원문정보
초록
영어
We present a method of recognizing hand gestures using RGB values, depth value, hand center coordinates, and finger counting data from Microsoft’s Kinect for implementing the immersive hand interface to overcome inconvenience using HMD. First, we set the RGB values and depth range to detect the hand. This processing can improve recognition rate. Then, through double labeling, outside labeling, and inside labeling, we detect the hand without noise. Then, we use the distance vector to obtain the hand center. It also removes everything except the hand area, including removal of the wrist. After detection of the hand, we use HNMA (Multi Information-Hippocampus Neuron Modeling Algorithm) to recognize the hand gesture. This algorithm helps to improve the recognize rate. It is difficult to use the interface when using an HMD (Head Mount Display) display machine. This algorithm can make an immersive environment.
목차
1. Introduction
2. Background Theory
2.1. Hand Detection Based on Kinect
2.2. Hippocampus Neuron Modeling Theory
3. Proposed Algorithm
3.1. Hand Detection Processing
3.2. Learning Algorithm Setting
3.3. Experimental Results Disseminate
4. Conclusion and Future Research
Acknowledgment
References