【摘要】 与稀疏表示(SR)模型相比,基于单个核函数的SR(KSR)模型可以有效减少数据维数、降低学习模型的计算复杂度并提高特征分类精度;但这种模型对核函数及其参数的选择通常不能包含恰当的、完整的分类信息。为了满足更高的特征分类精度需求,提出了一种基于多个核函数的SR(M-KSR)模型及其快速稀疏优化方法,并将其应用于掌纹图像的分类。测试结果证明了基于M-KSR模型的掌纹分类方法的有效性和实用性。更多还原

【Abstract】 Compared with sparse representation( SR) model,signal kernel function based SR( KSR) model,can reduce efficiently the number of dimensions of data and computational complexity of this learning model,as also as improve the feature classification accuracy. However,the proposed model usually doesn’t contain proper and complete classification information for the selection of kernel functions and their corresponding parameters,therefore,in order to meet the needs of higher feature classification accuracy, a multiple kernel function based KSR( M-KSR) model and its fast sparse optimization method are proposed here,and the proposed model is used to classify palmprint images. Test results prove that the M-KSR model based palmprint classification method is very efficient and applicable in practice.更多还原