Online hand-sketched flow chart recognition based on geometric features
ZHANG Xiaoliang, SUN Genzheng, LIAO Daxiong, WANG Shuxia(School of Mechatronics, Northwestern Polytechnical Univ., Xi’an 710072, China)
To reduce the complex operation steps in graphics designing and drawing software, an online hand-sketched graphics recognition method is proposed for flow chart based on geometric features of graphics unit. The approach recognizes user’s sketched graphics through firstly identifying the kinds of lines for inputting strokes, regulating the strokes to geometric lines and recording the attribute information of the strokes. Then user’s sketched graphics are identified by applying geometric features of the strokes and geometric relations between strokes in graphics unit. The whole recognition process is simple and agile with high veracity. And the method doesn’t require the fixed order of sketched strokes and is applicable to different users.