Single image defogging method based on sub-pixel and conditional antagonism generation network

Abstract

The invention discloses a single image defogging method based on sub-pixel and conditional antagonism generation network, which comprises the following steps: obtaining an original fogless image dataset and synthesizing the fogged data set according to a foggy day imaging model; inputting the fogged image to be processed into a generator G, wherein the network structure of the generator G is provided with a skip layer connection, a feature map with gradually reduced encoding output size is encoded, and the feature map is respectively obtained by deconvolution and sub-pixel in a decoding stage, and then the feature map is operated by convolution to obtain a fogless image output by the generator; inputting the non-fog image and the original non-fog image output from the generator G into thediscriminator D, and judging whether the non-fog image output from the generator D is true or not; the generators G and the discriminator D are constrained by antagonism at the same time, and the antagonism loss and L1 loss are calculated. The parameters of the generators G and the discriminator D are updated by back propagation according to the principle of stochastic gradient descent. When thetotal loss of the model converges, the training of the model is completed.

Description

technical field [0001] The invention relates to the technical field of image processing and pattern recognition, in particular to a single image defogging method based on sub-pixel and conditional confrontation generation network. Background technique [0002] Due to the refraction of water vapor and solid particles in the atmosphere, the captured images appear foggy visually. Image dehazing aims to restore a clean haze-free image from a hazy image. The generation of high-quality haze-free images is very beneficial to the later work of segmentation, detection and recognition. The representation of a foggy image can be modeled as: [0003] I(x)=J(x)t(x)+A(1-t(x)) (1) [0004] where I(x) and J(x) are the observed hazy image and clear scene irradiance, A is the global atmospheric light, and t(x) is called the medium transmittance. Assuming that the fog concentration is uniform, we can express t(x)=e -βd(x) , where β is the medium extinction coefficient and d(x) is the scen...

AI Technical Summary

Problems solved by technology

First, although the manual method can effectively dehaze the image, most of the manual methods are poor in dealing with the sky region of the image and the region whose brightness is close to the global atmospheric light A

Second, although the dehazing method based on the deep learning method is better than the traditional method, the main processing target of these methods (including the traditional method) is t(x), while ignoring the global atmospheric light A, if t(x ) is not properly estimated, the overall image defogging effect is not good enough

Although a complete end-to-end dehazing method, such as AOD-Net, has appeared in recent years, its main idea is still dealing with t(x), and it cannot jointly estimate t(x) and A well.

Images