Reliable Image Dehazing by NeRF

by   Zheyan Jin, et al.
Zhejiang University

We present an image dehazing algorithm with high quality, wide application, and no data training or prior needed. We analyze the defects of the original dehazing model, and propose a new and reliable dehazing reconstruction and dehazing model based on the combination of optical scattering model and computer graphics lighting rendering model. Based on the new haze model and the images obtained by the cameras, we can reconstruct the three-dimensional space, accurately calculate the objects and haze in the space, and use the transparency relationship of haze to perform accurate haze removal. To obtain a 3D simulation dataset we used the Unreal 5 computer graphics rendering engine. In order to obtain real shot data in different scenes, we used fog generators, array cameras, mobile phones, underwater cameras and drones to obtain haze data. We use formula derivation, simulation data set and real shot data set result experimental results to prove the feasibility of the new method. Compared with various other methods, we are far ahead in terms of calculation indicators (4 dB higher quality average scene), color remains more natural, and the algorithm is more robust in different scenarios and best in the subjective perception.


page 1

page 4

page 5

page 6

page 7

page 9

page 10

page 11


An Approach for Noise Removal on Depth Images

Image based rendering is a fundamental problem in computer vision and gr...

Subjective Quality Assessment for Images Generated by Computer Graphics

With the development of rendering techniques, computer graphics generate...

Deep Graphics Encoder for Real-Time Video Makeup Synthesis from Example

While makeup virtual-try-on is now widespread, parametrizing a computer ...

Hair Color Digitization through Imaging and Deep Inverse Graphics

Hair appearance is a complex phenomenon due to hair geometry and how the...

Simulation of Plenoptic Cameras

Plenoptic cameras enable the capturing of spatial as well as angular col...

SeaThru-NeRF: Neural Radiance Fields in Scattering Media

Research on neural radiance fields (NeRFs) for novel view generation is ...

The Theory of Unified Relativity for a Biovielectroluminescence Phenomenon via Fly's Visual and Imaging System

The elucidation upon fly's neuronal patterns as a link to computer graph...

Please sign up or login with your details

Forgot password? Click here to reset