Title

Technical Papers Session III: A novel image dehazing and assessment method

Abstract/Description

Images captured in hazy weather conditions often suffer from color contrast and color fidelity. This degradation is represented by transmission map which represents the amount of attenuation and airlight which represents the color of additive noise. In this paper, we have proposed a method to estimate the transmission map using haze levels instead of airlight color since there are some ambiguities in estimation of airlight. Qualitative and quantitative results of proposed method show competitiveness of the method given. In addition we have proposed two metrics which are based on statistics of natural outdoor images for assessment of haze removal algorithms.

Location

Room C9 (Aman Tower, 3rd floor)

Session Theme

Technical Papers Session III - Computer Vision

Session Type

Parallel Technical Session

Session Chair

Dr. Asim Ur Rehman Khan

Start Date

16-11-2019 3:10 PM

End Date

16-11-2019 3:30 PM

Share

COinS
 
Nov 16th, 3:10 PM Nov 16th, 3:30 PM

Technical Papers Session III: A novel image dehazing and assessment method

Room C9 (Aman Tower, 3rd floor)

Images captured in hazy weather conditions often suffer from color contrast and color fidelity. This degradation is represented by transmission map which represents the amount of attenuation and airlight which represents the color of additive noise. In this paper, we have proposed a method to estimate the transmission map using haze levels instead of airlight color since there are some ambiguities in estimation of airlight. Qualitative and quantitative results of proposed method show competitiveness of the method given. In addition we have proposed two metrics which are based on statistics of natural outdoor images for assessment of haze removal algorithms.