Design Of Rotating Polarization Detection System And Method Of Fog Image Reconstruction

Fog and haze are due to the scattering of light by solid and liquid particles in the atmosphere which results in poor visibility.This paper focuses on seriously deteriorated image quality caused by a strong scattering environment.To improve image quality,a rotational polarization detection system was designed.It was combined with a new polarized image reconstruction method which was tested successfully in numerous experiments.The key aspects are summarized as follows:(1)A polarization detection system was designed,which uses an embedded system for image acquisition,image processing and network data transmission.The system includes a revolver structure and is able to obtain polarization image data automatically.(2)Furthermore,a method was proposed which reduces the influence of a strong scattering environment on image quality.Polarized filtering method and adaptive light channel prior method were used to estimate reconstruction parameters of each pixel in the image.The results were used as input for the fog weather polarization imaging model to reconstruct the optimal image globally.In addition,texture information regarding the degree of polarization was used to further enhance the quality of the reconstructed image.(3)Field experiments were carried out to analyze the effect of reconstruction parameters for different visibility conditions.The experimental results show that the polarization reconstruction method is able to recover the target information from a strong scattering environment very well.The polarization enhancement method further improves background rejection.