Container-code pattern recognition based on attribute grid computing

A new container-code pattern recognition algorithm based on attribute grid computing is presented in this paper. The algorithm takes advantage of attribute grid computing, which is a new kind of calculator based on qualitative mapping. In this paper, character feature points are firstly modeled by qualitative criterion attribute grid computing. Then characteristics of each attribute are extracted and the corresponding attribute feature vector is established. Thus, the attribute feature vector can be used to train the model for each container-code character and finally to recognize the characters. By the attribute grid computing, our preliminary experimental results demonstrate an average recognition rate over 97% on hundreds of container-code characters. The results also demonstrate the feasibility of this method.