Restoring high-quality images from degraded hazy observations is a fundamental and essential task in the field of computer vision. While deep models have achieved significant success with synthetic data, their effectiveness in real-world scenarios remains uncertain. To improve adaptability in real-world environments, we construct an entirely new computational framework by making efforts from three key aspects: imaging perspective, structural modules, and training strategies. To simulate the often-overlooked multiple degradation attributes found in real-world hazy images, we develop a new hazy imaging model that encapsulates multiple degraded factors, assisting in bridging the domain gap between synthetic and real-world image spaces. In contrast to existing approaches that primarily address the inverse imaging process, we design a new dehazing network following the "localization-and-removal" pipeline. The degradation localization module aims to assist in network capture discriminative haze-related feature information, and the degradation removal module focuses on eliminating dependencies between features by learning a weighting matrix of training samples, thereby avoiding spurious correlations of extracted features in existing deep methods. We also define a new Gaussian perceptual contrastive loss to further constrain the network to update in the direction of the natural dehazing. Regarding multiple full/no-reference image quality indicators and subjective visual effects on challenging RTTS, URHI, and Fattal real hazy datasets, the proposed method has superior performance and is better than the current state-of-the-art methods.

In complex natural haze scenes, image haze removal still faces significant challenges in removing non-homogeneous and dense haze. The double complexity of haze distribution, on the one hand, is reflected in the interference of haze to the global image information, and on the other hand, it is reflected in the imbalance of image brightness and color caused by random haze distribution. In natural scenes with prominent edge and texture features, the above problems may cause severe degradation of image quality and performance of various tasks. Numerous studies on network learning show that the effect of haze removal is closely related to haze feature expression. Therefore, to improve the performance of dehazing, this paper proposes a sliding self-attention wavelet network. Specifically, we first design a sliding self-attention module to identify haze regions in images and capture rich haze-related feature information. Then, considering the uneven distribution of haze in images, discrete wavelet transform (DWT) and inverse transform (IDWT) are used for constructing a hierarchical encoder-decoder structure, which can fully use the multi-resolution characteristics of DWT, locally decompose feature maps of different scales, extract low and high-frequency information, and then gradually recover sharp edges and precise texture details from hazy images. Finally, to enable the proposed network to generate more realistic haze-free images on different complex haze scenes, we develop a DWT-based adversarial loss function to constrain the low and high-frequency components of generated images closer to the corresponding clear images. Experimental results on the relevant public benchmark datasets show that the proposed algorithm achieves favorable dehazing performance.

Haze can cause significant changes in the data domain of the image. Due to domain-shift problem, the model trained on the synthetic haze images has a weak or even invalid dehazing ef-fect on the real data. This paper proposes an unsupervised domain adaptation dehazing method based on the nearest-farthest subspace distance in response to this problem. For the middle convolved features, the matrix singular value de-composition is used to obtain the basis vectors of the support subspace. Narrowing the angle between the basis vectors of natural and synthetic haze domains could reduce the differ-ence between the data domains. In subspace, a newly defined distance with a new penalty term named subspace measure distance is employed to constrain the model. Furthermore, to avoid search blindness of the proposed method in the real data training phase, we propose the nearest-farthest subspace distance inspired by contrastive learning. In addition, we use a new training strategy. Finally, the experimental tests on the synthetic and real images prove the effectiveness of the pro-posed method.