Linear Feature Tracking for Robotic Obstacle Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional object tracking in environments using different object models for each type of identified object is computationally difficult and impractical for many applications, particularly in robotic vehicles where efficient and robust tracking of both static and dynamic obstacles is necessary.
Innovation Solution
The method involves identifying and tracking linear features such as lines and planes within sensor data using a locally adaptive algorithm, which adjusts its parameters based on data distribution, allowing for a single motion model to be used for all objects, and employing probabilistic mapping to persistently track features even if they are not continuously detected.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If different object models are used for each type of identified object, then tracking accuracy is improved, but computational complexity increases significantly
Solution Approach 1:
The patent applies universality by using a single linear feature model and motion model that can represent multiple types of objects (static obstacles, dynamic obstacles, vehicles, pedestrians) uniformly. Instead of creating separate models for each object type, the system uses linear features (lines, planes) as a universal representation that can capture the essential tracking requirements for all objects, thereby reducing computational complexity while maintaining tracking accuracy
Solution Approach 2:
The patent segments objects into linear features (one-dimensional lines or two-dimensional planes) extracted from sensor data, rather than processing complete three-dimensional object models. This segmentation reduces the complexity of object representation while preserving the key geometric properties needed for tracking, allowing efficient computation with simplified geometric primitives
2Loss of information
If conventional object models are used for tracking, then object identification is achieved, but tracking efficiency decreases
Solution Approach 1:
The patent extracts only the essential linear geometric features (lines or planes) from sensor data that are necessary for tracking, rather than processing complete object models. This extraction approach removes unnecessary computational overhead while preserving the key information needed for tracking motion and position, thereby improving tracking efficiency without losing essential object identification capability
Solution Approach 2:
Instead of identifying complete objects first and then tracking them, the patent inverts the approach by directly tracking linear features extracted from sensor data without requiring full object identification. This reversal of the conventional workflow eliminates the computationally intensive object identification step while maintaining effective tracking of moving entities
Data Source
AI summary
A method of tracking objects within an environment comprises acquiring sensor data related to the environment, identifying linear features within the sensor data, and determining a set of tracked linear features using the linear features identified within the sensor data and a previous set of tracked linear features, the set of tracked linear features being used to track objects within the environment.