先验采样约束结合扩展字典的细化稀疏识别算法

为解决可能存在遮挡环境下的模式识别问题，提出先验采样约束结合扩展遮挡字典的细化稀疏识别算法。针对训练样本无法包含测试样本遮挡变化的情况，首先需要构造遮挡字典（墨镜、围巾等），进而利用先验局部采样子模块稀疏表示分类原理判断可能存在的测试样本遮挡模式；然后对未被遮挡的局部子模块利用Borda计数投票方式，依据每类残差大小分配给不同的票数，计算样本类别信息；其次根据遮挡模式结果，利用全局稀疏表示通过构造样本遮挡扩展字典对测试样本进行全局分类投票；最后将两次分类投票结果进行融合，最终实现是否存在遮挡环境下的精细模式判别。实验结果表明，本文算法不仅能够给出准确的模式类别，还能给出遮挡类别信息，可得到精细化识别结果。 

Detailed Sparse Recognition Based on Priori Sampling Constraints and Extended Dictionary

To solve the problem of pattern recognition with occlusion, we propose a detailed sparse recognition algorithm based on priori sampling constraints and extended dictionary. Considering the problem that the training images can not span the occlusion variation under testing conditions, the possible occlusion dictionary (sunglasses, scarves ) is constructed at first, then priori local sampling sub-modular sparse classify is used to estimate the possible occlusion of testing images; After that, according to the residual size, we exploit borda count to assign different votes for local sub-modular which is unobstructed. Thus, the sample information can be obtained; Secondly on the basis of the result of occlusion pattern, global sparse representation are performed to assign votes for the testing images by constructing the sample occlusion extended dictionary; Finally ,two classification results are fused to accomplish pattern discrimination in the environment of occlusion. Experimental results show that the proposed method not only can give an accurate pattern category, but also obtain the information of occlusion category which is a detailed recognition result.