摘要
针对图像分割的复杂性和局限性,作者提出一种基于最小二乘支持向量机(LS-SVM)的木材表面缺陷网格化检测方法。首先将木材表面图像划分成互不重叠的矩形块,然后依次计算每个矩形块图像的特征向量,用于描述各个矩形块图像,其特征向量由颜色特征和纹理特征等参数共同组成。最后将特征向量归一化后送入LS-SVM分类器,利用特征向量的相似度来进行缺陷的定位和识别。实验结果表明,该方法可有效进行木材表面缺陷检测,检测准确率超过93%。
Due to the complexity and limitations of image segmentation,this paper proposed a wood surface defects gridding detection method based on least squares support vector machines（LS-SVM）.The wood surface image was first divided into non-overlapping rectangular blocks.And then,every block＇s feature vector,which consisted of color features and texture features,was calculated to describe the blocks accurately.Finally,the extracted feature vectors were normalized and inputted into the LS-SVM classifier to locate and detect the defects.The experimental results have shown that this method can effectively identify the defect regions and the detection accuracy is higher than 93%.
出处
《林业科技开发》
北大核心
2012年第6期73-76,共4页
China Forestry Science and Technology
基金
林业公益性行业科研专项(编号:201004007)
国家自然科学基金项目(编号:30972314)
