Recognition of handwritten similar Chinese characters by self-growing probabilistic decision-based neural networks

Abstract

Recognition of similar (confusion) characters is a difficult pattern recognition problem. In this paper, we introduce neural network solution, that is capable of modeling minor differences among similar characters, and is robust to various personal handwriting styles. The Self-growing Probabilistic Decision-based Neural Networks (SPDNN) is a probabilistic type neural networks, which adopts a hierarchical network structures with nonlinear basis functions and a competitive credit-assignment scheme. Based on the SPDNN model, we constructed a three stage recognition system. The prototype system demonstrates a successful utilization of SPDNN to similar handwritten Chinese recognition on the public database CCL/HCCR1 (5401 characters x200 samples). Regarding the performance, experiments on the CCL/HCCR1 database demonstrated a 90.12% of recognition accuracy with no rejection, and 94.11% of accuracy with 6.7% rejection rates respectively.