A remote sensing image adaptive feature selection segmentation method and system

Abstract

The invention provides a method and system for adaptive feature selection and segmentation of remote sensing images, including data preparation, cropping and data preprocessing of high-resolution images; an improved model, and an adaptive feature selection module based on the basic model network. The adaptive feature selection module is used to perform feature extraction on different feature maps. By performing global average pooling feature extraction on multiple feature maps of different scales or different receptive fields obtained by the basic model network, the fully connected layer and activation function are used to learn different features. The weight distribution of the importance of the graph is selected according to the weight distribution to realize the adaptive feature selection process; the optimized deep learning model is trained; the input visible light image is segmented according to the optimized deep learning model obtained from the training. The invention can not only effectively improve the accuracy of the segmentation model, but also reduce the calculation amount of the model, and has the advantages of universality, simple operation, strong performance and the like.

Description

technical field [0001] The invention belongs to the field of satellite remote sensing image processing, particularly relates to a segmentation scheme for high-resolution remote sensing images, and proposes a new deep learning model to realize an adaptive feature selection scheme for remote sensing images. Background technique [0002] In recent years, with the deepening of the application of deep learning in the field of image processing, semantic segmentation has been rapidly developed as an important research direction in image processing. The segmentation of remote sensing images is of great significance to the subsequent processing of remote sensing images. Especially on the on-board platform, the image is segmented in advance, and the target of the region of interest is extracted to provide prior knowledge for the understanding of the image. Applying artificial intelligence and computer vision to the on-orbit satellite platform can provide real-time intelligent service...

AI Technical Summary

Problems solved by technology

At present, there is no good solution to the problem of scale selection and authorized patents in China.

Images