location: Current position: Home >> Scientific Research >> Paper Publications

Semantic knowledge mining for human motion classification

Hits:

Indexed by:期刊论文

Date of Publication:2012-10-01

Journal:Journal of Information and Computational Science

Included Journals:EI、Scopus

Volume:9

Issue:10

Page Number:2761-2770

ISSN No.:15487741

Abstract:The current methods employed for capturing human motion data are based only on numerical strategies. It has been found that these methods reduce the dimension of sequence as well as the consumption of the calculation. However, these methods can neither be utilized in the field of implicit knowledge acquired from the sequences movement nor meet the processing requirements of a computer. This paper presents a conceptual analysis based on the movement of knowledge organization. For this purpose, the structure of the human body is anglicized, in addition to the physiology and sports science domain knowledge. Further, on the basis of a large number and poorly marked annotation of motion, the HowNet knowledge organization approach is imported for explaining the concept of movement, organization and the network for semantic description of human motion. Finally, a prototype is accomplished in order to illustrate how to establish the semantic network as well as the effectiveness of the model. ? 2012 Binary Information Press.

Pre One:Betterment of attitude estimation based on complementary filter applied in drumstick

Next One:An algorithm for mining high average utility itemsets based on tree structure