A novel algorithm for online handwriting Chinese document recognition

In this paper, a novel algorithm for online handwriting Chinese document recognition is proposed, which recognizes Chinese handwriting characters efficiently with the combination of handwriting character features and context information in handwriting document. First, we propose an improved dynamic time warping algorithm for similarity analysis of a handwriting document. Second, we propose an adaptive adjustment algorithm to improve the recognition accuracy by adjusting the handwriting document similarity analysis results. Experimental results on the HIT-OR3C database show that the proposed algorithm improves the recognition performance remarkably.