Abstract Pulmonary lobe segmentation is an important prerequisite for accurately quantifying pulmonary damage in many pulmonary diseases and planning treatment. However, due to the incomplete lobar structures and morphological changes caused by diseases, the lobe segmentation still encounters great challenges. In this study, a Learnable Interpolation and Extrapolation Network (LIE‐Net) is proposed to form complete and consecutive fissure surfaces by learning to extract information of the fissures from existing fissure points and absent points (unsegmented points belonging to fissures) to predict the z coordinate of the absent fissure points. The completed pulmonary fissures are further used for accurate pulmonary lobe segmentation. Specifically, LIE‐Net takes the coordinate information of existing fissure points (their ( x , y , z ) coordinates) and absent fissure points (their ( x , y ) coordinates) as two independent inputs, and predicts the z coordinates of absent points. The proposed LIE‐Net makes voxel‐wise predictions based on the spatial structure characteristics of the lung fissure, and is able to provide a consecutive fissure surface in space. According to the evaluation of radiologists, the lobe segmentation performance was remarkably enhanced in approximately 76% of patients in our additional dataset after the application of LIE‐Net, especially for those cases with large‐scale missing fissures.

Image downscaling has been a classical problem and has recently been linked to super-resolution (SR). In this paper, we aim to propose a learning-based image downscaling model, FastDownscaler, which can efficiently produce low-resolution (LR) images that not only preserve the rich details of the original high-resolution images but also be highly restorable for existing SR models. We first present two separate lightweight networks with different upsampling losses, the bilinear loss and the bicubic loss, which are better for SR restoration and LR downscaling, respectively. To produce versatile LR images, we then propose to distill bilinear loss guided network with bicubic loss guided one. To our best knowledge, we establish the first image downscaling quality assessment dataset to evaluate the downscaling performance. Experimental results demonstrate the superior performance of the proposed model on image downscaling and SR. Furthermore, our model can achieve over 600 FPS for downscaling a$1920\times 1280$image.

Recently, researchers have shown that coverless steganography is relatively safe. On this basis, to improve the payload of the coverless steganography, a novel semiconstruction coverless steganography algorithm is introduced in the paper. Firstly, web crawler technology is applied to crawl a wide range of small icons and hot news images from the Internet. These icons can be used as the training subset, and the hot news can be designed according to construction rules. Secondly, the Alex-Net network is introduced for training in the algorithm, and the adversarial samples are added to the training set. Thirdly, using the preset template, certain small icons and a hot news image are spliced into a secret carrier image according to the construction principle. The hot news image is in the top half of the carrier, and those small icons are in the bottom half. The image on the upper part of the carrier and the icons of the lower part can be connected by image and text semantics, and the semantic matching can be realized between image semantics and explanatory. The experimental results and analysis show that the proposed algorithm can resist steganalysis tools effectively and has good robustness against various image attacks. Meanwhile, the secret information payload has been greatly improved, the maximum payload can reach 180 bits of a single 512 × 512 image. This promising algorithm can be applied to build covert communications.

Synthesizing high dynamic range (HDR) images from multiple low-dynamic range (LDR) exposures in dynamic scenes is challenging. There are two major problems caused by the large motions of foreground objects. One is the severe misalignment among the LDR images. The other is the missing content due to the over-/under-saturated regions caused by the moving objects, which may not be easily compensated for by the multiple LDR exposures. Thus, it requires the HDR generation model to be able to properly fuse the LDR images and restore the missing details without introducing artifacts. To address these two problems, we propose in this paper a novel GAN-based model, HDR-GAN, for synthesizing HDR images from multi-exposed LDR images. To our best knowledge, this work is the first GAN-based approach for fusing multi-exposed LDR images for HDR reconstruction. By incorporating adversarial learning, our method is able to produce faithful information in the regions with missing content. In addition, we also propose a novel generator network, with a reference-based residual merging block for aligning large object motions in the feature domain, and a deep HDR supervision scheme for eliminating artifacts of the reconstructed HDR images. Experimental results demonstrate that our model achieves state-of-the-art reconstruction performance over the prior HDR methods on diverse scenes.

State‐of‐the‐art optimisation methods for salient object detection neglect that saliency maps of different images usually show different imperfections. Therefore, the saliency maps of some images cannot achieve effective optimisation. Based on the observation that the saliency maps of semantically similar images usually show similar imperfections, the authors propose an optimisation method for salient object detection based on semantic‐aware clustering and conditional random field (CRF), named CCRF. They first cluster the training images into some clusters using the image semantic features extracted by using a deep convolutional neural network model for image classification. Then for each cluster, they use a CRF to optimise the saliency maps generated by existing salient object detection methods. A grid search method is used to compute the optimal weights of the kernels of the CRF. The saliency maps of the testing images are optimised by the corresponding CRFs with the optimal weights. The experimental results with 13 typical salient object detection methods on four datasets show that the proposed CCRF algorithm can effectively improve the results of a variety of image salient object detection methods and outperforms the compared optimisation methods.

In order to improve the security and robustness of the Information Steganography Algorithm under strictly controlled environment, a new algorithm of modification-free steganography based on image and big data is introduced in this paper. In the proposed algorithm, a mapping relationship between the hot image entropy and the secret information is constructed and the payload information is expressed by the mapping relation. At the same time, turbo code is introduced in order to improve robustness, the hot image comes from Internet image big data, and the library of hot image is established. The performance of the proposed algorithm is analyzed using simulation experiment. Because of its none-modifying on carrier image, the results of experiment show that the proposed algorithm can achieve good performance in robustness analysis, dimension scaling attack, and rotation attack. In particular, in the test of dimension scaling attack and the rotation attack, the rate of data recovering can be over 95%. The proposed algorithm can be very valuable in the covert communication which requires high security and low volume, for example, the key exchange of symmetric encryption system.