About this document
Hand Posture Recognition Method by alaminmaxpro is a document available to read on EtoBox.
The paper presents a novel method for hand posture recognition that addresses the limitations of existing techniques reliant on color-based segmentation. It employs a four-step process including optical flow for movement detection, contour preprocessing, a contour signature for characterization, and a classification method utilizing principal component analysis and Mahalanobis distance. The proposed approach enhances recognition performance while maintaining low computational complexity, making it suitable
- Author
- alaminmaxpro
- Language
- EN