Plane-Based Localization for Textureless Indoor Tracking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing localization techniques for mobile devices, such as smartphones, are inefficient and consume high power due to reliance on sensors like IMUs and GPS, which fail to provide accurate and efficient localization in real-time environments, especially indoors, and struggle with environments lacking texture.
Innovation Solution
The method involves detecting and extracting planes from depth data using a two-stage process to correct IMU data, utilizing plane detection and extraction techniques to enhance visual-inertial odometry (VIO) for precise localization, incorporating plane normal vectors and distances to determine spatial transformation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If IMU integration is used for tracking device pose, then localization can be performed, but the estimates diverge over time due to sensor noise and bias
Solution Approach 1:
The system uses plane detection feedback to correct IMU integration drift. Detected planes provide reference information that feeds back into the localization system to reset and correct accumulated errors from IMU noise and bias, enabling long-term accurate tracking.
Solution Approach 2:
Planes serve as an intermediary reference object between the IMU and the environment. By detecting and tracking planes, the system obtains reliable geometric constraints that mediate the localization process, correcting IMU drift without requiring direct environmental features.
2Measurement precision
If point features are extracted from images for visual-inertial odometry, then additional motion information can be obtained, but the scene may contain very few point features due to lack of texture
Solution Approach 1:
The system changes the feature extraction parameter from point features (which require texture) to plane features (which rely on geometric structure). This parameter change enables operation in textureless environments while maintaining motion estimation accuracy through plane geometric constraints.
Solution Approach 2:
The system transitions from extracting 0D point features to extracting 2D plane features from the image data. This dimensional change allows the system to utilize surface geometry and structure information that persists even when texture information is absent, improving environment adaptability.
3Measurement precision
If existing localization techniques are used, then device position can be determined, but the computation is inefficient and power consumption is high
Solution Approach 1:
The system extracts only the essential plane geometric information (normal vectors and distances) needed for localization correction, rather than processing complete point clouds or performing exhaustive feature matching. This extraction approach reduces computational load and power consumption while maintaining localization accuracy.
Solution Approach 2:
The system performs partial processing by focusing computation only on plane detection and parameter extraction from depth data, rather than comprehensive environmental scanning or full 3D reconstruction. This partial action approach achieves sufficient localization accuracy with reduced computational effort and lower power consumption.
Data Source
AI summary
Various implementations disclosed herein include devices, systems, and methods that localize a device based on detecting planes in depth data acquired by the device. For example, an example process may include detecting first plane data in first sensor data acquired by a sensor at a first viewpoint location in a physical environment, detecting second plane data in second sensor data acquired by the sensor at a second viewpoint location in the physical environment, determining that the first plane data and the second plane data correspond to a same plane based on comparing the first plane data with the second plane data, and determining a spatial transformation between the first viewpoint location and the second viewpoint location based on the first plane data and the second plane data.