Premium
P‐4.1: A Hardware Optimized Dehazing Algorithm for Video
Author(s) -
Jiawen Chen,
Limin Yan
Publication year - 2021
Publication title -
sid symposium digest of technical papers
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.351
H-Index - 44
eISSN - 2168-0159
pISSN - 0097-966X
DOI - 10.1002/sdtp.14529
Subject(s) - computer science , flicker , inter frame , algorithm , computational complexity theory , channel (broadcasting) , frame rate , frame (networking) , artificial intelligence , computer vision , computer graphics (images) , reference frame , telecommunications , computer network
The dehazing technology has made great progress on a single image. But for the problem that the time complexity is high and cannot meet the real‐time requirement of HD video dehazing, a dark channel prior dehazing algorithm based on hardware architecture optimization is proposed in this paper. The hardwareoriented two‐scale joint filtering method is used to simplify the computational complexity of transmittance and atmospheric light values, while using interframe dependence constraints to suppress flicker noise in video dehazing. The experimental results show that the proposed algorithm is equivalent to the current mainstream dehazing algorithm under software and hardware platforms, even brighter and coordinated in the detail color, and the atmospheric light value of the algorithm is basically stable, indicating the proposed interframe dependency constraints can effectively adjust atmospheric light values.