论文部分内容阅读
要对高分辨率遥感影像进行分类,采用面向对象的遥感影像分析技术比传统的面向像元的遥感影像分析技术优越。要使用面向对象的遥感影像分析技术,关键的第一步是要对遥感影像进行分割,以便得到一系列与地物有密切联系的影像对象。分割的准确性与分割的尺度选择有关。本文针对成都平原高分辨率卫星影像分割尺度选择进行试验和研究,采用不同尺度对试验区不同分辨率遥感影像进行影像分割,并比较分割结果,得出成都平原高分辨率遥感影像数据分割最佳尺度与影像对象亮度均值标准差最大值所对应的分割尺度一致;并且遥感影像空间分辨率越高,最佳分割尺度越大,反之亦然。
To classify high-resolution remote sensing images, the use of object-oriented remote sensing image analysis technology is superior to the traditional pixel-oriented remote sensing image analysis technology. To use object-oriented remote sensing image analysis, the key first step is to segment the remote sensing images to get a series of image objects that are closely related to the features. The accuracy of segmentation is related to the scale selection of segmentation. In this paper, the scale selection of high-resolution satellite image segmentation in Chengdu Plain is studied and studied. Different scales are used to segment the remote sensing images with different resolutions in the experimental area. The results of the segmentation are compared. And the segmentation standard corresponding to the maximum value of standard deviation of the mean value of brightness of the image object is the same; and the higher the spatial resolution of the remote sensing image is, the larger the optimal segmentation standard is and vice versa.