一种自适应的木材表面缺陷图像分割方法

An adaptive segmentation method of wood surface defect image

木材表面缺陷图像的分割是木材表面图像处理中的基础性难题,针对这一问题提出了一种基于测地线活动轮廓模型的木材表面缺陷图像的分割算法。采用图像梯度信息对此模型进行了改进,使其更适合对木材表面缺陷的分割。实验表明:改进的测地线活动轮廓模型算法结合了图像梯度信息,对死节、虫眼、边缺及树脂纹几种常见木材表面缺陷图像的分割取得了比较理想的分割效果,相对于传统测地线活动轮廓模型有所提高。

Wood surface defect image segmentation is a basic problem in image processing. An algorithm for segmentation of wood surface defect images is proposed based on GGAC model, which is an effective image segmentation model. Image gradient information is used to improve the model, which makes it more suitable for the segmentation of wood surface defects. The dead-knot, worm holes, edge-missing and resin pattern of wood surface defects image segmentation, the improved GGAC model algorithm a- chieves good segmentation results combined with the gradient information of the image. GGAC model is a kind of improved model compared with the traditional model.