Abstract
In this paper, we present an extremely computation-efficient model called FAOD-Net for dehazing single image. FAOD-Net is based on a streamlined architecture that uses depthwise separable convolutions to build lightweight deep neural networks. Moreover, the pyramid pooling module is added in FAOD-Net to aggregate the context information of different regions of the image, thereby improving the ability of the network model to obtain the global information of the foggy image. To get the best FAOD-Net, we use the RESIDE training set to train our proposed model. In addition, we have carried out extensive experiments on the RESIDE test set. We use full-reference and no-reference image quality evaluation indicators to measure the effect of dehazing. Experimental results show that the proposed algorithm has satisfactory results in terms of defogging quality and speed.
Highlights
Many cities are shrouded in smog due to waste incineration, construction dust, and automobile exhaust
Deep learning techniques have been widely used in the field of image processing, such as image classification, object recognition, and face recognition. e third type is that the existing image dehazing algorithm in studies such as [7,8,9] based on deep learning mostly estimates the transmittance of foggy images through neural network model, estimates the atmospheric light value separately, and obtains the fog-free image according to the atmospheric scattering model
The proposed FAOD-Net is tested with both qualitative and quantitative analysis. e full-reference image quality evaluation indicators PSNR and SSIM and noreference image quality evaluation indicators spectral entropy-based quality (SSEQ) and BLIINDS-II are considered for the quantitative analysis
Summary
Many cities are shrouded in smog due to waste incineration, construction dust, and automobile exhaust. E third type is that the existing image dehazing algorithm in studies such as [7,8,9] based on deep learning mostly estimates the transmittance of foggy images through neural network model, estimates the atmospheric light value separately, and obtains the fog-free image according to the atmospheric scattering model. The pyramid pooling module is mainly included in [14, 15] He et al proposed SPPnet [14], which solves the problem that the input of the deep convolutional neural network must require a fixed image size and improve the efficiency of extracting features. For the real-world foggy test set, we use the no-reference image quality assessment indicators spatial-spectral entropy-based quality (SSEQ) [17] and blind image integrity notator using DCT statistics (BLIINDS-II) [18] to measure the dehazing effect
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.