| Summary: | In the discussion of computer vision, detection and recognition are an interesting topic to discuss. Basically, advanced computer vision technology requires a high-level interaction method above the text-based console interaction. Hand detection and gesture recognition is one of the interaction cases in computer vision. In this study, an experiment of hand detection and syntactic hand gesture recognition method are discussed. HSV (Hue Saturation Value) space color model is used as the basis of hand detection and segmentation. Then, the thinning method is used to get endpoint features of each fingertip.
The proposed design is designed to meet with real-time video processing. The experiment intended to find some issues usually happened when the ZS thinning method is used to gain the detection and recognition. The result shows that the proposed design able to detect and recognize some gesture, but unstable hand movement may lead into a fault called by extra endpoint. In this research, extra endpoints are considered as a challenge that must be anticipated when using thinning method especially ZS algorithm to perform syntactic hand gesture recognition.
|
|---|