论文部分内容阅读
以往面向对象影像分析的分割尺度主要依靠经验并结合目视来进行选择,带有一定的主观性。本文针对利用高分辨率遥感影像进行土地利用信息提取的目的,采用面向对象的方法完成了两个典型实验区域的多尺度分割。主要研究了分割参数的选择;重点提出了一种最优分割尺度计算模型。结果表明,此模型计算最优分割尺度方便快捷,而且计算出的最优分割尺度可信度高,避免了目视的主观性,能提高影像分割效率,为后续土地利用分类打下了良好的基础。
In the past, the segmentation standard of object-oriented image analysis mainly depends on experience and combined with visual to make a choice, with a certain subjectivity. In order to extract land use information from high resolution remote sensing images, this paper uses object-oriented method to accomplish multiscale segmentation of two typical experimental regions. The main study of the choice of segmentation parameters; focus on an optimal segmentation scale calculation model. The results show that this model is convenient and efficient in calculating the optimal segmentation scale, and the confidence of the calculated optimal segmentation scale is high, which avoids the visual subjectivity and improves the image segmentation efficiency, which lays a good foundation for the subsequent classification of land use .