원문정보
초록
영어
Images which are captured in outdoor scenes often degrade in quality due to atmospheric conditions induced by suspending particles like mist, haze, fog, etc. which scatter and absorb light before it reaches to the camera. The weather-degraded images directly affect the efficiency of computerized surveillance and monitoring system. In this paper, we propose an efficient haze removal technique which integrates dark channel prior with an ant colony optimization algorithm. Firstly, we use the dark channel prior method to compute dark channel in the hazy image and secondly, ant colony optimization algorithm is applied to make restoration factor for hazy images, adaptive between 0 to 1. The restoration factor in all the previous work is fixed value i.e. 0.1. Experimental-results show that proposed algorithm is effective, robust and is yielding high-contrast images.
목차
1. Introduction
2. Mathematical Model of Hazy Image
3. Proposed Method
3.1 Dark Channel Prior
3.2 Transmission Estimation
3.3 Estimating Atmospheric Light Value Based On Variogram
3.4 Recovering the Scene Radiance by Using ACO
4. Experimental Results
5. Qualitative Evaluation
6. Conclusion
References