Mixed Reality Object Tracking via Pose Transformation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing mixed-reality systems face challenges in effectively tracking and comprehending objects over time and space, particularly in environments with diverse sensory data.
Innovation Solution
A method and system for dynamic object comprehension, utilizing sensors to acquire observations of the environment, transforming sensor poses to track objects across different observations, and updating semantic comprehension based on trackable properties and observing perspectives.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If machine learning methods are applied to train mixed-reality systems to process physical environment data, then object detection and classification capabilities are improved, but computational complexity and processing time increase
Solution Approach 1:
The system performs preliminary actions by pre-processing sensor data to extract key features and characteristics before applying machine learning algorithms. This includes generating initial object hypotheses from raw sensor data, pre-computing spatial relationships, and preparing data structures that facilitate faster and more accurate object detection and classification during runtime.
2Reliability
If sensors continuously sample the physical environment to develop semantic comprehension, then object tracking accuracy is improved, but data processing load and energy consumption increase
Solution Approach 1:
The system implements periodic action by sampling the physical environment at optimized intervals rather than continuously. The sampling rate is dynamically adjusted based on object motion detection, scene changes, and tracking confidence levels. This allows the system to maintain reliable object tracking while significantly reducing sensor activation frequency and associated energy consumption during stable periods.
Solution Approach 2:
The system applies partial action by selectively processing only those sensor data samples that contain meaningful changes or updates to object states. Rather than processing every sensor frame, the system identifies and processes only frames with detected object entries, exits, movements, or attribute changes, reducing overall processing load while maintaining tracking accuracy.
3Loss of information
If the system transforms sensor poses to track objects across different observations, then object comprehension over time and space is improved, but computational requirements and processing complexity increase
Solution Approach 1:
The system introduces an intermediary coordinate transformation framework that mediates between different sensor poses and object representations. This intermediary layer maintains transformed object positions and orientations in a unified reference frame, enabling consistent tracking across multiple sensor observations without requiring complex real-time transformations for every comparison operation.
Data Source
AI summary
In an aspect, the present disclosure provides systems, methods, and computer readable mediums for tracking objects of interest in an environment. A method for tracking objects of interest may include, for example, acquiring, from a sensor, a plurality of observations of the environment, each observation associated with a pose of the sensor for use in transforming between location coordinates in the plurality of observations and location coordinates in the environment; identifying an object at an object location within an observation of the plurality of observation; transforming, based on the pose of the sensors associated with the observation, the location of the object within the observation to an object location in the environment; identifying the object in a different observation of the plurality of observations based on a correspondence between a location of the object within the different observation and the object location in the environment.