Abstract
A new method to quickly recognize hand posture from complex backgrounds based on depth image and appearance feature was proposed. First, hand posture region was quickly extracted from complex background via depth image. Then, appearance features were integrated to build the decision tree for hand posture recognition. Eight common postures under the complex background were tested in our experiments. The experimental results show that the recognition rate is 98.2 % and speed rate achieves 25 frames per second.
| Original language | English |
|---|---|
| Pages (from-to) | 88-91 |
| Number of pages | 4 |
| Journal | Huazhong Keji Daxue Xuebao (Ziran Kexue Ban)/Journal of Huazhong University of Science and Technology (Natural Science Edition) |
| Volume | 39 |
| Issue number | SUPPL. 2 |
| Publication status | Published - Nov 2011 |
| Externally published | Yes |
Keywords
- Complex background
- Depth image
- Hand posture
- Human-robot interaction
- Image recognition