2016
2016
-
Record 433 of
Title:Image dehazing technique based on polarimetric spectral analysis
Author(s):Xia, Pu(1,2); Liu, Xuebin(1)Source: Optik Volume: 127 Issue: 18 DOI: 10.1016/j.ijleo.2016.05.071 Published: September 1, 2016Abstract:Image took under hazy weather suffers from poor contrast and resolution, the haze particles will attenuate the light reflected by the targets and add unwanted scattering light. Based on the fact that the target reflection and scattering light have different polarimetric characteristics, light's power to penetrate the haze particles is linked with wavelength, this paper combines the polarimetric dehazing technique with spectral analysis, firstly proposed the polarimetric spectral dehazing method. A polarimetric spectral imager is used to obtain data under a continuously changing weather circle, the dehazing result is analyzed under five different spectral channels of 451.4 nm, 551.2 nm, 650.9 nm, 750.7 nm and 850.5 nm. The results show that our method can effectively recover the haze degenerated image under visible and infrared channels, the restoration quality of detailed information of the near-field and the far-field targets are in varying degrees under different channels. The dehazing process can enhance the image contrast by 1.68-3.64 times under different wavelengths. Two correction factors, which regularity is given for particle use, are introduced to revise an image restoration result. ? 2016 The Authors. Published by Elsevier GmbH.Accession Number: 20162302453314