提出一种基于主成分分析和支持向量机与线性判别分析结合算法的合成孔径雷达(synthetic aperture radar,SAR)图像目标鉴别方法. 利用主成分分析算法对SAR图像向量进行降维并提取其全局特征,对降维后的全局特征采用最小类内散度支持向量机算法进行变换,并对变换结果训练生成最佳分类器,进行分类完成目标鉴别. 实验结果表明该方法可以获得较高的分类正确率.
We propose a method for synthetic aperture radar images target discrimination based on the principal component analysis and an approach combining support vector machine (SVM) and linear discriminant analysis(LDA). Dimensionality of the image vector is reduced and the global features are extracted by using principal component analysis. The global features are transformed and the results are used to generate classifiers which complete target discrimination. The results show the high performance of the proposed method.
[1] Amit K, Mishra. Validation of PCA and LDA for SAR ATR //IEEE Region 10 Conference. Hyderabad, India, 2008:1-6.
[2] Wang C, Zhang B, Wen X Y, et al. Automatic target recognition in high resolution SAR image based on backscatter characteristics[J]. Chinese Journal of Radio Science, 2004, 19(4):422-426(in Chinese). 王超, 张波, 温晓阳,等.基于雷达散射特性的高分辨率SAR图像自动目标识别[J]. 电波科学学报,2004,19(4):422-426.
[3] Saurabh P, Lori M. Limitations of principal components analysis for hyperspectral target recognition[J]. IEEE Geoscience and Remote Sensing Letters, 2008, 5(4):625-629.
[4] Tatyana V, Lorenzo B, Gustavo C. Classification of hyperspectral images with regularized linear discriminant analysis[J]. IEEE Transactions on Geoscience and Remote Sensing, 2009, 47(3):862-873.
[5] Tao X, Vladimir C. A combined SVM and LDA approach for classification //IEEE Conf Neural Networks. Montreal, Canada, 2005: 1455-1459.
[6] Tefas A, Kotropoulos C, Pitas I. Using support vector machines to enhance the performance of elastic graph matching for frontal face authentication[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2001, 23(7):735-746.
[7] Gao J, Wang S T. Matrix pattern based minimum within-class scatter support vector machines[J]. Acta Electronica Sinica, 2009, 37(5):1051-1057(in Chinese). 皋军, 王士同. 基于矩阵模式的最小类内散度支持向量机[J]. 电子学报, 2009, 37(5):1051-1057.
