Search

Superpixel segmentation aims at dividing the input image into some representative regions containing pixels with similar and consistent intrinsic properties, without any prior knowledge about the shape and size of each superpixel. In this paper, to alleviate the limitation of superpixel segmentation applied in practical industrial tasks that detailed boundaries are difficult to be kept, we regard each representative region with independent semantic information as a subspace, and correspondingly formulate superpixel segmentation as a subspace clustering problem to preserve more detailed content boundaries. We show that a simple integration of superpixel segmentation with the conventional subspace clustering does not effectively work due to the spatial correlation of the pixels within a superpixel, which may lead to boundary confusion and segmentation error when the correlation is ignored. Consequently, we devise a spatial regularization and propose a novel convex locality-constrained subspace clustering model that is able to constrain the spatial adjacent pixels with similar attributes to be clustered into a superpixel and generate the content-aware superpixels with more detailed boundaries. Finally, the proposed model is solved by an efficient alternating direction method of multipliers (ADMM) solver. Experiments on different standard datasets demonstrate that the proposed method achieves superior performance both quantitatively and qualitatively compared with some state-of-the-art methods., Accepted by IEEE Transactions on Industrial Informatics, 2020

Stereo superpixel segmentation aims to obtain the superpixel segmentation results of the left and right views more cooperatively and consistently, rather than simply performing independent segmentation directly. Thus, the correspondence between two views should be reasonably modeled and fully considered. In this paper, we propose a left-right interactive optimization framework for stereo superpixel segmentation. Considering the disparity in stereo image pairs, we first divide the images into paired region and non-paired region, and propose a collaborative optimization scheme to coordinately refine the matched superpixels of the left and right views in an interactive manner. This is, to the best of our knowledge, the first attempt to generate stereo superpixels considering the parallax consistency. Quantitative and qualitative experiments demonstrate that the proposed framework achieves superior performance in terms of consistency and accuracy compared with single-image superpixel segmentation.

Based on Fully Convolutional Networks, recent salient object detection (SOD) methods achieve impressive results. Some studies improve the classical SOD frameworks by utilizing auxiliary information like fixation points, salient numbers and salient edges. These works adopt auxiliary information by embedding sub-networks into the main network. However, how to incorporate auxiliary information regardless of the specific structure with less coupling is unexplored in SOD. In this paper, we present DANet, a new Dynamic network to leverage arbitrary Auxiliary information for SOD. The proposed framework consists of 1) a Dynamic Weight Generator (DWG) which converts arbitrary auxiliary features into dynamic weights, 2) a Dynamic Bridge Block (DBB) which uses dynamic weight convolution to incorporate auxiliary information from DWG and then refines the fused features, and 3) a two-step training strategy to alleviate the side effect caused by drastic changes among different input images. Extensive experiments demonstrate the effect of different auxiliary information and the proposed framework is a universal method to improve SOD with auxiliary information. Comparison experiments show that DANet achieves state-of-the-art performance without any pre-processing and post-processing.

Objective The aim of this study was to investigate the differences in six anthropometric measurements of people born during and immediately after the 1959 to 1961 Great Chinese Famine using a regression discontinuity approach. Methods Data were drawn from the baseline of the China Kadoorie Biobank study, and a subset of data from 76,912 participants was analyzed. We performed regression discontinuity among participants who were born during the famine (October 1959 to October 1962) and immediately after the famine period (November 1962 to October 1964) by using local linear and parametric regressions. All analyses were conducted by sex and study area. Results Significantly, there were increases of 0.30 kg/m2 (P = 0.007) in BMI, 0.81 kg (P = 0.028) in weight, 8.57 mm (P = 0.004) in waist circumference, and 5.07 mm (P = 0.004) in hip circumference for rural women who were exposed to famine during their fetal period compared with those who were not exposed to famine in utero. However, such statistically significant increases in anthropometric values were not observed in local linear regression and most parametric models among rural men or in the urban population. Conclusions Rural Chinese women who were exposed to famine during the fetal period were observed to have higher levels of BMI, weight, waist circumference, and hip circumference in adulthood.

Superpixel segmentation aims at grouping discretizing pixels into high-level correlative units and reducing the complexity of subsequent tasks, e.g., saliency detection and object tracking. Existing superpixel segmentation algorithms mainly focus on maintaining the geometrical information, while neglecting the irregular structure of superpixels. In this paper, a superpixel segmentation method is proposed to generate approximately structural superpixels with sharp boundary adherence and comprehensive semantic information. The superpixel segmentation is formulated as a square-wise asymmetric partition problem, where the semantic perceptual superpixels are recorded in a square level to preserve abundant semantic information and save storage simultaneously. Moreover, in order to achieve regular-shape superpixel units to better adhere to image boundaries and contours, a combinatorial optimization strategy is devised to achieve an optimal combination of squares and isolated pixels. Experimental comparisons with some state-of-the-art superpixel segmentation methods on the public benchmarks demonstrate the effectiveness of the proposed method quantitatively and qualitatively. In addition, we have applied the method to brain tissue segmentation to illustrate superior performance.

