Single image dehazing using transmission estimation

Abstract

Image dehazing is a system which can reduce poor consequence of haze on pictures which increase competence of the picture/tape meting out method in the indistinct climate. The present study, presents a simple image dehazing technique. System deletes the initial transmittal, exactly depends on hidden area division which in turn reduces the initial transmittal. First, we build raw transmittal in Gaussian pyramid using selective transmittal extraction prior. Second, a patch based on Laplacian pyramid which gives the value of nonlinear re-trace function point by point. Third, is universal Laplacian map which states the segmentation-based re-traced image. All the three segmentation guide the performer for transmittal refinement. Finally, the dehazing output gets back from the refined transmittal and atmospheric scattered technique.