As a kind of low-dimensional visual structural feature, edge helps to highlight the basic information of the image, playing the key role in the pre-processing of subsequent advanced visual tasks. Edge detection models with VGG16 as the basic framework can achieve excellent performance through transfer learning, but such models suffer from problems such as large number of parameters and high computational costs. To address the challenge of the coexistence of accuracy and lightweight, an edge detection network inspired by the characteristics of biological visual receptive fields (BVRF-Net) was proposed in the paper. In the Global Pathway, the ON/OFF type convolution kernel was constructed by simulating the ganglionic ON/OFF centroid type receptive field, initializing the convolution kernel by pixel difference value. Meanwhile, simulating the sparse suppression property of complex neurons, the sparse surround suppression convolution kernel with center periphery adjustment was constructed to suppress the texture noise. In the Local Pathway, the asymmetric orientation selectivity convolution kernel was constructed by simulating the orientation selectivity and asymmetric surround suppression characteristics of primary visual cortex neurons, which helps to quickly optimize the orientation features and enhance the contrast features by giving the convolution kernel a priori knowledge, achieving the precise extraction of local detail features. Taking BSDS500 dataset, NYUD-v2 dataset and Multicue dataset as experimental objects, BVRF-Net can achieve competitive results with only 0.358M parameters required. The excellent performance proves the effectiveness of incorporating bio-vision characteristics to construct neural network models, promoting the development of biomimetic computational vision.

Although existing 3D super-resolution methods for magnetic resonance imaging (MRI) volumetric data can provide better visual images than some traditional 2D methods, they should face challenge of increasing network's parameters and computing cost for getting higher reconstruction accuracy. To address this issue, a lightweight 3D multi scale distillation volumetric Transformer, named Transformer-based dual-attention feature distillation (TDAFD) network, is proposed for 3D MRI by utilizing 3D information hiding in images sufficiently. Our TDAFD network contains several proposed dual-attention feature distillation (DAFD) modules and two designed recursive volumetric Transformers (RVT). Concretely, the proposed DAFD module contains a multi-scale feature distillation (MSFD) block for extracting global features under different scales and a feature enhancement dual attention block (FEDAB) for concentrating on the key features better. In addition, our RVT develops 2D Transformer to 3D and save network's parameters via recursion operations for capturing long-term dependencies in volumetric images effectively. Therefore, our proposed TDAFD network can not only extract deeper features via multi scale feature distillation and Transformer, but also realize the balance of performances and network's parameters. Extensive experiments illustrate that our proposed method achieves superior reconstruction performances than some popular 3D MRI SR methods, and saves number of weights and FLOPs.

Abstract Discriminative correlation filter is one of the quick and effective ways for studying visual tracking. However, discriminative correlation filter‐based methods still suffer from many challenging questions caused by environmental interferences, such as spatial boundary effect, temporal filter degradation, and tracking drift. A novel appearance optimisation model, named spatial and long–short temporal attention model, has been proposed based on a new spatial regularisation term and a long–short temporal regularisation term for learning the correlation filter to localise the target. On the one hand, our proposed method can improve the classical spatial regularisation term with a new weight matrix to alleviate the spatial boundary effect. On the other hand, two new temporal regularisation terms are designed: a short temporal regularisation term and a long temporal regularisation term. The short temporal regularisation term can enlarge the inner connections of the current frame and all foregoing frames to improve the tracking performances, and the long temporal regularisation term can address the influence of occlusion by using the similarity between the initial filter and the current one. Extensive experiments on various benchmarks illustrate that our proposed tracker performs favourably against several related popular trackers.

Abstract Recently, there have been many visual tracking methods based on correlation filters. These methods mainly enhance the tracking performances by considering the information of background, space, or time in the appearance model. This paper proposes an effective tracking method, named adaptive spatial–temporal regularized correlation filter (ASTRCF) tracker, based on the popular adaptive spatially regularized correlation filter (ASRCF) tracker. That is, the continuity of object's motion in the process of tracking is considered by introducing a temporal‐regularized term in the appearance model of ASRCF tracker. Furthermore, its solution is inferred by applying the alternating direction method of multipliers. The proposed appearance model contains a background‐awareness term, a spatially regularized term, an adaptive‐weight term, and a temporal‐regularized term. Therefore, it can not only keep the good performances of ASRCF tracker, such as learning the background information and the spatial information adaptively to enhance the discriminating ability, but also take advantage of the relation of correlation filters in the last frame and the current frame for addressing the complex cases, such as occlusion, and fast motion. Extensive experimental results on various challenging databases show that the proposed ASTRCF tracker achieves better tracking performances than some state‐of‐the‐art trackers.

