Occupancy Mapping That Excludes Moving Obstacles in Indoor Navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional approaches to building a map of a physical space for autonomous mobile devices (AMDs) often fail to accurately distinguish between stationary and non-stationary objects, leading to incorrect occupancy maps that can hinder autonomous movement by misclassifying moving objects as stationary, resulting in cluttered maps that obstruct path planning.
Innovation Solution
The use of sensor data from depth sensors and image sensors to classify objects as stationary or non-stationary based on characteristics such as velocity, aspect ratio, and size, with modules like point cloud processing, visual processing, and classification to filter out non-relevant objects and update occupancy maps accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional mapping approaches are used to build occupancy maps, then the map includes all detected objects, but moving objects are misclassified as stationary, resulting in cluttered maps that obstruct path planning
Solution Approach 1:
The system transitions from static object classification to dynamic tracking by continuously monitoring object positions across multiple sensor readings. Objects are classified as stationary or non-stationary based on their motion characteristics over time, allowing the occupancy map to dynamically update its representation of the environment while excluding transient moving objects
Solution Approach 2:
The mapping process is segmented into distinct stages: initial object detection, continuous tracking across multiple readings, motion analysis, and final classification. This segmentation allows the system to separate the detection of all objects from the identification of stationary objects, preventing moving objects from contaminating the occupancy map
2Measurement precision
If multiple sensor readings are processed to distinguish stationary from non-stationary objects, then classification accuracy improves, but computational complexity increases
Solution Approach 1:
The system processes multiple sensor readings (excessive action) to ensure accurate classification of all objects, but applies selective filtering to focus computational resources only on ambiguous cases where motion detection is uncertain. This approach maintains high classification accuracy while avoiding unnecessary computational overhead for clearly stationary or clearly moving objects
Solution Approach 2:
The tracking system uses the motion information naturally captured by sequential sensor readings to automatically distinguish stationary from non-stationary objects. The system's own repeated measurements serve the dual purpose of both detecting object presence and determining object motion state, eliminating the need for separate classification hardware
Data Source
AI summary
A physical space contains stationary objects that do not move over time (e.g., a couch) and may have non-stationary objects that do move over time (e.g., people and pets). An autonomous mobile device (AMD) determines and uses an occupancy map of stationary objects to find a route from one point to another in a physical space. Non-stationary objects are detected and prevented from being incorrectly added to the occupancy map. Point cloud data is processed to determine first candidate objects. Image data is processed to determine second candidate objects. These candidate objects are associated with each other and their characteristics assessed to determine if the candidate objects are stationary or non-stationary. The occupancy map is updated with stationary obstacles. During navigation, the occupancy map may be used for route planning while the non-stationary objects are used for local avoidance.