Multi-Device Skeleton Recognition via Coordinate Mapping and Synchronization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automatic gait recognition systems using single skeleton recording devices face limitations in field of view and accuracy due to noise and complexity in processing multiple devices, making real-time identification of individuals challenging.
Innovation Solution
A system and method utilizing multiple skeleton recording devices, such as Kinect devices, to capture 3D skeleton models, synchronize data using Network Time Protocol, and map data into a single coordinate system, extracting and processing gait feature vectors to identify individuals through a Support Vector Machine classifier and fuzzy C-means clustering.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If multiple skeleton recording devices are used to expand field of view, then coverage area is improved, but data complexity and noise increase making processing difficult
Solution Approach 1:
The patent merges skeleton data from multiple recording devices by mapping all device data into a single coordinate system defined by a master device. This combines the field of view benefits of multiple devices while unifying the data structure to manage complexity. The synchronization module coordinates multiple devices to work as an integrated system rather than separate independent systems.
Solution Approach 2:
The patent introduces a synchronization module and coordinate system mapping as intermediary components between multiple recording devices and the identification system. This intermediary layer handles the complexity of synchronizing timestamps and transforming coordinates, isolating the core identification algorithm from the complexity of multi-device data integration.
2Measurement precision
If multiple skeleton recording devices are used to capture more data, then identification accuracy is improved, but noise in skeleton data increases resulting in errors
Solution Approach 1:
The patent extracts and removes noise from skeleton data through the synchronization module that filters and validates data from multiple devices. By selectively processing only synchronized, valid data points and excluding out-of-sync or erroneous data, the system maintains high identification accuracy while eliminating the harmful noise that would otherwise accumulate from multiple devices.
3Area of stationary object
If skeleton data from multiple devices is processed, then wider coverage is achieved, but real-time identification becomes difficult due to processing complexity
Solution Approach 1:
The patent performs preliminary actions by pre-establishing a master coordinate system and pre-synchronizing multiple devices before actual identification occurs. Timestamp synchronization and coordinate system definitions are set up in advance, so that during real-time operation, data can be directly mapped and processed without complex runtime transformations, maintaining real-time identification speed while achieving wide coverage.
Data Source
AI summary
Method(s) and system(s) for identification of an unknown person are disclosed. The method includes receiving skeleton data comprises data of multiple skeleton joints of the unknown person from skeleton recording devices. The method further includes extracting G gait feature vectors from the skeleton data. Further, the method includes classifying each gait feature vector into one of N classes based on a training dataset for N known persons and computing a classification score for each class. The method also includes clustering the training dataset into M clusters based on M predefined characteristic attributes of the known persons, tagging each gait feature vector with one of the M clusters based on a distance between a respective gait feature vector and cluster centers of M clusters, and determining a clustering score for each M cluster. The method further includes identifying the unknown person based on clustering scores and classification scores.