| 基于空频变换的木材缺陷图像分割 |
| |
| 引用本文： | 白雪冰,王林.基于空频变换的木材缺陷图像分割[J].东北林业大学学报,2010,38(8). |
| |
| 作者姓名： | 白雪冰 王林 |
| |
| 作者单位： | 东北林业大学,哈尔滨,150040;东北林业大学,哈尔滨,150040 |
| |
| 基金项目： | 黑龙江省自然科学基金，东北林业大学研究生科技创新资助项目 |
| |
| 摘 要： | 针对木材缺陷这一自然纹理型事物,为了提取出其缺陷目标部分,进行下一步的分析和识别,采用一种空频变换方法对缺陷图像进行分割。选取虫眼、死节、活节3类木材缺陷图像样本各50个,构造一组多通道的Gabor滤波器对缺陷图像进行滤波,并提取出图像的多方向Gabor能量特征。最后结合模糊聚类算法和数学形态学后处理操作对缺陷图像进行了成功的分割。实验结果表明,此方法对3种木材缺陷图像的平均分割正确率分别达到了95.81%、94.58%、96.52%,证明了该方法的有效性。
|
| 关 键 词： | 木材缺陷 图像分割 空频变换 Gabor滤波器 模糊聚类 |
Image Segmentation of Wood Defects Based on Space-frequency Transformation |
| |
| Bai Xuebing,Wang Lin.Image Segmentation of Wood Defects Based on Space-frequency Transformation[J].Journal of Northeast Forestry University,2010,38(8). |
| |
| Authors: | Bai Xuebing Wang Lin |
| |
| Institution: | Bai Xuebing,Wang Lin(College of Mechanical , Electrical Engineering,Northeast Forestry University,Harbin 150040,P.R.China) |
| |
| Abstract: | Considering wood defects which are a type of natural texture things,a method based on space-frequency transformation was used to segment defect images to extract the defect parts of the wood for the analysis and identification afterwards.Three kinds of wood defects,worm-hole,dead knot and live knot,were separately selected in a sample of 50 images and filtered with a set of constructed multi-channel Gabor filters,and then the multi-directional Gabor energy features of the defect images were extracted.Finall... |
| |
| Keywords: | Wood defects Image segmentation Space-frequency transformation Gabor filters Fuzzy clustering |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
