Multispectral polarization image defogging method combined with sky light polarization model

Abstract

The invention provides a multispectral polarization image defogging method combined with a sky light polarization model, which comprises the following steps: under a foggy weather condition, firstly,acquiring polarization images under different spectrum channels by using a multispectral polarization imaging detection system, and simultaneously measuring the aerosol optical thickness of a corresponding spectrum band by using a sun photometer; secondly, inputting a detection waveband, a solar zenith angle, aerosol optical thickness, aerosol particle scattering information and the like into a sky light polarization model; obtaining a sky light polarization state distribution diagram through simulation, selecting a sky area in a horizontal field angle on a detection azimuth angle, and obtaining the average value of the light intensity and the average value of the polarization degree to serve as an atmospheric light intensity value and an atmospheric light polarization degree value at infinity respectively. And finally, defogging the polarization images of different spectrum channels by using a defogging method based on polarization difference. According to the method, the problem thatthe polarization degree atmospheric light information is estimated mistakenly when no sky area exists in a multispectral polarization image scene or some high-brightness objects exist in the scene issolved.

Description

technical field [0001] The invention belongs to the field of near-surface remote sensing image processing, and relates to a multi-spectral polarization image defogging method combined with a sky light polarization model, which is used to improve the image quality of optical detection system imaging under foggy weather and other harsh environmental conditions. Background technique [0002] In foggy weather conditions, due to the scattering phenomenon of atmospheric aerosol particles, the quality of images obtained by outdoor imaging systems is seriously degraded. Therefore, how to effectively remove the influence of fog, improve image quality, and obtain useful information that users are interested in has become a hot issue in the field of remote sensing and image processing in recent years. [0003] At present, the vast majority of image defogging methods are mainly divided into two categories based on computer vision and based on physical degradation models. Among them, th...

AI Technical Summary

Problems solved by technology

The model of this method is simple and easy to implement, and can effectively restore the visibility of the scene, but there are still some problems: when there is no sky area in the image or there are some high-brightness objects or strong light sources in the image, the sky area cannot be selected correctly , and then wrongly estimate the atmospheric light intensity and atmospheric polarization degree at infinity, which will affect the final dehazing result

This method is based on the partial polarization of skylight, but in dense fog conditions, where the degree of polarization of skylight is close to zero, this method will fail

Images