Method Research Of Fog Removal For High Resolution Remote Sensing Image

The preprocessing of remote sensing images is the basis for the application of remote sensing data.The removal of cloud and fog on the image data is an important part of image preprocessing.This article uses the "GF-1" satellite as an example to study the defogging method for high-resolution satellite remote sensing images;selecting the "GF-1" as the research object,mainly because it is the starting satellite of China's high score series,The method research of defogging the satellite image of "GF-1" has certain significance for defogging of other high-resolution images.In order to complete the defogging work of a single "GF-1" satellite image,the content of this paper is as follows:1.Introduce the imaging principle of high-resolution remote sensing images under fog conditions,study the image degradation model containing fog images,and study the spatial distribution characteristics,spatial frequency characteristics,and electromagnetic spectrum characteristics of mist-containing remote sensing images.It provides a theoretical basis for the subsequent defogging.2.According to the analysis of the fog image features of the analysis,several algorithm principles suitable for defogging single high-resolution images are summarized: including the Homomorphic filtering method,after the use of Fourier transform,the fog information mainly exists in the low-frequency part of the spatial frequency domain.High-resolution images containing smog are processed;the band fusion recombination method is used to process the high-resolution images containing smog using the different image levels of the different bands affected by smog;and based on the prior method of defogging of dark colors,The dark-color statistics a priori are used as a convergence condition for the defogging process to process the high-resolution images containing fog.3.Use the clear high-resolution remote sensing image to measure the dark color of true color and false color remote sensing images to verify the dark color priors of the remote sensing image.According to the statistical results,it is found that theclear-sky and non-fog remote sensing image meets the dark-color statistics priori,that is,the fog removal of the high-resolution remote sensing image is based on the dark-color statistics priori.4.Defogging the high-resolution satellite image using three methods of defogging,and evaluating the processed image by setting four image quality assessment parameters: entropy,energy,standard deviation,and average gradient,and using defogging Before and after the true color image,false color image to compare the image before and after the change of fog;at the same time with the adjacent phase of the clear sky fog-free image contrast detection algorithm defogging effect.5.Design and design a set of indicators based on NDVI to quantitatively analyze the defogging effect of the improved dark channel defogging algorithm.According to the results,the remote sensing data after defogging processing using the improved dark channel defogging algorithm can be applied to quantitative remote sensing applications,improving the availability and effectiveness of remote sensing data.Comparing the three methods of defogging,it is found that for high-resolution satellite remote sensing images,the homomorphic filtering algorithm defogging in the dense fog has poor defogging effect,and some detailed information is lost,and the edges are blurred;The method does not remove fog information from the image.It only uses long-wavelength near-infrared radiation to penetrate the atmosphere.It has a stronger ability to weaken fog in the image information and achieve the purpose of improving image quality.The improved method is based on the dark original color.The defogging side of the inspection can complete the defogging work on the satellite remote sensing image and improve the visual effect of the satellite remote sensing image data.The image becomes clearer and the ground objects are more easily identified,so that previously unavailable remote sensing image data containing fog can be applied.Improve the effectiveness of remote sensing image data content.