This study proposes a novel super‐resolution regularisation model based on adaptive sparse representation and self‐learning frameworks. The fidelity term in the model ensures that the reconstructed image is consistent with the observation image. The adaptive sparsity regularisation term constrains the reconstructed image with an adaptive sparse representation, which successfully harmonises the sparse representation and the collaborative representation adaptively via producing suitable coefficients. To construct a more effective dictionary, the high‐frequency features from the underlying image patches are extracted, and the dictionary learning and sparse representation are integrated. To this end, the alternating minimisation algorithm is used to divide this model into three subproblems, and the alternating direction method of multipliers and iterative back‐projection method are used to solve the subproblems. To illustrate the effectiveness of the proposed method, additional experiments are conducted on some generic images. Compared with some state‐of‐the‐art algorithms, the experimental results demonstrate that the proposed method achieves better results in terms of both visual quality and noise immunity.

Image super‐resolution (SR) reconstruction, which gains high‐pixel and multi‐detail image from single or several low‐pixel images, has attracted increasing interest in recent years. This study proposes a new SR method based on sparse representation, which made good use of the non‐local (NL) structure similarity and edge sharpness dictionary. Firstly, all the training patches are classified into different clusters according to diverse edge sharpness of patches. Secondly, different dictionaries are trained for different training patches in each cluster. Thirdly, the NL structure similarity is added into the constraint of NL structure similarity model, and the suitable dictionary is selected for current patch to achieve the coefficients according to the value of edge sharpness of patch. Finally, the high‐resolution (HR) image is obtained by integrating HR patches obtained by the product of HR dictionaries and coefficients. Moreover, by calculating edge sharpness, the different dictionaries which adapt to patches with different structure are obtained, and the NL similarity is well utilised and more details are added to HR patch. Compared to some classical and common methods, the proposed method possesses better reconstruction effects in numerical and visual aspects.

Segmentation of white blood cells (WBCs) image is meaningful but challenging due to the complex internal characteristics of the cells and external factors, such as illumination and different microscopic views. This paper addresses two problems of the segmentation: WBC location and subimage segmentation. To locate WBCs, a method that uses multiple windows obtained by scoring multiscale cues to extract a rectangular region is proposed. In this manner, the location window not only covers the whole WBC completely, but also achieves adaptive adjustment. In the subimage segmentation, the subimages preprocessed from the location window with a replace procedure are taken as initialization, and the GrabCut algorithm based on dilation is iteratively run to obtain more precise results. The proposed algorithm is extensively evaluated using a CellaVision dataset as well as a more challenging Jiashan dataset. Compared with the existing methods, the proposed algorithm is not only concise, but also can produce high-quality segmentations. The results demonstrate that the proposed algorithm consistently outperforms other location and segmentation methods, yielding higher recall and better precision rates.

Previous studies have shown that image patches can be well represented as a sparse linear combination of elements from an appropriately selected over-complete dictionary. Recently, single-image super-resolution (SISR) via sparse representation using blurred and downsampled low-resolution images has attracted increasing interest, where the aim is to obtain the coefficients for sparse representation by solving an l0 or l1 norm optimization problem. The l0 optimization is a nonconvex and NP-hard problem, while the l1 optimization usually requires many more measurements and presents new challenges even when the image is the usual size, so we propose a new approach for SISR recovery based on regularization nonconvex optimization. The proposed approach is potentially a powerful method for recovering SISR via sparse representations, and it can yield a sparser solution than the l1 regularization method. We also consider the best choice for lp regularization with all p in (0, 1), where we propose a scheme that adaptively selects the norm value for each image patch. In addition, we provide a method for estimating the best value of the regularization parameter λ adaptively, and we discuss an alternate iteration method for selecting p and λ . We perform experiments, which demonstrates that the proposed regularization nonconvex optimization method can outperform the convex optimization method and generate higher quality images.

As big data acquisition and storage becomes increasingly affordable, especially in the modern range sensing technology for the scans of complex objects, it is a challenge to reconstruct the surface of 3D geometric model effectively and precisely. In this paper, we describe a reduction method for the big data with noises in the 3D surface reconstruction based on partition of unity, Hermite radial basis functions, and sparse regularization. The proposed method not only provides an approach for pruning some redundant data according to the sparsity, but also contains a good robustness to the noises. This approach can be regarded as one of effective methods for processing big data. Experimental results are also provided.