Image–Depth SLAM Co-Location Through Transformation Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing SLAM systems face challenges in colocation when using different sensors due to differences in data types and formats, making it difficult to merge three-dimensional maps effectively.
Innovation Solution
A method and system for co-locating SLAM systems using an image sensor and a depth sensor by scanning an anchor to obtain feature points and coordinates, converting data formats, and computing a transformation matrix using algorithms like RANSAC or ICP to align the systems in a common coordinate system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If different sensors (image sensor and depth sensor) are used in SLAM systems, then the system can capture more diverse environmental information, but it becomes difficult to merge three-dimensional maps due to differences in data types and formats
Solution Approach 1:
The patent introduces a coordinate system transformation mechanism as an intermediary between image sensor data and depth sensor data. By converting depth information into image coordinate systems and using feature point matching as a mediator, the system bridges the gap between different sensor data formats, enabling seamless integration without requiring complex custom processing for each sensor type combination.
Solution Approach 2:
The patent transforms depth sensor data parameters (shooting directions and depths) into image coordinate system parameters (mappoint coordinates). This parameter transformation allows depth information to be expressed in the same coordinate framework as image data, enabling direct comparison and integration of features from both sensors without maintaining separate complex data structures.
2Measurement precision
If feature points from image sensor and sampling points from depth sensor are directly matched, then colocation can be achieved, but the large number of sampling points creates computational burden
Solution Approach 1:
The patent segments the large set of depth sensor sampling points by grouping them according to their corresponding image feature points. Instead of processing all sampling points individually, the system divides them into groups associated with each feature point, then selects representative points from each group. This segmentation dramatically reduces the number of points requiring detailed processing while preserving colocation accuracy.
Solution Approach 2:
The patent applies partial action by selecting only a subset of sampling points (fake mappoints) that are most representative of each feature point, rather than processing all sampling points. This selective approach performs sufficient processing to achieve accurate colocation without the excessive computational burden of processing every single sampling point, optimizing the balance between accuracy and efficiency.
Data Source
AI summary
Disclosed are a method and a system for co-locating simultaneous localization and mapping (SLAM) systems, adaptable for a first SLAM system using an image sensor and a second SLAM system using a depth sensor. The method includes: scanning an anchor in a space by the image sensor to obtain an image of the anchor and capture feature points of the image to create mappoints in the space; scanning the anchor by the depth sensor to obtain shooting directions and depths of sampling points on the anchor; converting the same into coordinates in the space, and downsampling the sampling points to create the fake mappoints based on a position of the depth sensor in the space; and computing a transformation matrix between three dimensional maps of the first and second SLAM systems. The transformation matrix is configured to perform colocation of the first and second SLAM systems.