Author(s)

Ling Gan, Ziwen Xiong

Corresponding Author

Ling Gan

ABSTRACT

The existing single image dehazing algorithm is mostly based on local prior, and the defogging results have problems such as block effect and color distortion. In response to the above problems, an improved algorithm is proposed. Firstly, the non-local prior is used to estimate the initial transmission. Then the L1 regularization method is used to optimize the transmission. The domain transform filter is used to further optimize the transmission to achieve smooth image and suppress noise interference. Simulation results have demonstrated that the improved algorithm can guarantee the readability of a hazed image after removing noise, and the color and the details near the observer in the dehazing image are better than that achieved by the primal method.

KEYWORDS

Image dehazing, Non-local dehazing, L1 regularization, domain transform