基于HSI空间的形态学单板缺陷检测  

Veneer Defects Detection Based on Mathematical Morphology in Space HIS

作　　者：解朦 戴天虹[1] 李琳[1] 

机构地区：[1]东北林业大学机电工程学院,哈尔滨150040

出　　处：《森林工程》2014年第2期65-67,70,共4页Forest Engineering

基　　金：黑龙江省自然科学基金(F200920)

摘　　要：单板材料在发展中国家的应用日趋广泛,优质的单板可用于模板,胶合板,贴面板等人造版的面板。为了检测并提取单板图像中的缺陷,提高单板质量,基于数学形态学,本文提出一种复合型的彩色图像边缘检测方法。在传统形态学算法的基础上,一是利用开闭运算的迭代提高了抗噪能力;二是利用双结构元分别对H、S、I三个分量进行形态学处理而后融合。与传统算法Canny和Sobel进行比较,结果表明该算法增强了图像分割的精确性和完整性,能够有效提高单板等级。The application of veneer material in developing countries is increasing. High-quality veneer can be used for template, plywood and other man-made version of the panel. In order to detect and extract the defects of veneer image and improve the quality of veneer, this paper presents a composite method based on mathematical morphology color image edge detection. On the basis of tradi- tional morphological algorithm, the opening and closing operation of iterations are used to improve the noise immunity, and double structure is used processing H, S, and I three components and then fusion. Compare with traditional algorithms canny, sobel, the experiment results show that the algorithm can enhance the accuracy and completeness of image segmentation, which can effectively improve the board level.

关 键 词：单板 彩色图像 HIS空间 数学形态学 缺陷检测 

分 类 号：S781.5[农业科学—木材科学与技术]