Aiming at assisting the document inspectors to identify the seal imprint time, a classification identification method is proposed based on BP NN (Neutral Network). This paper built a seal stamp characteristics index system aided by the recognition experience of famous inspectors, then, extracted and quantized the characteristic parameters. With these parameters, building BP NN training and testing models could help to identify the real range of seal imprint time. The experiment was implemented on the extracted seal characteristics data by using the Matlab Neural Network toolbox. Experiment results demonstrate the effectiveness and feasibility of the seal imprint time classification verification method based on BP NN.

No abstract