| Wood plays an important role in people’s production and life. With the development of economy, people also gradually have an increased demand for wood. But trees growth cycle is long, and people use tress unreasonably, intensify the contradiction between supply and demand of timber. How to use and save timber resources has become a pressing problem that the scientific researchers are facing. Using computer tomography (CT) technology for nondestructive testing of wood, can quickly and accurately judge the wood within all kinds of information, and find out defects that affect the properties of wood. Reasonable material selection and utilization, not only ease the contradiction between supply and demand, but also effectively protect resources, which is conductive to the sustainable development.This paper studies the common method of wood defects nondestructive testing, which is the principle and system of computer tomography (CT) technology. Getting wood defects of wood defects in CT image features from the defect category and the way of each defect feature extraction is analyzed. It also provides the evidence for the segmentation of CT images in wood defect.Image is influenced by all kinds of noise in the process of generation and transmission, lowering the quality of the image. Fuzzy C-means algorithm for wood defects segmentation of CT images, as the fuzzy C-means algorithm for noise robustness is poor, will lose some edge information of image segmentation. This paper improves from two aspects. One is at first to filter and process CT images of wood defect, and then is to segment fuzzy C-means algorithm. Our filter processes not only filter out the noise but also defect edge information. This paper gives experimental results including median filter, Butterworth low-pass filter, Butterworth high-pass filtering processing. So this article uses the median filter. Second, from the constraints of the improved fuzzy C-means algorithm itself, through the introduction of covariance matrix, we apply fuzzy C-means algorithm of mahalanobis distance to the wood defects in the CT image segmentation. Through the experiment, the wood defect in knot and cavity’s is highlighted, and edge defect continues, and also can retain a lot of details. The decayed and cracked wood defect’s effect is not obvious, but decaying segmentation is too big, and loses many of the details of the crack defects.The experimental results confirmed the fuzzy C-means algorithm for wood defect can segment effectively, which can provide effective assistance for the production and processing. It aims at selecting material reasonably, using material scientifically, improving the utilization rate of wood, saving timber resources effectively, and promising the sustainable development of forest ecology. |
