Handwritten Numerical String Recognition Based on SVM Verifier

Handwritten numerical recognition is an important branch of handwritten character recognition, The problem of isolated handwritten numerical recognition has been solved, and however, the segmentation and recognition of handwritten numerical string is a difficult problem in this area. An experiment model about handwritten numerical string recognition applied in this paper, which could deal with the problems such as over segmentation and under segmentation by setting verifier based support vector machine. The experiment results show that this method with verifier based support vector machine reduce the decrease of recognition rate because of inaccurate segmentation greatly.