特征提取是手寫(xiě)體數(shù)字識(shí)別研究中的重要問(wèn)題，有效、穩(wěn)定的特征是提高識(shí)別率和識(shí)別精度的關(guān)鍵。該文提出了一種基于分?jǐn)?shù)本征特征和核非線(xiàn)性分類(lèi)器的手寫(xiě)數(shù)字識(shí)別方法，首先找到時(shí)頻平面的一個(gè)軸進(jìn)行分?jǐn)?shù)傅里葉變換，使不同類(lèi)別樣本在這個(gè)軸上最大限度地分開(kāi)，然后用主元分析進(jìn)行降維，得到比較穩(wěn)健的低維特征，再將常用分類(lèi)器用于特征分類(lèi)，實(shí)現(xiàn)對(duì)手寫(xiě)數(shù)字的識(shí)別。對(duì)實(shí)際數(shù)據(jù)進(jìn)行實(shí)驗(yàn)，結(jié)果表明上述本征特征與核非線(xiàn)性分類(lèi)器相結(jié)合有較高的識(shí)別率和訓(xùn)練、分類(lèi)效率。Feature extraction is an important part in handwritten numeral recognition. Efficient and robust feature is a key to improving recognition rate and efficiency. This paper adopts fractional Fourier transform and principal component analysis to extract robust and compact features called fractional eigenafeatures. In classification, five kernel-based nonlinear classifiers, Parzen and robust Parzen classifiers, radial basis function classifier, support vector classifier, and kernel-based nonlinear representor are applied and compared. Experimental results show the effects and efficiency of the proposed algorithm.