Share Email Print
cover

Proceedings Paper

Destriping methods for high resolution satellite multispectral remote sensing image based on GPU adaptive partitioning technology
Format Member Price Non-Member Price
PDF $14.40 $18.00
cover GOOD NEWS! Your organization subscribes to the SPIE Digital Library. You may be able to download this paper for free. Check Access

Paper Abstract

The stripe noise is a key factor that affects imaging quality of satellite multi-hyperspectral remote sensing images, which also has a serious effect on the interpretation and information extraction of remote sensing images. Complex surface textures mixed with strip noises in the high-resolution multi-spectral remote sensing of satellite are extremely difficult to remove, this paper analyzes the Markov random field prior model method, combines the Huber function to propose a universal, fast and effective Huber Markov destriping method. According to the statistical characteristics of the image gray level variation, the distribution features and mutual relationship between each pixel and its neighborhood pixels in the image, the co-occurrence matrix reflecting the contrast gray characteristics of the image is connected with the threshold T of Huber function, which is automatically iteratively determined during the noise removal process, and will be able to remove image noises as well as preserving its edges and details effectively. In order to solve the time complexity of the algorithm caused by the pixel space information introduced by the Huber Markov random field algorithm, the GPU adaptive partitioning technique is adopted to accelerate the algorithm. Experimental results show that the destriping method based on Huber function Markov random field can remove the strip noise effectively, while preserving texture details of the image, which can be applied to a variety of noise-containing images. Meanwhile, GPUbased adaptive partitioning technology has been adopted, which has greatly improved the computational efficiency of processsing massive remote sensing images, and lays a foundation for the application of remote sensing satellite images in China.

Paper Details

Date Published: 24 October 2018
PDF: 9 pages
Proc. SPIE 10783, Remote Sensing for Agriculture, Ecosystems, and Hydrology XX, 107832I (24 October 2018); doi: 10.1117/12.2325311
Show Author Affiliations
Xue Yang, China Academy of Space Technology (China)
Feng Li II, China Academy of Space Technology (China)
Lei Xin III, China Academy of Space Technology (China)
Cheng Wang III, China Academy of Space Technology (China)
XiaoYong Wang, China Academy of Space Technology (China)
Henan Univ. (China)
Xing Chang, China Academy of Space Technology (China)
Lanzhou Jiaotong Univ. (China)


Published in SPIE Proceedings Vol. 10783:
Remote Sensing for Agriculture, Ecosystems, and Hydrology XX
Christopher M. U. Neale; Antonino Maltese, Editor(s)

© SPIE. Terms of Use
Back to Top