Building Attribute Domain of Arm Motions for the Representation of Gestural Information

Hand gesture recognition is one of the common sections in human motion analysis. It using camera to track hand movement and interpret into gesture database using image processing. High recognition performance requires every single coordinate projection is properly analyzed to obtain the trajectory of hand gesture. This research aim is to develop a hand gesture recognition system by using ontological approach. Ontology is the framework structure for organizing interconnected complex data model mainly function for information retrieval. In this research, ontology design is divided into three domains which are knowledge domain, attribute domain and process domain. Knowledge domain contains resampled and normalized raw gestural data from motion capture. The attribute domain is the stage where all the features of raw data was presented. However, the current challenge is to expand the variety of attribute in order to obtain higher recognition results where the raw gestural data only consists of $x$ and $y$ coordinate points. This paper has proposed the method to increase the number of attribute by converting the normalized position data into velocity, acceleration, and combination of them. Based on the plotted attribute elements as presented in results, it is practical and applicable to be used in the design of arm gesture recognition systems.