Abstract
Abstract. Multi-temporal remote sensing images classification is very useful for monitoring the land cover changes. Traditional approaches in this field mainly face to limited labelled samples and spectral drift of image information. With spatial resolution improvement, “pepper and salt” appears and classification results will be effected when the pixelwise classification algorithms are applied to high-resolution satellite images, in which the spatial relationship among the pixels is ignored. For classifying the multi-temporal high resolution images with limited labelled samples, spectral drift and “pepper and salt” problem, an object-based manifold alignment method is proposed. Firstly, multi-temporal multispectral images are cut to superpixels by simple linear iterative clustering (SLIC) respectively. Secondly, some features obtained from superpixels are formed as vector. Thirdly, a majority voting manifold alignment method aiming at solving high resolution problem is proposed and mapping the vector data to alignment space. At last, all the data in the alignment space are classified by using KNN method. Multi-temporal images from different areas or the same area are both considered in this paper. In the experiments, 2 groups of multi-temporal HR images collected by China GF1 and GF2 satellites are used for performance evaluation. Experimental results indicate that the proposed method not only has significantly outperforms than traditional domain adaptation methods in classification accuracy, but also effectively overcome the problem of “pepper and salt”.
Highlights
Multi-temporal remote sensing images, which are acquired by sensors mounted on board of satellites that periodically pass over the same geographical area, become an important tool for performing Earth monitoring
Two cases of multi-temporal high-resolution remote sensing images are considered in this paper: One is that all images cover different geographic areas, another is that all the images cover the same geographic area and are registered
simple linear iterative clustering (SLIC) segmentation is used and Majority voting manifold alignment is proposed for achieving this goal
Summary
Multi-temporal remote sensing images, which are acquired by sensors mounted on board of satellites that periodically pass over the same geographical area, become an important tool for performing Earth monitoring. Two main obstacles prevent multi-temporal technology from reaching a broader range of applications. There is generally a lack of labelled data at each acquisition. Multitemporal images obtained under different conditions show the spectral drift. Such drift generally happens due to differences in acquisition and atmospheric conditions or changes in the nature of the observed object (Tuia et al, 2016a). The obstacle of label scarcity can be solved by using available labelled samples from other temporal images. The distributions of source image and target image are significantly different. To classify un-labelled image using labelled image efficiently and accurately, modern processing systems must be designed to be robust for solving spectral drift
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have