Multi-Tracker Object Tracking Using Sensor Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional object tracking methods on portable computing devices are processor-intensive, leading to high power consumption and latency, and often require frequent image sampling, which can introduce substantial delay and diminish the user experience for real-time applications.
Innovation Solution
Combining lightweight tracking processes that run concurrently to determine the object's location, using sensor data such as accelerometers, gyroscopes, and ambient light sensors to optimize tracking, and selecting or combining results based on heuristics or methods like rule-based, classification-based, or estimation-based approaches.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional object tracking methods are used, then tracking robustness is improved, but power consumption increases and latency increases
Solution Approach 1:
The patent segments the object tracking task into multiple independent lightweight tracking processes, each handling specific aspects of tracking. This allows the system to distribute computational load across multiple simple processes rather than relying on a single complex processor-intensive method, thereby reducing overall power consumption while maintaining tracking robustness through the combination of multiple lightweight trackers.
Solution Approach 2:
The patent combines results from multiple lightweight tracking processes to achieve robust object tracking. By merging the outputs of several simple trackers and selecting or combining their results based on heuristics or classification methods, the system attains the reliability of conventional methods while consuming less power, as each individual lightweight process requires minimal computational resources.
2Reliability
If conventional object tracking methods are used, then tracking robustness is improved, but latency increases
Solution Approach 1:
The patent divides the tracking task into multiple lightweight parallel processes that can execute simultaneously with minimal computational overhead. This segmentation enables real-time processing by avoiding the latency associated with sequential execution of complex tracking algorithms, while the combination of results from these segmented processes maintains tracking robustness.
Solution Approach 2:
The patent employs dynamic selection and combination of results from multiple lightweight tracking processes based on current conditions and heuristics. This dynamic approach allows the system to adaptively choose the most appropriate tracking results in real-time, reducing latency by avoiding fixed complex processing pipelines while maintaining robustness through adaptive result selection.
3Measurement precision
If frequent image sampling is used, then tracking accuracy is improved, but user experience deteriorates due to delay
Solution Approach 1:
The patent implements periodic execution of the lightweight tracking processes at optimized intervals that balance tracking accuracy with real-time responsiveness. By using periodic action rather than continuous heavy processing, the system achieves sufficient tracking precision while minimizing delay, thereby maintaining responsive user experience in real-time applications.
Solution Approach 2:
The lightweight tracking processes are designed to be self-sufficient and efficient, requiring minimal processing resources and executing quickly without needing frequent heavy computational intervention. This self-service capability allows the system to maintain accurate tracking at lower sampling frequencies, reducing delay while preserving tracking precision through the efficiency of each individual tracking process.
Data Source
AI summary
Systems and approaches are provided for tracking an object using multiple tracking processes. By combining multiple lightweight tracking processes, object tracking can be robust, use a limited amount of power, and enable a computing device to respond to input corresponding to the motion of the object in real time. The multiple tracking processes can be run in parallel to determine the position of the object by selecting the results of the best performing tracker under certain heuristics or combining the results of multiple tracking processes in various ways. Further, other sensor data of a computing device can be used to improve the results provided by one or more of the tracking processes.