论文部分内容阅读
对合成孔径雷达(synthetic Aperture Radar,SAR)图像提取得到的几何、灰度、纹理特征共66个特征量,采用封装模式算法进行特征选择,降低特征维度并提高对溢油及疑似溢油样本的识别率。特征选择采用二进制离散粒子群优化(binary particle swarm optimization,BPSO)和支持向量机(support vector machine method,SVM)的封装模式算法(BPSO-SVM)进行,该方法在特征选择的同时可对支持向量机模型中的参数进行优化。论文采用BPSO-SVM算法和序列前向搜索(sequential forward selection,SFS)算法、序列后向搜索(sequential backward selection,SBS)算法与SVM算法相结合特征优化算法(SFS-SVM和SBS-SVM算法)进行实验。并将BPSO-SVM算法、SFS-SVM算法、SBS-SVM算法和直接使用SVM算法的分类识别结果进行比较。实验结果表明,BPSO-SVM算法在SAR图像上溢油特征量筛选与识别效率方面行之有效。
Sixty-six feature vectors were extracted from the synthetic aperture radar (SAR) images, and the feature extraction was performed using the encapsulation mode algorithm to reduce the feature dimension and improve the performance of the oil spill and suspected oil spill samples Recognition rate. The feature selection is based on the BPSO-SVM (BPSO-SVM), which is based on binary particle swarm optimization (BPSO) and support vector machine method (SVM) Machine model parameters are optimized. In this paper, we use BPSO-SVM algorithm and SFS-SMP algorithm, SBS-SVM algorithm and SBS-SVM algorithm, conduct experiment. The classification results of BPSO-SVM, SFS-SVM, SBS-SVM and SVM are also compared. Experimental results show that the BPSO-SVM algorithm is effective in screening and identifying spilled oil features on SAR images.