AGV Camera Localization Using Ground-Touching Corner Edges
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional localization methods for autonomous ground vehicles (AGVs) using perception/vision sensors like LiDAR and cameras are either expensive or lack accuracy, especially when using low-cost cameras, and are not universally applicable across all environments.
Innovation Solution
A method and system that utilizes a camera-mounted AGV to receive a line drawing of the 2D camera scene, determines ground-touching corner edges, and calculates 3D points based on a mapping relationship derived from camera calibration, generating 2D occupancy data to accurately localize the AGV.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If perception/vision sensors like LiDAR and radar are used for localization, then measurement precision is improved, but device cost increases
Solution Approach 1:
The patent replaces expensive perception sensors (LiDAR, radar) with a low-cost camera for localization. The camera captures 2D images that are processed to extract ground-touching corner edges and generate occupancy data, achieving accurate localization without the high cost of traditional sensors.
Solution Approach 2:
The patent substitutes the mechanical/optical measurement systems (LiDAR, radar) with a computational vision system using a standard camera. The localization is achieved through image processing algorithms that detect edges, corners, and occupancy patterns rather than direct distance measurement.
2Ease of manufacture
If a low-cost camera is used for localization, then device cost is reduced, but measurement precision deteriorates
Solution Approach 1:
The patent transforms the 2D camera image into a 3D occupancy map by detecting ground-touching corner edges and calculating spatial relationships. The system infers depth and distance information from 2D image features (edges, corners, lines) to create a structured representation of the environment.
Solution Approach 2:
The patent performs preliminary calibration to establish a mapping relationship between camera image coordinates and real-world coordinates. This pre-established mapping enables accurate distance and position calculation from 2D images without requiring expensive depth-sensing hardware.
3Measurement precision
If conventional perception sensor methodologies are used, then localization accuracy is improved, but adaptability to different environments deteriorates
Solution Approach 1:
The patent creates a universal localization method that works across diverse environments using only a standard camera. The approach detects ground-touching corner edges and generates occupancy data that can be applied to various settings (indoor, outdoor, structured, unstructured) without requiring environment-specific sensors or complex pre-mapping.
Data Source
AI summary
The disclosure relates to method and system for localizing an autonomous ground vehicle (AGV). In an example, the method includes receiving a line drawing corresponding to a two-dimensional (2D) camera scene captured by a camera mounted on the AGV, determining a plurality of ground-touching corner edges based on a plurality of horizontal edges and a plurality of vertical edges in the line drawing, determining a plurality of three-dimensional (3D) points corresponding to a plurality of 2D points in each of the plurality of ground-touching corner edges based on a mapping relationship between an angular orientation of a ground touching edge of an object in real-world and in camera scene and a set of intrinsic parameters of the camera, generating 2D occupancy data by plotting the plurality of 3D points in a 2D plane, and determining a location of the AGV based on the 2D occupancy data and the mapping relationship.