基于神经网络方法的语义级图像分级分类
Semantic classification Of Image Using The Neural Network Method
本文提出了首先根据空间颜色特征采用基于自组织稀疏RAM的n-tuple神经网络（SSn-tuple）进行室内和室外图像的场景分类，再根据Canny算子检测得到的边缘方向信息采用SSn-tuple神经网络进行城市和乡村图像的场景分类，取得了较好的分类效果。
In this paper, firstly the spatial color features of a image are imported into the n-tuple neural networks with self-organizing sparse RAM to classify it as indoor or outdoor. Then, the edge direction information of a image are imported into SSn-tuple neural network to classify it as city or landscape. The method is confirmed by the test that can get good classify and retrieval performance.
吴乐南、周兆捷
计算技术、计算机技术电子技术应用
基于内容的图像检索语义分类自组织映射N-tuple分类器稀疏分布存储器
ontent-based image retrievalSemantic classificationSelf-Organizing MapsN-tuple ClassifierSparse Distributed Memory
吴乐南,周兆捷.基于神经网络方法的语义级图像分级分类[EB/OL].(2004-06-07)[2025-08-11].http://www.paper.edu.cn/releasepaper/content/200406-35.点此复制
评论