Design of Remote Sensing Image Processing Algorithm Based on Machine Learning

Remote sensing (RS) image classification is one of the most basic problems in RS image information processing. Its classification technology is the key technology in RS application system. In the practical application of RS image classification processing, a large number of training data are needed to obtain high-precision classification results, marking these training data requires a lot of manpower and material resources, and it is also time-consuming. Traditional hyperspectral image processing methods can only extract the shallow features of the image, but ignore the deep features of the image. In this paper, an RS image processing algorithm based on deep learning (DL) is proposed, which can fully mine the useful information of a large number of unlabeled samples in RS image classification, so as to expand a small number of labeled samples, enhance the classifier’s ability, and improve the classification accuracy. Experiments show that the RS image processing algorithm in this paper has high efficiency and short response time, with an average of about 0.3 s; at the same time, the root mean square error (RMSE) of the algorithm is low, which can be stabilized at about 0.523. The research in this paper is of great practical value and practical significance for improving the classification accuracy of RS images with few labeled training samples.