Year: 2024
Type: article
Abstract: This paper presents a deep learning method for image dehazing and clarification.The main advantages of the method are high computational speed and usingupaired image data for training. The method adapts the Zero-DCE approach for the image dehazing problem and uses high-order curves to adjust the dynamicrange of images and achieve dehazing. Training thhe proposed dehazing neuralnetwork does not require paired hazy and clear datasets but instead utilizes a setof loss functions, assessing the quality of dehazed images to drive the trainingprocess. Experiments on a large number of real world hazy images demonstratethat our proposed network effectively removes haze while preserving details andenhancing brightness. Furthermore, on an affordable GPU-equipped laptop, theprocessing speed can reach more than 1000 FPS for images with 2K resolution,making it highly suitable for real-time dehazing applications.
Source: Machine Vision and Applications
Authors: Hongyi Qin, Alexander G. Belyaev
Institutions: Imperial College London, Heriot-Watt University
Language: English
Topic: Image Enhancement Techniques
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
Estimated APC paid: $3,390
OpenAlex ID: W4402013460