Offline Handwritten Numeral Recognition Based on Principal Component Analysis
Wan Junli, Yuehua Huang, Guohua Zhang, Cheng Wan · 2007
To overcome the difficulty of fusing statistical feature and structural feature in the research on handwritten numeral recognition, Principal Component Analysis is used to reconstruct numeral model and estimate the numeral reconstructive error based on the statistical information of digit structural feature. At the same time, the height-width ratio and Euler value of numeral is extracted. Recognition of the digit character is completed through combining the neural network and Bayes classifier respectively corresponding to the three type features. The recognition rate of this method is 90.73% on handwritten numeral database.