摘要
基于内容的医学图像分类是一个复杂的非线性问题,分类器的性能主要取决于提取的特征和模式识别算法。讨论了医学图像基本特征提取方法和多特征融合技术的发展,以及常用的模式分类算法。最后指出了支持向量机在医学图像分类中应用时需要解决的问题。
Content-based medical image classification is a complicated non-linear problem. The performance of classification depends on extracted features and pattern recognition method. The developments of basic feature-extract methods, multi-feature combination and pattern recognition method are discussed. At last, several problems which need to be resolved when support vector machine is applied to medical image classification.
出处
《科学技术与工程》
2007年第1期85-90,共6页
Science Technology and Engineering
基金
广东省自然科学(05100514)基金资助
关键词
医学图像分类
特征提取
多特征融合
模式识别
支持向量机
medical image classification feature extraction multi-feature combination pattern recognition support vector machinel
