Title :
Video image preprocessing based on neural network
Author :
Wang, Jing ; Yao, Yi ; Chen, Dan
Author_Institution :
Coll. of Inf. Eng., Sichuan Univ. of Sci. & Eng., Zigong, China
Abstract :
The difference between two video images acquired in the same scene under different atmospheric conditions is great because the quality of images is directly affected by the atmospheric conditions. We can suppose the differences of frequency spectrum are merely induced by the atmospheric modulation transfer function. The atmospheric modulation transfer inverse system of the bad weather based on neural network can be acquired by the relationship of the image extracted from the bad weather video and the one from good atmospheric condition as the input and the output of the inverse system, and the bad weather effects can be eliminated which lead the video images degenerated.
Keywords :
feature extraction; neural nets; video signal processing; atmospheric condition; atmospheric modulation transfer inverse system; bad weather video; frequency spectrum; image extraction; neural network; video image preprocessing; Atmospheric modeling; Degradation; Equations; Mathematical model; Meteorology; Modulation; Neural networks; atmospheric modulation transfer inverse system; maximum entropy; neural network; video images preprocessing;
Conference_Titel :
Natural Computation (ICNC), 2012 Eighth International Conference on
Conference_Location :
Chongqing
Print_ISBN :
978-1-4577-2130-4
DOI :
10.1109/ICNC.2012.6234753