木材表面缺陷特征轮廓提取算法研究

虫眼、活节和死节是最常见的木材表面缺陷，是木材分选过程主要的识别目标，精确提取木材表面缺陷轮廓特征能大幅提高木材分选的准确率。本研究提出一种针对木材表面虫眼、活节、死节缺陷轮廓提取方法。针对木材表面常见黑点和纹理等非线性噪声，使用中值滤波方法平滑图像。然后分别应用OTSU算法与全局阈值分割算法分离图像背景与目标，对结果二值图像使用数学形态学方法进行滤除和填充，最终用sobel算子提取缺陷边缘。结果表明，采用OTSU算法分割和数学形态学相结合的方法可以很好地提取木材表面缺陷特征，用sobel算子能够提取到比较完整、准确、连续的木材表面缺陷边缘轮廓，提高了目标图像的可视性和精准性。

Research on Feature Extraction Algorithm of Wood Surface Defects

Worm holes,live and dead knots are the most common wood surface defects and they are the main identification targets in the wood sorting process.Accurately extracting the contour features of wood surface defects can greatly improve the accuracy of wood sorting.In this paper,a method for the contour extraction of worm hole,live and dead knots on wood surface was presented.The median filter method was used to smooth the image for common nonlinear noise such as black spot and texture on wood surface.Then OTSU algorithm and global threshold segmentation algorithm were used respectively to separate the image background and target,then mathematical morphology was adopted to filter and fill the binary image,and finally Sobel operator was applied to extract the defect edge.The experimental results showed that the combination of OTSU algorithm segmentation and mathematical morphology could extract the wood surface defect features well,and Sobel operator could extract relatively complete,accurate and continuous wood surface defect edge contour,improving the visibility and accuracy of the target image.