Background Disease leaf segmentation in color image is used to explore the disease shape and lesion regions. It is of great significance for pathological diagnosis and pathological research. Objective This paper proposes a superpixel algorithm using Non-symmetry and Anti-packing Model with Squares (NAMS) for color image segmentation of leaf disease. Methods First of all, the NAMS model is presented for color leaf disease image representation. The model can segment images asymmetrically and preserve the characteristics of image context. Second, NAMS based superpixel (NAMS superpixel) algorithm is proposed for clustering pixels, which can represent large homogeneous areas by super squares. By this way, the impact of complex background and the data redundancy in image segmentation can be reduced. Results Experimental results indicate that compared with segmenting the original image directly and manipulating by Simple Linear Iterative Clustering (SLIC) superpixel, the proposed NAMS superpixel performs more excellently in not only saving storage but also adhering to the lesion region edge. Conclusions The outcome of NAMS superpixel can be regarded as a preprocess procedure for leaf disease region detection since the method can segment the image into superpixel blocks and preserve the lesion area.

In this paper, a novel concept factorization (CF) method, called CF with adaptive neighbors (CFANs), is proposed. The idea of CFAN is to integrate an ANs regularization constraint into the CF decomposition. The goal of CFAN is to extract the representation space that maintains geometrical neighborhood structure of the data. Similar to the existing graph-regularized CF, CFAN builds a neighbor graph weights matrix. The key difference is that the CFAN performs dimensionality reduction and finds the neighbor graph weights matrix simultaneously. An efficient algorithm is also derived to solve the proposed problem. We apply the proposed method to the problem of document clustering on the 20 Newsgroups, Reuters-21578, and TDT2 document data sets. Our experiments demonstrate the effectiveness of the method.

Image segmentation plays a crucial role in various biomedical applications. In general, the segmentation of brain Magnetic Resonance (MR) images is mainly used to represent the image with several homogeneous regions instead of pixels for surgical analyzing and planning. This paper proposes a new approach for segmenting MR brain images by using pseudo-color based segmentation with Non-symmetry and Anti-packing Model with Squares (NAMS). First of all, the NAMS model is presented. The model can represent the image with sub-patterns to keep the image content and largely reduce the data redundancy. Second, the key idea is proposed that convert the original gray-scale brain MR image into a pseudo-colored image and then segment the pseudo-colored image with NAMS model. The pseudo-colored image can enhance the color contrast in different tissues in brain MR images, which can improve the precision of segmentation as well as directly visual perceptional distinction. Experimental results indicate that compared with other brain MR image segmentation methods, the proposed NAMS based pseudo-color segmentation method performs more excellent in not only segmenting precisely but also saving storage.

A novel image sparse representation method, named NAMlet transform, is proposed by fully considering the structure of the nature images.The NAMlet transform is based on the asymmetric homogeneous blocks, which could preserve the detail information of the images.The NAMlet transform can remove the restrictions of the size of the images. Image sparse representation methods have been widely applied in many image processing fields, such as computer vision, image de-noising, super resolution, and visual tracking. An efficient sparse representation method can improve the accuracy. However, few of the traditional representation methods consider from the point of the anti-packing problem. Thus, these methods are not only restricted by the size of the image, but also lose a great amount of detail information by using a symmetric blocking method. In this paper, we have proposed an image sparse representation method, called NAMlet Transform. The NAMlets are haar-type wavelets, which are based on the non-symmetric homogeneous blocks obtained by the non-symmetry and anti-packing model. In homogeneous blocks, all the pixels are in the same bit-plane. The NAMlet transform can reduce the lost detail information and remove the restrictions of image size. The experiment results show the strong superiority of the NAMlet transform for image representation in comparison with some state-of-the-art image sparse representation methods.

Saliency modeling has become one of the most popular studies in computer vision. Many previous works adopted distinctness to compute saliency score of an image element, which usually need point-to-point distances calculation and it is quadratic complexity. In this paper, a visual saliency model based on one time sampling outlier detection is proposed, and the time complexity is linear to image size, further analyses and experiments demonstrate that our model is robust and efficient. This model is parameter insensitive, without learning, and easy to implement. Extensive experiments on four benchmark datasets show that our model is competitive compare with state-of-the-art models under shuffled AUC metric.

