Motion estimation of lips in pronouncing Korean vowels based on fuzzy constraint line clustering

The fuzzy constraint line clustering (FCLC) method for optical flow estimation is proposed. The proposed method represents the spatial and temporal gradients as fuzzy sets. Based on these sets, several constraint lines with different membership values are generated for the pixel whose velocity is to be estimated. The method estimates the optical flow from the intersection of those weighted constraint lines. The proposed method is tested with mouth image sequences in pronouncing Korean vowels. The result is compared with that of constraint line clustering (CLC). In addition, the statistical distribution of the motion vectors for pronouncing each vowel is shown