Hand Posture Classification via Skeleton Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Controller-free interactive systems, such as gaming systems, struggle to detect and distinguish subtle hand gestures like open and closed hands due to limitations in motion estimation routines, which affect the accuracy and intuitiveness of user interaction.
Innovation Solution
The system receives a depth image from a capture device, estimates skeleton information, segments the hand region, extracts a shape descriptor, and classifies the hand state using training data to determine the posture of the hand, enabling more refined and intuitive interactions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If skeleton estimation is used to detect user motions, then major joints can be estimated, but subtle gestures cannot be detected
Solution Approach 1:
The patent segments the hand region from the depth image by identifying pixels within a bounding box defined by skeleton joint positions (wrist and finger tip). This segmentation isolates the hand region for detailed analysis, enabling detection of subtle gestures that would be invisible in full-body skeleton estimation.
Solution Approach 2:
The patent transitions from 3D skeleton joint coordinates to 2D shape descriptor space by computing normalized pixel coordinates and distance metrics within the hand region. This dimensional transformation enables classification of hand states (open/closed) based on shape characteristics rather than relying solely on 3D joint positions.
2Adaptability or versatility
If traditional motion estimation is used, then system complexity remains low, but interactivity and user experience are limited
Solution Approach 1:
The patent performs preliminary actions by first estimating skeleton information to locate hand regions, then segmenting those regions before extracting shape descriptors and classifying hand states. This multi-stage preliminary processing enables refined gesture detection while maintaining system modularity and manageable complexity.
Data Source
AI summary
Systems and methods for estimating a posture of a body part of a user are disclosed. In one disclosed embodiment, an image is received from a sensor, where the image includes at least a portion of an image of the user including the body part. The skeleton information of the user is estimated from the image, a region of the image corresponding to the body part is identified at least partially based on the skeleton information, and a shape descriptor is extracted for the region and the shape descriptor is classified based on training data to estimate the posture of the body part.