the edges of the curbs.