Improved Multiple Fusion Techniques Based Single Image Dehazing

Abstract

In view of the unfavorable climate circumstances like the presence fog or overwhelming rain,
computerized pictures are simply put through to a broad variety of assortment during
acquisition, which may diminishes visual effect & affects post-processing of the picture.
Pictures debased by mist adversely influence the class of vision-based physical security
system. This prompts to accidents in air, on sea & on the road. In any case, for enhancing the
picture appearance to a human viewer, or to change a picture to a format more suitable to
machine processing, improving techniques ought to be utilized. After improvement is again
re-established the improved picture via restoration techniques. Here, we enhance our image
by applying propose method provide us better results when contrasted with base techniques.
The output image is clearer in compare of base paper image and hollow effect is also
reduced. In our propose algo in view of another, non-local prior. The algorithm depends on
the suspicion that shades of a fog-free picture are well approximated via a pair of hundred
particular colors that frame tight groups in RGB space. In this algorithm we use low pass
filter & band pass filter for smoothing the image. Calculate the PSNR and Time.