Motion Capture Camera for Synchronized Marker and Scene Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional motion capture systems face suboptimal asynchronous capture of scene data and marker data, which hampers precise correspondence between marker position data and scene images, especially in applications requiring high fidelity.
Innovation Solution
A motion capture system with integrated marker tracking and removal subsystems processes digital image data onboard cameras to generate synchronized marker data and scene images, allowing for high-fidelity training data generation for markerless object tracking in machine learning systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If separate cameras are used for marker tracking and scene recording, then the signal-to-noise ratio for marker detection is improved, but the temporal synchronization between marker data and scene images deteriorates
Solution Approach 1:
The patent combines marker tracking and scene recording capabilities into a single motion capture camera system. The image sensor captures both scene light and marker light simultaneously, and the processor separates marker data from scene data through image processing algorithms, achieving both high marker detection precision and temporal synchronization.
Solution Approach 2:
The motion capture camera is designed with multi-functionality, serving both as a scene recorder and a marker tracker using the same hardware components. The image sensor and processor handle dual functions, eliminating the need for separate specialized cameras and ensuring synchronized data capture.
2Measurement precision
If marker tracking optical filter is used to enhance marker light signal, then marker detection accuracy is improved, but the ability to record normal scene visibility is reduced
Solution Approach 1:
The patent segments the captured light signal into different wavelength components using optical filters. The image sensor captures multiple wavelength bands simultaneously, allowing separation of marker light (enhanced by filters) from scene light (captured in visible spectrum), achieving both high marker accuracy and scene visibility.
Solution Approach 2:
The system adds a wavelength dimension to the capture process by using multiple optical filters for different wavelength ranges. This allows the same camera to extract marker information from one wavelength band while preserving scene information from other wavelength bands, resolving the trade-off between marker accuracy and scene visibility.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system provides accurate, synchronized marker data and scene images, enhancing the training of AI systems for precise object tracking and reducing processing burdens by processing data locally, thereby improving the fidelity and efficiency of markerless motion capture.
Implementation Method 1
an image sensor that is operable to generate a series of frames of digital image data representing a scene that is visible to the motion capture camera
Implementation Method 2
a plurality of reflective or light-emitting markers attached to a moving subject in the scene
Implementation Method 3
a plurality of reflective or light-emitting markers attached to a moving subject in the scene
Data Source
AI summary
Motion capture systems and methods involve processing a series frames of digital video image data on-camera to determine the position of markers attached to a moving subject in the scene. Compressed video and corresponding marker position data or object model data are transmitted by each camera while preserving correspondence or synchronization information between each frame of compressed video and the corresponding marker data or object model data. Each frame of the digital image data may be altered on-camera, before compression and transmission, to paint out the markers in the scene before the series of frames of digital image data, so altered, are encoded by a compression algorithm. The encoded and compressed video data and the corresponding marker data sets, or object data based thereon, may be utilized to train machine learning systems or other AI systems for markerless motion capture.