Research on characters extraction algorithm of offline handwritten Chinese character based on procedure neuron networks

To extract the characters of Chinese efficiently and simulating the information of writing process, a method of characters extraction of off-line handwritten Chinese which is based on procedure neuron network is advanced in this paper. The stroke type and location, and the topology of the handwritten Chinese character are analyzed, the stroke writing order is simulated, and a procedure neuron network model which is to get fault-tolerant features of stroke type number, stroke type location, topology type number, the information of stroke order is constructed. A digital vector about the features of handwritten Chinese is established. The simulation result shows that this method can extract the characters of off-line handwritten Chinese efficiently.