Sketched symbol recognition using Latent-Dynamic Conditional Random Fields and distance-based clustering

摘要: In this paper we propose a two-stage method for recognizing sketched symbols that combine the use of discriminative model, labeling symbol strokes and distance-based clustering algorithm, grouping labels belonging to same symbol. first stage, employ Latent-Dynamic Conditional Random Field (LDCRF), model able analyze features unsegmented sequences by taking into account spatio-temporal information, classify parts considering contextual information. second obtained from LDCRF are grouped using algorithm which takes geometric relationships among strokes. The effectiveness our has been evaluated on domain electric circuit diagrams achieving accuracy values varying between 81.3% 91.0%. HighlightsA two stage methodology recognition is proposed.The approach combines graphical algorithm.The applied recognition.A complete evaluation in provided.The achieved performance compared with baseline system.