This letter proposes a bottom-up saliency model to predict eye fixation locations. Unlike traditional models that measure saliency by computing local or global distinctness, the proposed model considers saliency as the prediction error, because we believe that image patches or pixels with higher prediction error are more salient than others. The prediction error consists of both mispredicted error and unpredicted error. We propose a new algorithm called sparse residual to compute the mispredicted error. We then adopt outlier detection to compute the unpredicted error. Finally, we obtain the saliency map from merging the two results together via a guided filter. Extensive experiments on three benchmark databases show that our model is superior to 12 state-of-the-art models.

This paper makes an in-depth analysis of the characteristics, the motivation and the intervention effect of government intervention in executive compensation in state-owned enterprises (SOEs). The research shows that the local government always makes and implements differentiated compensation regulation policies according to the marketization degree, the income gap, the financial deficit level, the unemployment rate and the GDP growth rate. However, from the perspective of intervention effect, such government intervention in executive compensation weakens the compensation performance sensitivity of SOEs executives, effects the future performance of Chinese SOEs, and it has a greater negative impact on competitive SOEs. The conclusion indicates that the government should improve the pertinence and validity of government behavior and try to establish differentiated executive’s compensation incentive mechanism that matches the governance behavior of modern SOEs with Chinese characteristics.

The goal of salient region detection is to identify the regions of an image that attract the most attention. Many methods have achieved state-of-the-art performance levels on this task. Recently, salient instance segmentation has become an even more challenging task than traditional salient region detection; however, few of the existing methods have concentrated on this underexplored problem. Unlike the existing methods, which usually employ object proposals to roughly count and locate object instances, our method applies salient objects subitizing to predict an accurate number of instances for salient instance segmentation. In this paper, we propose a multitask densely connected neural network (MDNN) to segment salient instances in an image. In contrast to existing approaches, our framework is proposal-free and category-independent. The MDNN contains two parallel branches: the first is a densely connected subitizing network (DSN) used for subitizing prediction; the second is a densely connected fully convolutional network (DFCN) used for salient region detection. The MDNN simultaneously outputs saliency maps and salient object subitizing. Then, an adaptive deep feature-based spectral clustering operation segments the salient regions into instances based on the subitizing and saliency maps. The experimental results on both salient region detection and salient instance segmentation datasets demonstrate the satisfactory performance of our framework. Notably, its APr@0.5 and Apr@0.7 reaches 73.46% and 60.14% in the salient instance dataset, substantially higher than the results achieved by the state-of-the-art algorithm.

The binary feature descriptors such as binary robust independent elementary features (BRIEF), oriented rotated binary robust independent elementary features (ORB), and fast retina keypoint (FREAK) usually perform binarisation on the intensity comparisons, thus they lose some useful information. In this study, the authors propose an effective binary image descriptor which is called significant bit-planes-based local binary pattern for visual recognition. First, the authors divide an image into several sub regions according to the intensity orders to incorporate the spatial information. Then the authors extract the higher bit planes for all the sub regions and sort the adjacent neighbour bits based on the corresponding intensity orders, which make the descriptor invariant to rotation. In order to further improve the discriminative ability, the authors sample the multi-scale neighbours and average the adjacent pixels and extract the feature descriptor from the higher bit planes. Since the authors directly perform operation on the significant bit planes without quantisation, the authors decrease the information loss to some extent. The descriptor has demonstrated a better performance over the state-of-the-art binary descriptors as well as scale invariant feature transform on two recognition benchmarks (i.e. Kentucky and ETHZ) and PASCAL 2007 for image classification.

Saliency modeling has played an important part in computer vision studies over the past 30 years. Many state-of-the-art models adopted complex mathematical and machine learning theories. In this paper, a simple and effective visual attention model is proposed. We find that a single fixed template is enough for saliency map generation; this idea is inspired by the receptive field of the human visual system. All that is needed is to convolve the input image with this template with additional post-processing. Experiments show that our model is extremely fast and performs better than state-of-the-art models in human eye fixation prediction.

Designing efficient and effective keypoint descriptors for an image plays a vital role in many computer vision tasks. The traditional binary descriptors such as local binary pattern and its variants directly perform a binarization operation on the intensity differences of the local affine covariant regions, thus their performance usually drops a lot because of the limited distinctiveness. In this paper, we propose a novel image keypoint descriptor, namely local derivative quantized binary pattern for object recognition. To incorporate the spatial information, we first divide the local affine covariant region into several subregions according to the intensity orders. For each sub region, we quantize the intensity differences between the central pixels and their neighbors in an adaptive way, and then we order the differences and use a hash function to map the differences into binary codes. The binary codes are histogramed to form the feature of each subregion. Furthermore, we utilize multi-scale support regions and pool the histograms together to represent the features of the image. Our approach does not need prior codebook training and hence it is more flexible and easy to be implemented. Moreover, our descriptor can preserve more local brightness and edge information than the traditional binary descriptors. Also, our descriptor is robust to rotation, illumination variations and other geometric transformations. Finally we conduct extensive experiments on three challenging datasets (i.e., 53 Objects, ZuBuD, and Kentucky) for object recognition and the experimental results show that our descriptor outperforms the existing state-of-the-art descriptors.

