Classified Segmentation of Online Complex Freehand Sketching

Online freehand sketching is an efficient way for exchanging design idea at the conceptual design stage of products and a precondition for human-computer interface. It is one of the important factors for checking human-computer interactive performance that the recognition rate of online freehand sketching. The excessively long recognition time will cause a breakdown in online recognition. Based on convex-concave characteristic of online complex freehand sketching, this paper classifies stroke into two parts depending on whether it is non-ultra compound primitivity or ultra compound primitivity, and educes the corresponding segmentation algorithm. This paper produces a dense fold-line processing method for ultra compound primitivity. This algorithm effectively avoids the question of excessively long time of complex freehand sketching recognition. The experiment results show the method is effective and builds a foundation for human-computer interactive sketching input.