【摘要】 对于一些昂贵的珍稀木材,如何提高成材率是一个值得深入研究的课题,本文将支持向量机的多分类方法引入到对木材图像的缺陷识别中,通过采样提取木材缺陷数据和相关统计信息,并标记缺陷的类别信息,将这些信息组成一个向量作为支持向量机训练的输入信息,用训练后的支持向量机网络分类木材图像,取得了较好的分类效果。更多还原

【Abstract】 How to improve the utilization of the expensive wood is an interesting research topic that needs deeply investigation.This paper proposes a novel method that introduces the SVM’s multi-classification algorithms into the recognition of the images of wood.First,the deficiency data and its related statistical information can be obtained from sampling and will be labeled into different classes.Then these data will be input into a SVM as training data.At last,the resulting SVM can be used to classify the images of wood.Experimental result shows that the method is satisfactory.更多还原