基于支持向量机的目标图像识别技术

Target Image Recognition Technology Based on Support Vector Machine

介绍了目标图像的前期处理、目标图像的特征提取和基于支持向量机的目标图像识别方法。为了验证方法的正确性,采用了一批太空目标图像对其算法进行了检验。数据实例表明,支持向量机识别的正确率比其它两种神经网络方法的正确率要高很多。

This paper introduces a target image recognition technology based on support vector machine. The author describes the pretreatment method of target image, the eigenvector pick-up technique of target image and the target image recognition technology based on support vector machine. In order to validate the technology of this paper, the author gives a set of space image to test the proposed algorithm.