Ensemble learning method based on self-adaptive sample expansion

Abstract

The invention discloses an ensemble learning method based on self-adaptive sample expansion. On the one hand, multiple weak classifiers are integrated through adopting a manner of bootstrap feature sampling and dynamic weighted voting, and the advantages of high classification accuracy and good repeatability of the ensemble learning method are inherited; and on the other hand, the method can realize self-adaptive expansion of samples through iteration classification and sample screening based on spatial dispersion and attribute similarity on the basis of a small number of ground survey samples, and the model under-learning problem caused by small / medium-sized samples in classification is solved. The ensemble learning method to which the scheme relates adopts a manner of self-adaptive sample expansion, can effectively solve the problem of insufficient samples in a remote-sensing classification process, and reduces manpower and time spent by researchers for obtaining samples at the sametime.

Description

technical field [0001] The invention relates to the technical field of remote sensing, in particular to a method for extracting remote sensing information under the condition of small samples by adaptively expanding a small number of collected real samples. Background technique [0002] With the improvement of imaging technology, the means of remote sensing data acquisition are becoming more and more diverse. One of the important applications of remote sensing data is image classification. Early remote sensing image classification mainly includes supervised classification and unsupervised classification. The supervised classification method requires model training with the support of ground object samples, and then classifies the entire image. There are two main ways to obtain ground feature samples: one is to rely on prior knowledge to determine different types of ground features through visual interpretation on images, and select samples of various types from the images; t...

AI Technical Summary

Problems solved by technology

[0004] In view of the deficiencies in the prior art, especially the problem that it is difficult to obtain real samples in the remote sensing image classification process, the purpose of the present invention is to propose an integrated learning method based on sample self-adaptive expansion, the application of which can effectively Solve the problem of insufficient samples, and then realize the classification of remote sensing images by using the integrated learning method in machine learning to improve the timeliness and accuracy of sample classification

Images