Effective background removal method based on generative adversary networks

2020 ◽  
Vol 29 (05) ◽  
2014 ◽  
Vol 989-994 ◽  
pp. 4107-4110
Author(s):  
Shao Jun Guo ◽  
Zhe Wang

The space-based visible observation imaging platform for sky targets is influenced by many factors, a serious factor is the light of background too bright. A image with the bright stray light background has some high gray areas those may submerge the targets info. Aimed at the shortcomings of traditional background removal method in target extraction under the bright stray light background, according to the differences of bright stray light background and sky targets imaging characteristics, this paper has made some research of the algorithms about how to remove the bright stray light background but not delete the targets info. The algorithm we got that give us great results will be shown in the paper. It solves the problems of the bright background light removal and greatly retain the targets info which submerged in the bright areas.


1993 ◽  
Vol 32 (S2) ◽  
pp. 125 ◽  
Author(s):  
Matthew Newville ◽  
Pěteris Līviņ\us ◽  
Yitzhak Yacoby ◽  
John J. Rehr ◽  
Edward A. Stern

2016 ◽  
Vol 31 (3) ◽  
pp. 767-772 ◽  
Author(s):  
Qingdong Zeng ◽  
Lianbo Guo ◽  
Xiangyou Li ◽  
Meng Shen ◽  
Yining Zhu ◽  
...  

An approach of portable laser-induced breakdown spectroscopy based on a fiber laser with a background removal method was proposed.


Author(s):  
Hengqian Zhao ◽  
Lifu Zhang ◽  
Xia Zhang ◽  
Jia Liu ◽  
Taixia Wu ◽  
...  

2014 ◽  
Vol 926-930 ◽  
pp. 3050-3053
Author(s):  
Zhao Fei Li ◽  
Jiang Qing Wang

In the task of the image processing and analysis, the background noise removal is a important step. In the image background noise removal, there are many methods which is popular for the researchers. For example, the gray threshold methods are commonly taken to remove the noises which have large contrast to the interest objects. However, there are many noises with no variance with the interest objects in the gray level. For these noises, the gray level based noise removal method is totally futile, while the contour feature has its super performance for reducing this sort of noise. For the contour feature based image background removal method, the contour model is the key. This paper proposes a novel method for modeling the contour feature of the interest objects. With this method, a novel background noise which has the same gray level to the background noise is totally removed.


Sign in / Sign up

Export Citation Format

Share Document