A Single Image Dehazing Method Based on Dark Channel Prior

Abstract

The invention discloses a dark channel prior based single-image defogging method, for the purposes of improving the defogging effect of a conventional dark channel prior algorithm and improving the algorithm efficiency. The method comprises the following steps: solving a dark channel graph Idark comprising a fog image I; performing adaptive threshold segmentation on the dark channel graph to obtain a binary graph Ybinary of a sky area and a non-sky area; re-estimating an atmosphere optical value A by use of the segmented sky area; roughly estimating transmissivity of the segmented sky area and the non-sky area by use of different modes, then merging the estimated transmissivity of the sky area and the non-sky area by use of the binary graph Ybinary to obtain rough transmissivity t(x,y)<~>; refining the transmissivity by use of guided filtering; obtaining a restored de-fogged image J by use of an atmosphere scattering model; and further enhancing edge information of the restored image J by use of the guided filtering, and obtaining a finally restored image Jfinary. The defogging algorithm provided by the invention can effectively improve the degeneration phenomenon of foggy-day images, improves the image sharpness and also obviously improves the processing efficiency.

Description

technical field [0001] The invention relates to the technical field of image processing, in particular to a single image defogging method based on dark channel prior. Background technique [0002] At present, the processing methods for foggy images are mainly divided into two categories: foggy image restoration and foggy image enhancement. The enhancement method of foggy image does not consider the essential cause of image degradation, but is based on subjective visual experience, which can effectively improve the contrast of foggy image, highlight the details of the image, improve the visual effect of the image, and has a wide range of applications. information may cause certain losses. The image restoration method in foggy weather establishes an image degradation model based on the influence mechanism of foggy weather on image imaging, deduces the image restoration model inversely, and calculates the optimal estimated value not affected by fog to obtain a clear image. In...

AI Technical Summary

Problems solved by technology

[0003] The deficiencies of the dark channel prior dehazing method: 1) The traditional dark channel prior algorithm uses soft matting to refine the transmittance. Since it uses high-precision interpolation to solve large sparse matrices, it brings a huge amount of calculation, which is not conducive to engineering. 2) The dark channel prior algorithm is based on the dark channel assumption. For bright areas that do not meet this assumption, the estimated transmittance of the algorithm is too small, and the restored image appears color distortion and overall dark phenomenon.

Images