基于近义词分配的铁路扣件状态检测

摘要

针对传统“视觉词包模型”在进行底层特征编码时存在较大的量化误差的问题,提出一种基于近义词分配的铁路扣件状态检测模型.首先,利用潜在狄利克雷分布模型分析得到语义主题在某一视觉单词上的概率分布,并引入相对熵衡量视觉单词间的语义距离,从而获取语义相关的近义词;然后,在“视觉词包模型”的基础上,结合柔性分配策略将底层特征映射至若干近义词上;最后,利用支持向量机实现扣件检测.对4类扣件图像的分类实验证明该模型能够有效提高扣件分类精确度.%Aiming at the problem of large quantization error in traditional Bag of words (BOW) model when performing the low-level feature coding,a model of railway fastener detection based on homoionym-assignment is proposed.Firstly,the Latent Dirichlet Allocation (LDA) model is used to analyze and obtain the latent topic distribution induced by the visual words.Secondly,the relative entropy is introduced to measure semantic distance between visual words and obtain semantically related words.And then,soft-assignment is adopted to realize the mapping of low-level features on some homoionym.Finally,the Support Vector Machine (SVM) is applied to fulfill fastener inspection.The experiment on four types of fasteners shows that the proposed model can improve effectively the accuracy of fastener classification.