Computer Vision SLAM Landmark Outlier Detection Using Position Uncertainty
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing SLAM systems face challenges in accurately distinguishing between landmarks that are outliers and those that are part of the environment, particularly in dynamic environments, which affects the precision and stability of localization and mapping processes.
Innovation Solution
An electronic device and method that determines whether a landmark is an outlier by obtaining and analyzing position and uncertainty information across multiple image frames, using both two-dimensional and three-dimensional coordinate systems, and employing techniques like binocular stereo matching and inertial sensor data to assess reliability and error ranges.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional SLAM systems use simple outlier detection methods, then computational load is reduced, but accuracy in distinguishing outliers from environment landmarks deteriorates
Solution Approach 1:
The patent segments the outlier detection process into multiple independent stages: (1) obtaining landmark position information from image frames, (2) calculating uncertainty information based on position data, (3) comparing uncertainty against threshold values, and (4) making outlier determination. This segmentation allows each stage to be optimized independently, improving overall detection accuracy while managing computational complexity.
Solution Approach 2:
The patent performs preliminary calculations of uncertainty information based on landmark position data before the actual outlier detection decision. By pre-computing uncertainty metrics and comparing them against predetermined thresholds in advance, the system prepares detection criteria that can be quickly applied during runtime, reducing real-time computational burden while maintaining high detection accuracy.
2Measurement precision
If SLAM systems process more image frames and sensor data to improve landmark positioning accuracy, then measurement precision improves, but processing time and computational load increase
Solution Approach 1:
The patent extracts only the essential uncertainty information from the processed image frames and sensor data, rather than analyzing all available data. By calculating uncertainty metrics based on key position parameters and selectively using this extracted information for outlier detection, the system achieves accurate landmark positioning without the computational overhead of processing every detail of the image and sensor data.
Solution Approach 2:
The patent changes the parameter representation from raw image and sensor data to uncertainty metrics that quantify position accuracy. By transforming position data into uncertainty parameters and comparing these against thresholds, the system efficiently evaluates landmark reliability without requiring extensive processing of the original high-dimensional image and sensor data.
3Reliability
If SLAM systems use uncertainty-based outlier detection, then reliability of landmark identification improves, but computational complexity increases
Solution Approach 1:
The patent uses predetermined threshold values for uncertainty that can be set in advance and reused across multiple detection operations. These threshold parameters act as disposable, pre-computed references that improve reliability without requiring complex real-time calculations, as the same thresholds can be applied repeatedly to different landmarks and image frames.
Solution Approach 2:
The uncertainty calculation mechanism is self-service in that it automatically computes reliability metrics from the landmark position data itself. The system uses the position information to derive uncertainty values without requiring external validation or additional complex processing, allowing the data to assess its own reliability.
Data Source
AI summary
An electronic device and method for computer vision are provided. The electronic device for computer vision includes a memory including instructions and a processor configured to execute the instructions, in which the instructions, when performed by the processor, cause the electronic device to determine a first position corresponding to a position of a landmark in a first image frame acquired at a first collection time prior to a second collection time in a second image frame acquired at the second collection time, determine uncertainty information corresponding to the first position based on the determined first position, determine the position of the landmark in the second image frame as a second position, and determine whether the landmark is an outlier based on the first position, the second position, and the uncertainty information.