Browsing by Author "Biswas, Debalina"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Image Dehazing using Contrast Limited Adaptive Histogram Equalization and Guided Filter
    (Department of Computer Science & Engineering, NITS, 2019) Biswas, Debalina
    Due to the presence of haze in the atmosphere, the contrast as well as the true color of the real image taken in the outside, gets degraded. Therefore, it is much important to remove hazy effect from the real image to get clear or haze free image. So, here an image dehazing technique, based on combining Contrast Limited Adaptive Histogram Equalization (CLAHE) and Guided filter has been presented. Before applying the proposed algorithm, white balancing is applied to the input image, which helps in adjusting the actual color of the image with respect to white color, to make the output image look more natural. Experiments are carried out on different hazy images, collected from HazeRD dataset, having different haze levels like 50, 100, 200 etc. Results obtained shows that the proposed method can dehaze the hazy image quite effectively and can restore the contrast of the degraded image as well. Experiments are also performed for selecting the desired value of the clip limit for the proposed algorithm. Graphs have been plotted for the results obtained from the experiment, which shows why a particular clip limit is chosen in the proposed algorithm. The purpose of the presented work is to dehaze any hazy input image and get a clear image. Images affected by haze degrades the quality of the image which can create problem in many cases like vision-based physical security system, surveillance, in some medical cases etc. This can also lead to accidents in air, on sea and on the road. So, to avoid such problems and for improving the appearance of the degraded image, dehazing method should be adopted.