In this paper, a novel label propagation (LP) method is presented, called the manifold adaptive label propagation (MALP) method, which is to extend original LP by integrating sparse representation constraint into regularization framework of LP method. Similar to most LP, first of all, MALP also finds graph edges from given data and gives weights to the graph edges. Our goal is to find graph weights matrix adaptively. The key advantage of our approach is that MALP simultaneously finds graph weights matrix and predicts the label of unlabeled data. This paper also derives efficient algorithm to solve the proposed problem. Extensions of our MALP in kernel space and robust version are presented. The proposed method has been applied to the problem of semi-supervised face clustering using the well-known ORL, Yale, extended YaleB, and PIE datasets. Our experimental evaluations show the effectiveness of our method.

Multi-frame super-resolution image reconstruction aims to restore a high-resolution image by fusing a set of low-resolution images. The low-resolution images are usually subject to some degradation, such as warping, blurring, down-sampling, or noising, which causes substantial information loss in the low-resolution images, especially in the texture regions. The missing information is not well estimated using existing traditional methods. In this paper, having analyzed the observation model describing the degradation process starting with a high-resolution image and moving to the low-resolution images, we propose a more reasonable observation model that integrates the missing information into the super-resolution reconstruction. Our approach is fully formulated in a Bayesian framework using the Kullback---Leibler divergence. In this way, the missing information is estimated simultaneously with the high-resolution image, motion parameters, and hyper-parameters. Our proposed estimation of the missing information improves the quality of the reconstructed image. Experimental results presented in this paper show improved performance compared with that of existing traditional methods.

In this paper, we propose a novel graph-based semisupervised learning framework, called joint sparse representation and embedding propagation learning (JSREPL). The idea of JSREPL is to join EPL with sparse representation to perform label propagation. Like most of graph-based semisupervised propagation learning algorithms, JSREPL also constructs weights graph matrix from given data. Different from classical approaches which build weights graph matrix and estimate the labels of unlabeled data in sequence, JSREPL simultaneously builds weights graph matrix and estimates the labels of unlabeled data. We also propose an efficient algorithm to solve the proposed problem. The proposed method is applied to the problem of semisupervised image clustering using the ORL, Yale, PIE, and YaleB data sets. Our experiments demonstrate the effectiveness of our proposed algorithm.

Display Omitted The TV model is combined with H1 model to form the lmix model.The lmix model can not only remove the noise, but also preserve the edges.An lmix based prior model for super resolution reconstruction is proposed.The convergence and the efficiency of the proposed method are also analyzed. Multi-frame Super-Reconstruction (SR) is a technique for reconstructing a High-Resolution (HR) image by fusing a set of low-resolution images of the same scene. One of the most difficult problems in SR is to preserve the edges while removing the noise. Therefore, in this paper we propose a novel l mix model, which combines the total variation model and the H 1 model by using a pair of different weighting parameters. Also, a hierarchical Bayesian framework is used and the weighting parameters can be modeled together with the HR image and other parameters. Thus the weighting parameters are updated according to the global features of the HR image in iterations. In this way, the proposed l mix model can not only preserve the edges but also it can remove the noise. Our experimental results show that the proposed method has much improvement over the existing methods.

A high-resolution image is obtained by fusing the information derived from blurred, sub-pixel shifted, and noisy low-resolution observations. In this paper, a novel regularization model based on an Anisotropic Fractional Order Adaptive (AFOA) norm is proposed and then we apply the AFOA model into the Super-Resolution Reconstruction technology. Compared with the existing models, the proposed AFOA model can remove the noise and protect the edges adaptively according to the local features of the images. Meanwhile, the proposed AFOA model can avoid the staircase effect effectively in the smooth region. To obtain the solution to the proposed AFOA model, the Gradient Descent Method is used in this paper. Finally, the experimental results show that the proposed method has much improvement than the existing methods in the respect of the Peak Signal-to-Noise Ratio and the visual quality.

In this paper, a non-symmetry and anti-packing image representation model (NAM) has been proposed. NAM is a hierarchical image representation method and it aims to provide faster operations and less storage requirement. By taking a rectangle sub-pattern, for example, we describe the idea of NAM and its encoding algorithm. In addition, an approach for adaptive area histogram equalization for image contrast enhancement based on a NAM image is presented. The contrast enhancement approach is designed to meet the NAM image representation and it can be duplicated with faster operation. The complexity analysis and the experimental results show that the NAM based algorithm for image contrast enhancement is faster and more effective than that based on matrix image.