The Research On The Modeling Of Airlight Gradient Prior Characteristic And The Dehazing Method Based On Polarized Layering And Analysis

In low visibility weather such as fog and haze,the photoelectric imaging detection system is affected by atmospheric scattering and the imaging quality is degraded.It is of great research significance and application value to study the method of target polarization reconstruction under fog and haze to improve the clarity and contrast of the detected image.Accurate estimation of airlight is one of the key issues in polarized dehazing research.In order to improve the accuracy of airlight estimation in polarized defogging studies,this paper starts from the statistical characteristics of foggy images and statistically analyzes the gradient histograms of a large number of foggy images.It is found that the gradient distribution of airlight images approximately follows the narrow Gaussian Characteristics,and then proposed a model of airlight gradient channel prior that characterize airlight gradients.Secondly,combining atmospheric scattering imaging model and polarization imaging theory,a polarization layering and analysis defogging method that does not depend on the polarization degree of airlight is proposed.The prior model of airlight gradient channel is used as the characteristic constraint condition of airlight estimation to construct an image layering function.The original foggy polarized image is layered to estimate the airlight polarized image.Finally,the airlight is analyzed from the polarized airlight image,and the atmospheric light at the infinity is estimated from the atmospheric light image to realize the defog reconstruction of the foggy image.The experimental results show that the method in this paper improves the accuracy of airlight estimation,which makes the reconstructed image clearer and the target reduction is higher;and it is suitable for the reconstruction of foggy images at different concentrations.The main work of this paper are as follows:1)By analyzing the statistical characteristics of the gradient histogram of foggy images at a large number of different concentrations,the typical stable characteristics of airlight in the gradient domain are explored,and the airlight gradient distribution has a narrow Gaussian distribution form.The model is verified and solved to realize the analysis and modeling of airlight gradient distribution characteristics.2)Combining atmospheric scattering imaging model and polarization imaging theory,a polarized layering and analysis defog model is constructed.Taking the airlight gradient channel priori as a feature constraint,an image layering method based on the airlight gradient prior information is proposed.The airlight polarization image is estimated from the original foggy polarization image and the atmosphere is solved from the airlight polarization image,and estimate the atmospheric light at infinity according to the gradient characteristics of atmospheric light to achieve the solution of key parameters in the model.3)By carrying out actual measurement experiments and comparative experiments in different concentrations and different scenarios,the effectiveness of the polarized layering and analysis dehazing method proposed in this paper is verified,and the atmospheric light polarization degree is discussed in detail.The discussion results show that the atmospheric light polarization degree is used as a Global constants can improve the accuracy of atmospheric light estimation.