论文部分内容阅读
为了寻找快速准确、方便调查病态性入侵植物金钟藤分布的方法和技术,该文以海南省定安县为研究区,结合光谱信息与纹理信息,运用支持向量机的分类方法,提取该地区金钟藤信息,并将分类的结果与最大似然法、单源数据(B3波段图像)的支持向量机分类结果进行定性与定量的比较分析。结果表明:基于比值植被指数、B3波段为纹理信息源的支持向量机方法有利于金钟藤信息的提取;通过结合纹理信息和支持向量机的方法,实现了分类精度的提高。
In order to find a fast and accurate method and technique for investigating the distribution of Morphological trapping plant Apis mellophytus, aiming at the research area of Ding’an County in Hainan Province, this paper combined the spectral information and texture information, using the classification method of support vector machine, The classification results are compared qualitatively and quantitatively with SVM classification results of maximum likelihood method and single source data (B3 band images). The results show that the SVM based on the ratio of vegetation index and the B3 band as the texture information source is favorable to the extraction of the information of the Tricholoma giganteum. By combining the texture information and support vector machine, the classification accuracy is improved.