摘要
遥感图像具有信息大、灰度级大等特点,传统简单组合特征出现特征冗余、维数高等缺陷,造成图像分类精度差。为提高分类的准确性,提出一种多目标优化人工蜂群算法的遥感图像自动分类算法(ABC-SVM)。首先提取遥感图像的颜色、纹理特征,然后采用人工蜂群算法对特征进行选择和优化,最后采用支持向量机对优化特征进行训练,建立遥感图像自动分类模型。仿真结果表明,ABC-SVM克服了传统组合特征算法的缺陷,提高了遥感图像分类准确率,加快分类速度,可以满足遥感图像分类的实时性要求。
This paper proposed a remote sensing image classification algorithm by combination improved artificial bee colony algorithm with support vector machine ABC - SVM). Firstly, the color and texture features of remote sensing image color were extracted, and then the artificial bee colony algorithm was used to select and optimize the features. Finally, the optimized features were input to support vector machine to learn building remote image automat- ic classification model. The simulation results show that ABC - SVM can overcome the traditional algorithm, improve the remote sensing image classification accuracy rate, and increase the speed of classification, it can satisfy the real time requirement of remote sensing image classification.
出处
《计算机仿真》
CSCD
北大核心
2013年第6期378-381,385,共5页
Computer Simulation
关键词
遥感图像
人工蜂群算法
支持向量机
偏好区域
Remote sensing image
Artificial bee colony ( ABC ) algorithm
Support vector machine ( SVM )
Preference region