摘要
本文提出了一种基于矩不变量和支持矢量机对遥感图像中的目标进行识别的方法，该方法提取目标七个Ｈｕ矩不变量作为特征矢量，应用支撑矢量机方法对其进行分类识别，结果表明，这一方法对仅含有目标和背景的遥感图像具有很好的分类识别结果。此外，我们发现对于不同的图像，其二值化取值范围对识别结果有着直接的影响。
An Efficient target recognition method for remote sensing image is proposed in this paper, which based on moment invariant and support vector machines. Firstly, it extracts the seven Hu's invariant moments as feature vectors, then uses the support vector machines to recognize plane and ship targets on a binary remote sensing. The experiment results show that the new method has a better recognition results. Moreover, we observed that the range of the binary values of image affects the performance of recognition results.
出处
《信号处理》
CSCD
2002年第1期1-4,共4页
Journal of Signal Processing
