An Online Algorithm for 3D Signal Recognition Based on Ethesis Evaluation
Chenhui Yang · Computer Engineering and Science · 2009
Nowadays,there are so many kinds of sensors which have been widely used in intelligent systems.The signals of these sensors are recognized by the traditional pattern recognition methods,which need huge training sample sets or experts' knowledge.In the article,we propose a new method which is mostly based on ethesis evaluation.It obtains original signals,does segmentation and recognition of the signals and then recognizes the action what the signal means.In the stage of signal recognition,it uses comparison and similarity of two curves.We turn the curves to weight graphs which we call simple linear graphs.Using this structure,we construct an evaluation function for comparison.Our method uses the space topological information and very little samples instead of huge training sets or experts' knowledge.Moreover,it is an online algorithm,which benefits many real-time applications.Finally,we present the results of our experiment.