基于非负稀疏编码和RBPNN的掌纹图像识别方法

The Method of Palmprint Recognition Based on Non-negative Sparse Coding and Radial Basis Probabilistic Neural Network

主要探讨了一种新颖的基于非负稀疏编码(NNSC)和径向基概率神经网络(RBPNN)模型的掌纹图像识别方法。使用NNSC算法可以成功地提取掌纹图像的特征,利用RBPNN模型可以有效、快速地实现掌纹图像的分类。与RBFNN和BPNN模型相比,实验结果表明RBPNN模型具有更高的识别率和更好的分类能力。

The paper mainly discusses a novel palmprint recognition method based on Non-negative Sparse Coding （NNSC） and Radial Basis Probabilistic Neural Network （RBFNN）. Palmprint features can be extracted successfully by using the algorithm of NNSC, and the classification task can be implemented efficiently and fast by the RBPNN model proposed. Moreover, compared with the classification methods of Radial Basis Function Neural Network（RBFNN） and BPNN, experimental results also show that the RBPNN achieves higher recognition rate and better classification efficiency.