ckwx018计算机视觉和图像处理技术与木材解剖特征量分析.pdfVIP

中国林学会木材科学分会第九次学术研讨会论文集 计算机视觉和图像处理技术与木材解剖特征量分析 保昆雁 范晓宁 (西南林学院木质科学与装饰工程学院,昆明 650224) 要：木材横截面图像 包含了诸如胞壁厚度、胞腔面积及直径、胞壁率、壁腔比等木材解剖特征量，本文 提出了运用计算机视觉系统和图像处理技术，采用显微照相设备获取木材横截面图像并转化为数字信号，采用先进 的计算机硬件与软件技术对图像数字信号进行处理、分析，得到所需要的木材解剖特征量的方法并编制了相应的软 件，基于MATLAB 实现了木材切片图像读入、变换、增强、边缘探测和二值化处理等，在此基础上提取木材横截面切 片图像所反应的木材相关解剖特征量 （包括：细胞数目、胞腔面积及其分布、胞壁率、壁腔比等）。 关键词：计算机视觉 图像处理 木材解剖特征量 边缘探测 插补运算 尺寸标定 Computer Vision and Anatomy Feature of Wood Bao Kun-yan Fan Xiao-ning (Department of Wood Science Technology, Southwest Forestry College, Kunming 650224) Abstract: Anatomy features of wood (such as: thickness of cell wall, area of cell, diameter of cell, wall-cell radio, wall-indiameter radio etc) includes in the image of wood cross section. Approach for extracting them from images of wood cross section was studied and a computer aided image analysis system was programmed with Matlab. A friendly exchange interface and some utilities such as Interpolated Filling and Subtracting Tools and Zoom Tools were included. Some basic data analysis and visualize function was also involved in. Three wood samples include Cunninghamia lanceolata Betula luminifera and Betula luminif era was tested with this 2 system. Basic anatomy features of these three samples are: average cell area - 0.079967, 0.002619, 0.100567mm , thickness of cell wall -0.07364, 0.09693, 0.10351mm, wall-cell radio 0.505087, 0.682087, 0.465153, wall-indiameter radio -0.19409, 0.407383, 0.187736. Keywords: Computer v