基于机器视觉的木材特征提取与树种识别研究综述

我国是木材及木制品加工大国，近年来家具、装修等市场需求的快速增长推动了木材加工行业的发展。由不同树种制作而成的木材材料性质与价值大相径庭，因此准确识别木材树种具有重要意义。相较于传统人工识别，基于机器视觉的木材树种识别大幅度提高了准确率。文中通过分析近5年来木材识别领域的相关文献，总结了木材特征提取的相关技术与树种识别的各种方法，提出要深度融合木材的多个特征并加强各种算法间的配合使用；此外，针对机器视觉在木材树种识别中的应用普遍停留在学术研究阶段的问题，提出木材树种识别应向装备数字化方向发展，以期提高木材树种识别的工作效率。

Research Advance in Wood Feature Extraction and Tree Species Identification Based on Machine Vision

China is a large country in the processing of wood and wood products. In recent years, the wood processing industry has been boosted in its development by the rapid growth of market demand for furniture and decoration. The wood products made of different wood species have large differences in properties and values. Therefore, the accurate identification of wood species is of great significance. Compared with traditional manual identification, the identification of wood species based on machine vision has greatly improved the accuracy. By analyzing the relevant literature in the field of wood recognition in the past five years, the paper summarizes the technologies related to wood feature extraction and various methods of tree species identification, and proposes to deeply integrate multiple features of wood and strengthen the cooperation among various algorithms. It suggests that wood species recognition should proactively develop in the direction of equipment digitalization to address the problem that the application of machine vision in wood species recognition generally stays in the academic research stage, so as to improve the efficiency of wood species identification.