Analysis of the Mutli-scale Characteristics with Objects Extraction

Abstract

Image information with different displays distinct spatial structure. Scale is different for each remote sensing image analysis. Image analysis with single could not meet the heterogeneity and dynamic of pattern and process and objects extraction in the same resolution or is unreasonable. To acquire the high quality data of the surface, multi-scale image analysis is necessary. Object-oriented image analysis approach offers a solution to extract different objects from various images. Object-oriented image analysis creates meaningful objects and builds a hierarchy levels close to surface character using multi-scale segmentation. Objects could be analysed in their suitable levels. The patient advantage of this method is obvious that classification is based on the meaning objects, rather on the pixels. In object-oriented image analysis scale is not the pixel size, but the segmentation threshold. This paper presents a snapshot of work using multi-scale image analysis with several resolution images. The results show the characteristics of objects multi-scale analysis, the relationship of spatial and resolution. Image object levels from multi-scale segmentation are different from other images with various spatial resolutions. Different objects in surface could be represented in corresponding image object levels. Context information and texture information are very important factors for image information extraction.