中文摘要：

提出了一种基于断层图像的木材腐朽程度自动评价方法。首先利用Graph Cuts算法对木材断层图像进行分割后,对所得图像进行灰度化处理。然后利用累加直方图技术获取木材断层图像中缺陷部分面积。最后,将测得的木材缺陷面积值与真实值进行比较,从而实现腐朽程度的自动评价。分别采用了含人工模拟腐朽的木材样本和含自然腐朽的木材样本进行实验,结果表明：本文提出的方法具有较高的检测精度,为木材断层图像智能处理和无损检测技术的在线应用提供了新的方法。

英文摘要：

This paper presents an automatic evaluation method for the wood decaying degree based on tomographic images. Firstly, the wood tomographic image was partitioned by Graph Cuts, and conversed to grayscale. Then, the cumulative histogram technology was used to obtain the defect part area of the grayseale image. Finally, the measured values were compared with the real values, and the automatic evaluation for the wood decaying degree was implemented. The wood samples containing artificial decay or nature decay were used in our experiments. The experimental results show that the accuracy of the method we proposed is high, and this study provides a new method for the intelligent processing of wood tomographic images and the online application of nondestructive detection technology.