Humanoid Skeleton Extraction from Depth Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for processing depth maps struggle to efficiently capture and extract high-level information, particularly in tracking moving humanoid forms, which hinders the development of applications that rely on 3D mapping technology.
Innovation Solution
A method and system that process temporal sequences of depth maps to locate and estimate the dimensions of a humanoid form, reconstructing a 3D skeleton model of the body, allowing for accurate tracking of movements without the need for markers or sensors, by identifying key features like arms, shoulders, and head, and registering with 2D images for face recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing methods process depth maps to identify objects, then object segmentation is achieved, but efficiency in capturing and extracting high-level information deteriorates
Solution Approach 1:
The patent segments the depth map processing into distinct stages: initial object segmentation, skeleton extraction, and hierarchical feature analysis. By dividing the complex processing task into manageable segments, the system achieves both accurate object identification and efficient information extraction, resolving the contradiction between precision and productivity
Solution Approach 2:
The patent extracts skeletons as simplified representations from full depth maps, removing unnecessary detail while preserving essential structural information. This extraction process maintains measurement precision for tracking while significantly improving processing efficiency by working with reduced data complexity
2Reliability
If depth maps are processed to track humanoid movements, then movement tracking capability is improved, but device complexity increases
Solution Approach 1:
The patent creates simplified skeleton copies that represent the essential structure and movement of humanoids without requiring full-depth-map processing. These skeleton models serve as lightweight representations that maintain tracking reliability while reducing the computational complexity of the processing system
Solution Approach 2:
The patent transitions from processing full 3D depth map data to extracting 1D skeletal representations and 2D joint positions. This dimensional reduction allows reliable movement tracking while significantly simplifying the processing requirements, resolving the contradiction between reliability and device complexity
3Measurement precision
If skeletons are extracted from depth maps, then tracking accuracy is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary skeleton extraction and dimension estimation on the first frame before processing subsequent frames. This preliminary action establishes accurate tracking references that can be maintained across multiple frames, improving tracking accuracy while reducing the processing time required for each individual frame
Solution Approach 2:
The patent maintains continuous tracking by updating skeleton positions and dimensions across temporal sequences of depth maps. Once skeletons are extracted, the system continuously tracks their movement with minimal re-processing, preserving tracking accuracy while minimizing additional processing time through efficient frame-to-frame updates
Data Source
AI summary
A method for processing data includes receiving a temporal sequence of depth maps of a scene containing a humanoid form having a head. The depth maps include a matrix of pixels having respective pixel depth values. A digital processor processes at least one of the depth maps so as to find a location of the head and estimates dimensions of the humanoid form based on the location. The processor tracks movements of the humanoid form over the sequence using the estimated dimensions.