基于最大似然法的钓鱼岛航空遥感监视监测信息提取

摘要

This paper proposed the maximum likelihood method for supervising the categorization of monitoring and sur-veying images on Diaoyu Dao using the high resolution image data from National Island Monitoring and Surveying System. The basic types of land coverage on Diaoyu Island were extracted accurately, and the geomorphological features on Diaoyu Dao were also inversed clearly by using this method. The categorizing resolutions were further improved by adding denoising method based on primary and secondary analysis. This paper indicated that the proposed methods had good cat-egorizing effects.%基于国家海岛监视监测系统的高精度影像数据，运用最大似然分类法对钓鱼岛监视监测信息图像进行监督分类，较为准确地提取出钓鱼岛基本地物类型，清晰反演钓鱼岛地貌特征，并引用了主次要分析去噪法进一步提高了分类精度，取得较好的分类效果。