论文部分内容阅读
现有的应力波木材检测仪只能测定木材内部是否存在缺陷,无法对木材缺陷类型进行分类。笔者提出了一种结合应力波无损检测技术和支持向量机(SVM)的木材缺陷识别分类方法,该方法首先测量木材内部的应力波传播速度,以此作为分类特征,利用支持向量机对木材的内部缺陷进行分类。为了验证该方法的有效性,选取健康的以及含有不同缺陷的山核桃木试样31件、松木试样28件,采集山核桃木试样应力波传播速度数据117组、松木试样应力波传播速度数据80组,以应力波传播速度为分类特征,利用支持向量机对木材的缺陷类型进行分类。结果表明:山核桃木试样缺陷分类准确率达到93.75%,松木试样缺陷分类准确率达到95%。该方法不仅能识别木材内部是否存在缺陷,还能对木材的空洞、裂缝、腐朽等缺陷进行准确分类。
The existing stress wave wood detector can only determine the existence of defects within the wood, timber types of defects can not be classified. In this paper, we propose a classification method of wood defect recognition based on non-destructive testing of stress wave and support vector machine (SVM). The method first measures the propagation speed of stress wave in wood as a classification feature and uses support vector machine Internal defects are classified. In order to verify the effectiveness of the method, 31 samples of hickory wood with different defects and 28 samples of pine wood were collected, 117 samples of stress wave propagation velocity of hickory wood samples were collected, stress wave propagation of pine samples 80 sets of velocity data, the stress wave propagation speed as a classification feature, the use of support vector machine classification of wood defects. The results showed that the classification accuracy of pecan tree samples was 93.75%, and that of pine samples was 95%. The method can not only identify the existence of defects within the wood, but also on the wood hollow, cracks, decay and other defects accurately classified.