Equirectangular Projection for SLAM Drift Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current SLAM technologies face limitations in field of view (FOV), execution efficiency, and accuracy of localization, particularly with single lens cameras and multiple lenses, which can result in drift errors and reduced map accuracy.
Innovation Solution
A method and device for processing panoramic maps using equirectangular projection, capturing and transforming images to extract feature points, establishing 3D sparse point maps, and optimizing camera pose, which improves feature point extraction efficiency and stability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of moving object
If multiple lenses are used to expand field of view, then the field of view is improved, but drift errors increase due to image distortion and deformation
Solution Approach 1:
The patent introduces an equirectangular projection as an intermediary transformation layer between the multiple camera lenses and the localization system. This projection method uniformly maps images from multiple lenses onto a common spherical coordinate system, establishing a consistent reference frame that eliminates drift errors while preserving the expanded field of view coverage.
Solution Approach 2:
The patent segments the panoramic image processing into distinct equirectangular projection steps for each camera lens, transforming individual lens images into a unified spherical coordinate system. This segmentation allows each lens to be processed independently while maintaining global consistency, preventing drift accumulation.
2Measurement precision
If entire map image is processed by LSD-SLAM algorithm, then localization is achieved, but memory consumption increases
Solution Approach 1:
The patent extracts only the essential feature points from the equirectangular projected panoramic images for localization processing, rather than processing the entire map image. This extraction approach significantly reduces memory consumption while maintaining localization accuracy through the use of distinctive and informative feature points.
Solution Approach 2:
The patent transforms the 2D map image processing into a 3D spherical coordinate system through equirectangular projection. This dimensional transformation allows the system to represent the environment more efficiently, reducing the data volume required for processing while preserving spatial relationships for accurate localization.
3Device complexity
If single lens camera is used, then device complexity is reduced, but field of view is limited
Solution Approach 1:
The patent makes the equirectangular projection system universal by providing a unified processing framework that can accommodate any number of camera lenses. The same projection and feature extraction algorithms work consistently across single lens or multiple lens configurations, allowing the system to scale from simple to complex setups without requiring different processing pipelines.
Data Source
AI summary
A method of processing a panoramic map based on an equirectangular projection is provided. The method comprises the following steps. A first and a second equirectangular projected panoramic images are captured through at least one lens at different time, respectively. The first and the second equirectangular projected panoramic images are perspectively transformed based on at least one horizontal angle, respectively. A plurality of first and second feature points are extracted from the first and the second equirectangular projected panoramic images, respectively. A plurality of identical feature points in the first and the second feature points are tracked. A camera pose is obtained based on the identical feature points. A plurality of 3D sparse point maps, in binary format, are established based on the at least one horizontal angle. The camera pose and the 3D sparse point maps are exported to an external system through an export channel.