FPD Net: Feature Pyramid DehazeNet

Abstract

We propose an end-to-end dehazing model based on deep learning (CNN network) and uses the dehazing model re-proposed by AOD-Net based on the atmospheric scattering model for dehazing. Compare to the previously proposed dehazing network, the dehazing model proposed in this paper make use of the FPN network structure in the field of target detection, and uses five feature maps of different sizes to better obtain features of different proportions and different sub-regions. A large amount of experimental data proves that the dehazing model proposed in this paper is superior to previous dehazing technologies in terms of PSNR, SSIM, and subjective visual quality. In addition, it achieved a good performance in speed by using EfficientNet B0 as a feature extractor. We find that only using high-level semantic features can not effectively obtain all the information in the image. The FPN structure used in this paper can effectively integrate the high-level semantics and the low-level semantics, and can better take into account the global and local features. The five feature maps with different sizes are not simply weighted and fused. In order to keep all their information, we put them all together and get the final features through decode layers. At the same time, we have done a comparative experiment between ResNet with FPN and EfficientNet with BiFPN. It is proved that EfficientNet with BiFPN can obtain image features more efficiently. Therefore, EfficientNet with BiFPN is chosen as our network feature extraction.

Advancements in Remote Sensing Image Dehazing: Introducing URA-Net with Multi-Scale Dense Feature Fusion Clusters and Gated Jump Connection

“…To circumvent the constraints associated with estimated parameters, researchers have harnessed the formidable learning capabilities of neural networks, ushering in a new wave of direct repair dehazing algorithms [64][65][66][67][68]. This new approach shows high robustness and generalization ability when dealing with various types and degrees of haze images [69][70][71][72].…”

CLGA Net: Cross Layer Gated Attention Network for Image Dehazing

“…The MSDBN-DFF [18] network derives from U-Net [30], and uses back projection technology to realize the effective integration of multi-layer features in space and reduce the loss of low-level information. The FPD network [12] applies the FPN network structure in the field of object detection for dehazing, which can effectively integrate high-level and low-level semantics. Zhang et al [31] combine the dehazing algorithm with an iterative fine-matching algorithm derived from motion structure to perform 3D reconstruction of dehazing images to improve the accuracy and accuracy of dehazing.…”

“…With the rise of deep learning, researchers have put efforts on neural networks to restore fog-free images, including residual learning ResNet [7,8], DehazeNet [9], multi-scale convolutional neural networks (CNN) [10], and pyramid dehazing network [11][12][13]. Compared with the traditional methods, deep learning methods try to directly retrieve the transmission image or the final fog-free image for end-to-end defogging, using a large number of training data sets.…”