Hand Gesture Recognition Using Depth Map and 3D Point Cloud Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current hand gesture recognition technologies lack practical methods for accurately recognizing hand gestures, particularly in vision-based systems.
Innovation Solution
A method and device that acquire a depth map of a hand, estimate joint positions, create a 3D point cloud, match the joints with a stored 3D hand model to obtain refined positions and degree of freedom parameters, and recognize gestures using a trained image-hand gesture mapper, which includes neural networks or random forest classifiers, with iterative updates for improved accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If vision-based hand gesture recognition is implemented, then the technology is leading in the field, but it lacks any practical hand gesture recognition method
Solution Approach 1:
The patent segments the hand gesture recognition process into distinct modules: depth map acquisition, 3D point cloud generation, joint position estimation, and gesture classification. This segmentation transforms the abstract vision-based approach into concrete, implementable steps that can be executed sequentially, resolving the contradiction between leading technology status and lack of practical method.
Solution Approach 2:
The patent performs preliminary actions by first acquiring depth maps and generating 3D point clouds before gesture recognition. It also pre-trains image-hand gesture mappers using sample data, and updates shape parameters in advance. These preliminary steps establish a solid foundation for accurate recognition, making the overall system practical and reliable.
2Measurement precision
If depth map acquisition and 3D point cloud creation are performed, then joint position estimation accuracy is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary actions by first acquiring depth maps and generating 3D point clouds before gesture recognition. It also pre-trains image-hand gesture mappers using sample data, and updates shape parameters in advance. These preliminary steps establish a solid foundation for accurate recognition, making the overall system practical and reliable.
Solution Approach 2:
The patent creates a 3D point cloud as a digital copy of the hand from the depth map, and uses an image-hand gesture mapper that has been trained with sample data copies. This copying approach allows the system to work with simplified representations rather than raw data, improving processing efficiency while maintaining accuracy.
3Measurement precision
If the image-hand gesture mapper is updated with shape parameters, then recognition accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent implements feedback by updating the image-hand gesture mapper and shape parameters based on the matching results between estimated joint positions and 3D point clouds. The system continuously refines its model by comparing predictions with actual data, improving accuracy over time while managing complexity through iterative optimization.
Solution Approach 2:
The patent changes parameters by optimizing shape parameters (bone lengths, joint positions) and updating the image-hand gesture mapper based on matching results. This parameter optimization approach improves recognition accuracy by adapting the model to specific hand geometries while maintaining computational tractability through focused parameter adjustment.
Data Source
AI summary
A method and device for use in hand gesture recognition is applicable to image processing. The method includes: acquiring a depth map of a hand in a current image; estimating first positions of joints of the hand according to the depth map of the hand; creating a 3D point cloud of the hand according to the depth map of the hand; matching the first position of the joints of the hand and a stored 3D hand model to the 3D point cloud of the hand to obtain second positions of the joints and first degree of freedom parameters of the joints; and recognizing the hand's gestures according to the second positions of the joints and the first degree of freedom parameters of the joints. The method achieves a practical hand gesture recognition technique and recognizes hand gestures accurately.