The best way to conference proceedings by Francis Academic Press

Web of Proceedings - Francis Academic Press
Web of Proceedings - Francis Academic Press

Research on Denoising Method of Remote Sensing Images Based on Convolutional Neural Network

Download as PDF

DOI: 10.25236/ciais.2019.034

Author(s)

Rui Jiang

Corresponding Author

Rui Jiang

Abstract

Remote sensing image is different from general image, it has a high feature dimension, and the current remote sensing image detection technology is difficult to express its high-dimensional features well. Once the remote sensing image has noise, the difficulty of feature extraction is further improved. With the continuous development of science and technology, the research of image denoising is not limited to the professional field. Now photography has become an indispensable part of everyone's life, and the demand for image denoising is also increasing. However, there are still some shortcomings in the traditional image denoising algorithm. Based on the above problems, a method of remotely sensed image denoising based on convolution neural network is proposed in this paper. Compared with traditional denoising methods, convolution neural network is more inclined to consider the local spatial characteristics of pictures. Through network training, the features of pictures are learned, and the noisy pictures are compared by using the learned features, so as to achieve the purpose of denoising. Experiments show that the method effectively improves the image quality of remote sensing images and has good noise reduction effect.

Keywords

Convolutional Neural Network;Remote Sensing Images;Picture Noise Reduction; Feature Extraction