AI RESEARCH
Multi-Branch Non-Homogeneous Image Dehazing via Concentration Partitioning and Image Fusion
arXiv CS.CV
•
ArXi:2605.00885v1 Announce Type: new Existing single image dehazing methods have nstrated satisfactory performance on homogeneous thin-haze images; however, they often struggle with non-homogeneous hazy images that exhibit spatially varying haze concentrations and abrupt density transitions across different regions.