A multi-feature joint classification method for remote sensing images based on openstreetmap

Abstract

The invention discloses a remote sensing image multi-feature classification method based on OpenStreetMap. The invention extracts a variety of spatial features from the remote sensing image to describe the characteristics of different aspects of the image, and uses morphological corrosion filtering to remove the confused pixels of the object boundary. Intra-class cluster analysis, remove wrong labels, obtain accurate classification samples, combine acquired image features and classification samples, and combine support vector machine classifiers to calculate the probability of each pixel belonging to different categories under different features, combined with reliable classification The weighted fusion of category probabilities corresponding to different features can be realized to complete the classification of remote sensing images. The invention effectively integrates multiple classic spatial feature extraction algorithms by mining the spatial information of remote sensing images, can enhance the separability of different categories, and obtain more accurate classification results of each category and higher classification precision.

Description

technical field [0001] The invention relates to the technical field of remote sensing image processing, in particular to a multi-feature joint classification method, especially an OpenStreetMap-based multi-feature joint classification method for remote sensing images. Background technique [0002] With the development of WEB 2.0, the public can not only acquire knowledge from the Internet, but also upload their own knowledge and experience to the Internet to share information with others. Among them, the information with geographic location contributed by the public is called Volunteered Geographical Information (VGI). Among them, OpenStreetMap (OSM) is one of the most typical VGI projects, and its goal is to draw the power of the public to draw an open-source and free world map, and ultimately serve the public. The public can use portable GPS devices to record trajectories as a means of measurement, and they can also use Bing images as a base map for mapping based on local...

AI Technical Summary

Problems solved by technology

The current multi-feature fusion method often uses vector superposition to form a higher-dimensional feature input classifier for interpretation, which fails to consider the contribution of different features to the classification result, which is not conducive to the further improvement of classification accuracy.

Images