Two-Pass Adaptive Histogram based method for Restoration of Foggy Images

In this paper, we present a Local Histogram Equalization (LHE) based method to improve the visual quality of foggy images. From the fog-degraded image we separate all three color components and apply the enhancement algorithm to each of these components. To mitigate the issue of blocking artifacts generated from Local Histogram Equalization with non-overlapping blocks, we propose to process blocks by overlapping them. Processing smaller block size yields better visual quality at the expense of relatively higher processing time, therefore, we also propose to use statistically optimal block size to obtain a good visual quality in reasonable time duration.